WineHQ

Final Fantasy XI Online

Final Fantasy XI is the first online Final Fantasy RPG from Square Enix.

Application Details:

Version: Final Fantasy XI
License: Retail
URL: http://www.playonline.com/ff11...
Votes: 414
Latest Rating: Silver
Latest Wine Version Tested: 9.6

Maintainers: About Maintainership

Link Square Enix account managment Free Download Installation Media (EU) Free Download Installation Media (US) Free Download Installation Media (JP)

Test Results

Old test results
The test results you have selected are very old and may not represent the current state of Wine.
Selected Test Results

What works

PlayOnline

FFXI Game

What does not

Workarounds

What was not tested

High end raiding content

Hardware tested

Graphics:

  • GPU: AMD
  • Driver: open source

Additional Comments

selected in Test Results table below
Operating systemTest dateWine versionInstalls?Runs?Used
Workaround?
RatingSubmitter
ShowGentoo LinuxApr 05 20249.6Yes Yes NoSilverChiitoo 
ShowArch Linux x86_64Jun 20 20238.10Yes Yes NoPlatinumPaul C. 
CurrentArch Linux x86_64Dec 14 20228.0-rc1Yes Yes NoPlatinumPaul C. 
ShowGentoo Linux x86_64Nov 15 20227.20Yes Yes NoPlatinumPandoraxero 
ShowUbuntu 22.04 "Jammy" (+ variants like Kubuntu)Nov 15 20227.20Yes Yes YesGoldPandoraxero 

Known Bugs

Bug # Description Status Resolution Other apps affected
28861 Final Fantasy XI hangs after character selection NEW View
32931 FFXI Amazon Downloader crashes when trying to download install files NEW View
40794 PlayOnline Viewer crashes when using a game controller. UNCONFIRMED View
41641 Final Fantasy XI: incorrect face culling; no glitches when disabled UNCONFIRMED View
45520 Install Shield fails on 64bit prefix with Silent install UNCONFIRMED View
47577 Final Fantasy XI crashes UNCONFIRMED View
52989 Final Fantasy XI Setup window is not rendered properly and appears stuck (msiexec) UNCONFIRMED View
53009 Final Fantasy XI Config has some text rendered/converted incorrectly UNCONFIRMED View
53711 Final Fantasy XI crashes randomly at multiple points UNCONFIRMED View
55007 Multiple applications fail to run if Wine is compiled with CFLAGS="-march=znver4" (PlayOnline Viewer, Steam). RESOLVED NOTOURBUG View
56447 PlayOnline Viewer: The "splash screen" is blank/grey without its image. UNCONFIRMED View
56448 Final Fantasy XI Online: No application icon in taskbar/panel. UNCONFIRMED View
56462 PlayOnline Viewer crashes when Wine is built without gstreamer support UNCONFIRMED View
56530 Final Fantasy XI Online: Memory leak when Wine is built with CFLAGS="-g -mno-avx". UNCONFIRMED View
56531 Final Fantasy XI Online: Some textures are transparent, malformed, or misplaced. UNCONFIRMED View
56532 Final Fantasy XI Online: Unexpected Scroll Lock behaviour. UNCONFIRMED View
56533 Final Fantasy XI Online: Incorrect/corrupt textures shown on models. UNCONFIRMED View
57207 Fallout 3: Regression spams console with errors UNCONFIRMED View
57418 PlayOnline Viewer throws an application error at launch. UNCONFIRMED View

Show all bugs

HowTo / Notes

HOWTO

Prelude

Regarding installation of PlayOnline Viewer and Final Fantasy XI, everything should be working out of the box at the time of writing (2024-08), though some runtime issues, that do not prevent actual use of the software, do exist (see the list of open bugs).

The information that follows is kept for historical purposes, and in case regression or new issues appear where the old information might be handy.

DirectX

It is not necessary to install DirectX via the installer, which it does offer.  If for some reason it is needed, it should be reported as a bug.

FFXI ON Linux (Ubuntu)

updated 7/19/20 by Christopher Strom - Kbtaru

in progress

Install Wine

apt install wine

Installing FFXI

Unzip the ffxi .zip file that was downloaded from the official website above, browse to that folder within the terminal window and enter

wine ffxisetup.exe

Follow normal installation.

FFXI ON MAC OS X

updated 7/19/20 by Christopher Strom - Kbtaru


***FFXI Setup files can be downloaded free from Official Website


Getting Starting 

You will need to install Java Development Package, Xcode, X11 and Homebrew


Installing Wine

Open a terminal window and enter 

brew install wine

Installing FFXI

Unzip the ffxi .zip file that was downloaded from the official website above, browse to that folder within the terminal window and enter

wine ffxisetup.exe

Follow normal installation.

Updating FFXI Settings

In the terminal, type

regedit

Navigate to

HKEY_LOCAL_MACHINE\SOFTWARE\PlayOnlineUS\SquareEnix\FinalFantasyXI

(Replace xx with your region, Ex EU/JP/US)


The following is a quick reference for the values found in the above
registry key:
Value Description
0000 MIP mapping
0001 Screen resolution width (in pixels)
0002 Screen resolution height (in pixels)
0003 Background resolution width (in pixels)
0004 Background resolution height (in pixels)
0007 Music and Sound effects (0=Off, 1=On)
0011 Environmental Animation (0=Off, 1=Normal, 2=Smooth)
0017 Bump Mapping (0=Off, 1=On)
0018 Texture Compression (0=High, 1=Low, 2=Uncompressed)
0019 On-Screen Maps (0=Compressed, 1=Uncompressed)
0022 Intro Movie (0=Disabled, 1=Enabled)
0034 Window Mode (0=Full Screen, 1=Window Mode) 

You must update registers 1-4

Registers 0001 and 0002 are your display window size. These should be smaller than your screen resolution. ex Resolution of 1920x1080 could set to 1280x1024.


Registers 0003 and 0004 are background resolution, these need to be set to a square ex 2048x2048.


Gamepad Settings

in progress


Even Older Archive

updated 7/19/20 by Christopher Strom - Kbtaru

Installation

The PlayOnline Viewer and Final Fantasy XI installation/data files can be downloaded from the PlayOnline website(s):

During the installation process of Final Fantasy XI, a window with only borders, and no rendered content may be seen, and the 'msiexec' process can be using up to 100% of a single CPU core while possibly processing the 'FINAL_FANTASY_XI.msi' file probably. This can take up to at least four (4) minutes, but should eventually complete.

Last updated until this point: 2019-06-14

Gameplay

Windows 7 needs to be set as the reported Windows version, or the game client will crash after accepting the license agreement.

Everything imaginable in-game works, though the game can be quite heavy, depending on settings (especially due to the shadows).  Using CSMT from Staging helps with FPS by allowing more CPU to be used.

Gamepad

A PS2-controller via a simple USB-converter works flawlessly (aside from sometimes when entering the game, and the analog mode is off, the player character and the camera will start moving by themselves, but they stop when the analog mode is enabled (whether or not this would happen on Windows as well is unknown at this time)).

Japanese Input

Requires the Japanese PlayOnline Viewer software, as is intended (using 'uim' to switch input methods).

Works while in PlayOnline Viewer, but not while in-game.  This may be due to how things are built-in.  It is possible to paste text from outside the PlayOnline Viewer into it, but the same is not true for the actual game client.

Known Issues

  • The PlayOnline Viewer splash image is "transparent" (works with Wine Staging).
  • The opening movie does not play correctly, but the audio can be heard (for about 10 seconds, then it's skipped).  This seems to be regression (bug 42008).
  • Sometimes, usually only after having gone back to the PlayOnline Viewer from the game, the application will crash when re-launching the game from the viewer (backtraces speak of the (proprietary) nvidia-drivers).
  • The PlayOnline Viewer graphics are not updated when moved to a secondary monitor, looking like a hang (audio still plays, and it is possible to interact with buttons and such).  The actual game client doesn't have this issue.
  • Using the Japanese client, typing Japanese in-game doesn't work via 'uim', probably because the thing is heavily built-in (text can't be 'pasted in' even).
  • Using the Japanese client, having the opening movie enabled will crash the client when launched from the PlayOnline Viewer. 

The following information may be largely out of date since 2016-12-19 or earlier.

HOWTO - Final Fantasy XI: The Ultimate Collection 2011

This HOWTO will help you get Final Fantasy XI setup on your machine using "The Ultimate Collection 2011" retail package, or equivalent media. Here you will also find a guide to get the PlayOnline Viewer running. This HOWTO is complete to the best knowledge of its creators, however, there may be some errors that have not been resolved.

Please note that in order to run this game, you will most likely NEED the proprietary drivers for your graphics card(s). Open-source drivers are known to have problems with 3D rendering, and may render many things incorrectly or slowly. Please bear this in mind when submitting test results, as well.


Installation

  • Insert "The Ultimate Collection 2011" dvd into your dvd drive.
  • In a terminal session, type in "wine /path/to/dvdrom/PlayOnline/setup.exe".
  • From here, you will be guided by a setup wizard. Choosing all defaults is acceptable.
  • After the PlayOnline Viewer is finished installing, click finish on the wizard if you haven't done so already.
  • Type in "cd/path/to/dvdrom/INST1" to go into the install directory for Final Fantasy XI.
  • Type in "wine setup.exe".
  • A wizard will appear that will guide you through setting up Final Fantasy XI. Again, choosing all defaults is acceptable.
  • After installation is complete for Final Fantasy XI, you may be given a choice to restart your machine. Choose "No, I will restart my machine later" and click finish.
  • From here, type in "cd/path/to/dvdrom/INST2".
  • Type in "wine setup.exe".
  • Follow the wizard for setting up Rise of the Zilart, choosing all defaults, and selecting to restart your computer later, if prompted.
  • Type in "cd/path/to/dvdrom/INST3".
  • Type in "wine setup.exe".
  • Follow the wizard for setting up Chains of Promathia, choosing all defaults, and selecting to restart your computer later, if prompted.
  • Type in "cd/path/to/dvdrom/INST4".
  • Type in "wine setup.exe".
  • Follow the wizard for setting up Treasures of AhtUrhgan, choosing all defaults, and selecting to restart your computer later if prompted.
  • Type in "cd/path/to/dvdrom/INST5".
  • Type in "wine setup.exe".
  • Follow the wizard for setting up Wings of the Goddess, choosing all defaults, and selecting to restart your computer later if prompted.
  • You may now umount your installation media. Be sure to store it in a safe place.

Overriding DXDIAGN.DLL

  • Obtain a copy of the file "Dxdiagn.dll" from a Windows XP install (It'll be in the System32 Directory)
  • Copy said file into your WINE installation (.wine), place it either in the Windows folder or the System32 folder, either is fine.
  • Open a terminal and navigate to the directory you placed Dxdiagn.dll into.
  • Run the following command: "wine regsvr32 dxdiagn.dll". Additionally try running "wine regsvr dxdiagn.dll".
  • Open winecfg and navigate to the overrides.
  • Set a new override for Dxdiagn.
  • You should be set to go.

Configuring/Updating/Running

  • Here the PlayOnline Viewer will be updated. Follow the instructions to update the viewer.
  • After PlayOnline Viewer is updated, the viewer will ask you to click next to restart it. Do so, and PlayOnline Viewer Should restart.
  • Once logged in, you may run Final Fantasy XI. It will be updated upon running them for the first time, this is the perfect opportunity to eat some cheesecake, because the initial update could take hours. Do note that if you have registered the registration codes for any of the Abyssea expansions it will be automatically downloaded and installed in the update.
  • Once the update has completed, Final Fantasy XI should launch with your desired settings.
  • Go to a heavily populated area, and type "/yell Final Fantasy XI Runs In Linux!!" into Final Fantasy, and press Enter

Running more than one instance of FFXI:

  • In a shell do:
    WINEPREFIX="/home/$USER/.wine-second" wine /path/to/first/full/FFXI/Install/pol.exe
  • This will create a new instance directory for wine in your home folder. It will be named .wine-second
  • You will get a few errors. You can dismiss all of them.
  • Use your favourite file handling utility and copy the complete PlayOnline folder from your previous (/home/yourname/.wine/drive_c/Program Files/) wine instance to your new instance. (Yes you will need two installs of FFXI for this to work)
  • Make sure you copy your original install from it's root folder in the Program Files directory to the Program Files directory of your new wine instance. The folder named PlayOnline is all you need to copy.
  • Now you should have one original install in:
    /home/yourname/.wine/drive_c/Program Files/PlayOnline/
    and one copy of that install in:
    /home/yourname/.wine-second/drive_c/Program Files/PlayOnline/
    
  • In a shell do:
    regedit
    
  • Now make sure that the root of the registry is selected in the tree structure on the left hand side. In English the root is called 'My Computer'
  • Now click the following sequence in the Regedit menu bar:
    'Registry -> Export registry file'
  • Select a location to save your registry export. (Suggest saving to desktop for easy removal after you are done)
  • Close the registry editor.
  • In a shell do:
    WINEPREFIX="/home/$USER/.wine-second" regedit
    
  • You are now editing the registry information for your new instance of wine.
  • Once again; select the root of the registry (My Computer). Then click the following sequence in the registry editor:
    'Registry -> Import registry file'
  • Select the file you saved in the previous regedit instance. (The one I suggested you save to your desktop)
  • Once the file has been imported... (might take a while depending on size of registry) you are ready to start using two instances of FFXI at the same time on the same computer.
  • You may now delete the .reg file that you saved to your desktop.
  • To start the first instance of FFXI use:
    wine /path/to/first/full/FFXI/Install/pol.exe
    
  • To start the second instance of FFXI use:
    WINEPREFIX="/home/$USER/.wine-second" wine /path/to/copy/of/full/FFXI/Install/pol.exe
  • The default path to an install is: /home/username/.wine/drive_c/Program Files/PlayOnline/
  • This guide is crude and in an initial stage. In time it will be >optimised.

Other Recommendations/Info:

  • Disabling shadows altogether causes a speedup.
  • Desktop Effects (Compiz) should be disabled if flickering of the screen occurs.
  • Please use the latest version of wine if possible, if not, there is a list of known working versions at the bottom of this guide.

Christopher Strom

Comments

The following comments are owned by whoever posted them. WineHQ is not responsible for what they say.

Some Updates...
by Blink on Tuesday April 23rd 2024, 5:00
So naturally this whole page needs a massive update as far as the game working on Macs using wine. I'll first confirm that the individual game installer files work great, along with a default directX install provided in the latest complete collection. I have tested on both a quad i5 iMac with AMD Pro graphics, as well as M1 and M2 series macs. There was nothing special needed for the wine install or prefix to installing or running the game. I installed wine through brew, did not change the prefix to 32bit, and simply installed and ran the shortcut (.lnk) the installer put on my mac's desktop via wine/terminal, the game launched with no issues.

This was also tested on the M series using wine as well as crossover, both performed the same as the intel with much faster install and loading times. I did not experience long installation times that others here have said they had during the space calculation when the installer starts, it took a few mins but not that bad. If it does take a while this has to do with .net on the wine side but should continue either way.

First I tested on intel and M series Ventura, then I tested on Sonoma on M2 and can confirm everything works right out of the box on the current versions as of me writing this. Hope this helps :)
RE: Some Updates...
by Chiitoo on Tuesday August 6th 2024, 15:22
Thank you for commenting on the macOS side of things. :]

It would be good to have that information as a test results as well, especially since these comments do get cleaned away from time to time.
Above HOWTO steps no longer necessary
by Paul C. on Sunday December 18th 2022, 20:06
Just chiming in to say that in a default 64bit prefix, the game can be installed by simply invoking the installer.
RE: Above HOWTO steps no longer necessary
by Chiitoo on Monday December 19th 2022, 0:10
Indeed, most of it has been kept around for historical purposes so to speak.

The very top of the notes does suggest no workarounds on the Linux side are needed, and that some extra packages may be needed on the MAC OS X side.

Regarding the settings part, I would say it's better to use the FFXI Config Utility as one normally would.

I'll look into cleaning them up when I find some time for it.

Thanks!
Regarding Installer issues
by Pandoraxero on Wednesday December 14th 2022, 16:39
During install, the installer appears to freeze while installing FFXI itself. This same issue seems to occur on Windows, but it is much quicker to resolve, and certainly does not cause Windows to come up with the "(Not Responding)" thing in the title bar. As far as the EULA not appearing on some distros, I noticed that, on Windows, the EULA uses an unusual font - something like MS Mincho or SimSun are the closest I can find - half-width monospace fonts, and I'm guessing if there's no alias for the font it's looking for - you get rectangles in those distros. If I had it in my mind to give ubuntu another go, I'd check to see if there are any crucial differences in the /usr/share/fonts dir between it and my main Gentoo install. A missing font or setting, perhaps, on ubuntu.
RE: Regarding Installer issues
by Pandoraxero on Wednesday December 14th 2022, 17:18
I should add that, for the install process even on Windows it seems to take a VERY substantial amount of time. All in all, I would say the only issue with the install that Wine itself might be amplifying is the freeze at the start of the FFXI Install... And I think that may be due to what the installer is trying to do at that point: "Computing Space Requirements" - it's trying to determine if there's enough space on the disk for the game. As for how it does that - I don't know. maybe there's some kind of windows system call that gives a program the approximate amount of disk space available, and that call doesn't translate one-for-one in Wine - because it kind of CAN'T. As far as Wine is concerned, there's the odd scenario of C: - where programs usually install to - being additionally a subset of Z: ... I dunno, I'm really just spitballing at this point.
Sound issues - but too many variables in play
by Pandoraxero on Tuesday May 31st 2022, 20:57
Okay, I'm noticing pops on my sound like an old vinyl record. But I'm unable to isolate the cause due to too many factors at play:
I'm running straight ALSA, no pulseaudio
I'm using HDMI out
HDMI is going to an A/V Receiver.

And then there's the fact that it's FFXI... in all its poorly-optimized glory.
Anyone else having these kinds of issues?
RE: Sound issues - but too many variables in play
by Chiitoo on Wednesday June 1st 2022, 1:07
I'm also using plain ALSA, with my motherboard's chip, and can't say I've noticed this here.

I do have the occasional underruns, rare'ish, usually seeing them in the log when I have had the PlayOnline Viewer running in the background and minimised for a time (not hearing anything weird with those either).

I am using a custom '.asoundrc' to create a loopback device for recording audio, but don't really do anything else with it aside from setting 'rate' to '48000'.
keyboard/gamepad issue.
by Pandoraxero on Saturday May 14th 2022, 23:41
Alright, that's it. I'm done for today. I got absolutely NO testing done, but I did find out something interesting.

The way my kernel is set up, my Logitech Y-BN52 Keyboard will generate 3 event devices. One of them reads as a joystick, and since it was plugged directly into my PC it registered before my actual gamepad.

It just so happens FFXI and FFXIPadConfig.exe are both, well... kinda dumb. They just look for the first gamepad device they can find, and only poll that device for input.

As a result, I have been chasing down an issue with my gamepad for over 8 hours... that was literally an issue with this keyboard.

Tomorrow, I'm going to see if it does the same thing in Windows, AND see if there's something I can disable in the Kernel so it doesn't do that.
RE: keyboard/gamepad issue.
by Chiitoo on Sunday May 15th 2022, 8:19
That does seem quite annoying. >.<

I wonder if you could do something about it via the Game Controllers config in the Wine Control Panel?
RE: keyboard/gamepad issue.
by Pandoraxero on Sunday May 15th 2022, 11:31
huh. I didn't think to look there, but, yeah, disabling the device in the Game Controllers config seems to have solved the issue.

It doesn't appear to be an issue at all in Windows, so I'm just left to wonder exactly what is causing this device to register as a joystick device - probably just a hyperactive udev config or something.

No matter, though. I mean, that's one of three keyboards I have connected, which may sound crazy, but one's for the couch, one's for up close, and one's wireless. The wireless one is sometimes not reliable.

This is basically a couch gaming setup ... I even installed every emulator I could. the TV's 75 inches, sitting 16 feet away, which is about the same as a 24 inch monitor at 3 or 4 feet. I run at 1080p so I can see what I'm doing from the couch - if that helps game performance, then, well, so much the better, I suppose.
Testing Variety
by Pandoraxero on Saturday May 14th 2022, 16:36
Looking at the testing history on this AppDB entry, I've done a breakdown of our testing variety, broken down by category (due to rounding errors, percentages may not add up perfectly)
Total: 158 (100%)
-Debian+variants: 95 (60.13%)
- Ubuntu: 84 (53.16%)
- Debian Proper: 7 (4.43%)
- Linux Mint: 3 (1.90%)
- Knoppix: 1 - That's me! :) (0.63%)
-Gentoo+variants: 34 (21.52%)
- Gentoo Proper: 32 (20.25%)
- Sabayon: 1 (0.63%)
- Funtoo: 1 (0.63%)
-RPM-based: 10 (6.33%)
- SUSE/openSUSE: 5 (3.16%)
- Fedora: 4 (2.53%)
- Mandriva (which I'm pretty sure is dead now): 1 (0.63%)
-Arch: 9 (All Arch Proper) (5.70)
-BSD: 4 (2.53%)
- FreeBSD: 3 (1.90)
- PC-BSD: 1 (0.63%)
-MacOS: 4 (2.53%)
-Slackware: 1 - Me Again! :) (Slackware Proper) (0.63%)

And last but not least, the award for most certifiably insane tester goes to.... Demo, for testing this on Linux From Scratch on Oct 29, 2009 ... look, I've tried 6 different distros from this list, and even I'm not THAT crazy. (0.63%)

Mind you, this is based on the currently-available test results, dating all the way back to 16 years ago. It seems Ubuntu is in the lead by a large margin - no surprise, there - and Gentoo is trailing quite far behind. RPM distros are narrowly edging out Arch 10 to 9, and BSD in general is tied with MacOS at 4 entries
In relation to my previous post, I think we could use some testing on one of the new Apple Silicon Macs. My pockets aren't deep enough to get one which might be able to run FFXI, though. ...That and I'm not a fan of Apple by any stretch of definition - I might consider getting an Apple computer if they abandoned the Lightning port on their phones.

We need testing outside of Linux in general, whether that testing is in MacOS, BSD, or perhaps something more crazy like illumos (assuming wine will even run on it). The simple fact is we have a proven working product on Linux, with very few bugs... which is honestly a level of functionality I would have thought impossible 14 years ago. There's only two ways to push the envelope at this point, and of the two, only one is in scope for this AppDB entry: Testing on UNIX. Whether the UNIX testing is taking place on something based on a BSD- or SysV-based.

I suppose, additionally, testing could take place for how well the game works with, say, Wayland vs. X11, or if there's any difference between audio quality and/or response between straight ALSA, and other intermediaries like pulseaudio or pipewire. Testing such as this could play an important role not only in ensuring this game works as it should, but also that wine is up to speed with these newer technologies... whether they'll catch on as systemd seems to have or fizzle out like compiz did is anyone's guess at this point.
HOWTO Maintenance
by Pandoraxero on Friday May 13th 2022, 20:57
The HOWTO seems to be suffering from an abundance of - for lack of a better word - cruft. A lot of bloat from outdated info that can be easily removed.

It can be reasonably assumed that the "Archive Below" section is proper cruft.
I am also uncertain precisely how much of the Registry hacking part is strictly necessary given how the FFXI config works these days.
Given Apple's move to ARM-based silicon, I am uncertain how much longer the Apple part will be workable, but I do not own any Apple devices on which to test.

Basically we can assume that, given current instructions, people will be installing with the current - or, at the very least, a recent - download.
I've still not taken the time to test JP input. I assume, perhaps correctly, that it doesn't work, but I cannot say for certain. Regardless, even on Windows, JP input requires either JP POL (which, to the best of my knowledge, DOES work), or Windower (which is out of scope for this appdb entry, and recent tests show that it does NOT work)

The Links at the top of the page work, so linking to the downloads in the HOWTO is sort of redundant, but the links in the HOWTO can stay if need be.

Wine install instructions for Ubuntu and its derivatives should stay, as those tend to be the ones new users gravitate towards.
It can be reasonably assumed that anyone using Gentoo knows how to install wine, and I suspect the same can be said of Arch users.
I am uncertain, based on Apple's move to ARM-based silicon, if this will work on the new Macs, but I will address that in an upcoming post, likely tomorrow.
Basic Testing
by Pandoraxero on Sunday May 1st 2022, 11:50
My basic testing thus far seems quite promising, barring some minor gamepad config issues. I've half a mind to try pushing my luck by testing if I can get JP POL installed and get JP input.
RE: Basic Testing
by Pandoraxero on Sunday May 1st 2022, 20:06
My testing is complete for the time being. I simply cannot justify rating this any less than gold on the platform with which I tested it. If not for the "Space Requirements" bug on install, it would be Platinum. As far as the gamepad oddities, I'm fairly certain that's either a hardware issue, a driver selection issue.... or a "me" issue.
For Thousands of Years I Laid Dormant
by Pandoraxero on Tuesday April 26th 2022, 20:21
I'm back after some time away - and with a brand-new PC to boot.
My specs (no, this is not intended to be a flex):
Case: Fractal Design Node202 (It's scary what I can do even with such a small case)
CPU: AMD Ryzen 5600X (underclocked to 2.8GHz for thermals, might try higher if temps stay reasonable)
GPU: Radeon RX6700XT, open source drivers.
RAM: 64GB DDR4
OS: Gentoo, of course
Wine Ver: 7.7 (vanilla)

I have yet to give the game a proper test, but Install seemed to have a few issues. It froze for somewhere between 2 and 5 minutes (exact time not available. left PC to get coffee.) on "Computing Space Requirements", then popped back to life. Odd, but okay, I guess. The update also ran fine and the opening movie even played. We shall see how she runs this weekend, I guess.

In the meantime, is there any way to get my FPS and thermals on an overlay in the corner of the screen or something?
RE: For Thousands of Years I Laid Dormant
by Chiitoo on Wednesday April 27th 2022, 7:19
Welcome back!

The "hang" during the space requirements computing is a very old issue, and I think I have managed to still not file a bug for it... thanks for the reminder!

You might notice tiny stuttering every second or so (bug filed) as well as random'ish crashes with "GL_OUT_OF_MEMORY in glGetTexImage" and such (discussed this one with a developer a little, but no bug filed yet... waiting for the next crash to happen for that).

For the overlay, I guess there are multiple ways one could achieve that. I've only used WINEDEBUG="fps" myself, so I haven't tested these, but see for example:

linuxhint.com/show_fps_counter_linux_games/
RE: For Thousands of Years I Laid Dormant
by Pandoraxero on Wednesday April 27th 2022, 8:34
Actually, now that I think about it, I was reading something about using 'screen' to monitor one terminal from another. It was useless to me at the time, since I was trying to monitor an emerge from another location, and it requires some setup prior to being used... but that just might be what I'm looking for.
Using that, and a simple script that refreshes 'sensors' every second or so, I should be able to monitor status via SSH.
The thing about building in the Fractal Design Node 202 is it does NOT leave much room for a cooling solution.
I'm running a Noctua NH-LH9a-AM4 Heatsink with a 92x25mm fan attached, and still getting temps in the 80s at CPU base clock.

I need the FPS monitoring for obvious reasons, and the thermals to make sure the thing doesn't explode in the process. Well, "explode" may be a bit hyperbolic, but you get the picture. I need to run some sort of stress test on this system, and FFXI under Wine is perhaps the most practical stress test I can do.
RE: For Thousands of Years I Laid Dormant
by Chiitoo on Wednesday April 27th 2022, 9:32
Yeah, 'screen' is pretty handy. I use it for a lot of things, such as CLI IRC clients which for example allows me to restart the DE without losing the IRC sessions (and well, now that they removed the little scrollback it used to have from the kernel, it helps with doing stuff in CLI only a lot).

For CPU and RAM stress testing, emerging something like qtwebengine works pretty well I think. ^^;

Those temps seem a bit high, yeah, but I'm used to my first generation Ryzen 7 1700, which tends to be very cool I think, or at least it stays quiet. Haven't really been monitoring its temps actually...
RE: For Thousands of Years I Laid Dormant
by Pandoraxero on Saturday April 30th 2022, 19:17
I ended up using a VERY different approach to capture the FPS on a SSH session: Named Pipes.
I use 'mkfifo' to create a named pipe. have wine push all its output to the named pipe, then in an SSH session, direct the pipe to 'cat'
maybe there's a more "elegant" way to do this, but my way is simple, easy to understand, and most importantly, it works.

Shadows on High bottomed out at 16FPS in PJeuno. Shadows on Normal bottomed out at about 23, same location, and that's the one factor that causes the game to choke regardless of system, so aside from installation issues and the gamepad config issue, the game seems to run near-flawlessly. Not sure if it's better software on Wine's end, or better hardware/drivers on my end.

Tomorrow, though, will be the real test, because I have an Echad Ring Charge active, and another ready to pop off. And if I'm not mistaken, I was leveling BLU in Sea.

As far as my temps go, on the test run I just did with the CPU capped at 2800MHz It got up to 62, and ambient is around 27.
Delta-T of 35 isn't bad considering what I'm working with. I think I'll eventually bump it up to 3GHz, but these tests take precedence.
RE: For Thousands of Years I Laid Dormant
by Pandoraxero on Friday April 29th 2022, 20:16
in the event you need more debug output, I might be inclined to do another install - i mean, the issue ONLY pops up on the FFXI installer, so it's not something you'd notice once it's installed, let alone be able to get debug info for in any other situation.

Honestly, I thought something screwed up. Then I went and made my coffee, came back and it was ready. A less patient person might have just given up, but I run Gentoo, and there is no distro - aside from perhaps LFS - which requires so much patience. I have reason to suspect users of Mint - by way of example, not exclusion - might just assume it's all jacked up and just give up.

I'd be curious how well this runs on one of the new M1 Macs, also, but I'm not foolish enough to buy one that would be equal to the task, spec-wise.
RE: For Thousands of Years I Laid Dormant
by Chiitoo on Friday April 29th 2022, 20:53
It's easy to re-produce so no need for more info on the installer issue methinks. Thank you though!

And yes, that happened: there were reports about the installer just hanging and not working, even though it does after a while. I'll try to file a bug about it soon. And yeah, Gentoo definitely is not for the impatient ones. :]

One thing I would like you to check and report back in on, is if you see all the text in the config utility, or if the tabs and some of the drop-downs have ????? shown only?
RE: For Thousands of Years I Laid Dormant
by Pandoraxero on Friday April 29th 2022, 21:31
Well, one thing I noticed on my first attempt at an install (which I was NOT patient with, and I nuked the whole ~/.wine directory as a result) - if you run it in Windows 10 mode, there's a chance the text for the Checkboxes on the installer won't appear until you select them. weird.

Additionally, the POL gamepad config doesn't seem to recognize the second shoulder buttons (R2/L2 equivalent) on my Logitech Dual Action gamepad (as far as USB Gamepads go, it's pretty much the dictionary definition of Legacy PS1/PS2-like gamepads), as far as the FFXI gamepad input I had to use the xinput switch... which was weird. I'll have to 2xcheck my useflags.
Mistake in my own report
by Aurora Robb Kristiansen on Tuesday November 30th 2021, 16:03
Hi there! I submitted a report earlier, and erroneously said status effects don't show their remaining time. However this is my own fault, as it turns out I didn't enable the setting correctly. They totally work now, how can I edit my report, or could a maintainer do it for me? Thanks!
RE: Mistake in my own report
by Chiitoo on Wednesday December 1st 2021, 2:43
Yes, we can edit reports, or you can submit a new one.

In addition to the timers, I was going to send you a message regarding a few others things as well:

- While yes, the installer does seem to hang with a blank window, it does eventually proceed, or at least it still did when I last tested not long ago (I wonder how long it takes on Windows?).
- DXVK, to my knowledge (and after re-testing it just now), still does not seem to do anything with FFXI (probably due to the age of the DirectX bits it uses). How have you tested this?

Finally, since a fellow maintainer already approved the test report, I'm willing to let it be as is, but if I had handled it, I would have asked for the report to be re-submitted without any of the mentions to third-party applications and unofficial servers. I consider these off-topic, as they are in violation of all kinds of terms and/or services of SqEX, and I think of them as akin to cracked/pirated software which go against the AppDB Test Results Guidelines [1].

Furthermore, when they modify the way the game client runs in any way, the test becomes "dirty", and may no longer reflect the actual state of the application. If anything, such applications should have their own AppDB entries, although personally I would rather not see such things here at all.

In any case, thank you for your contribution!

1. wiki.winehq.org/AppDB_Test_Results_Guidelines
RE: Mistake in my own report
by Aurora Robb Kristiansen on Wednesday December 1st 2021, 9:40
Hey Chiito, thanks for the reply! Yeah, this all makes sense, so I'll re-submit my test results the way you described it later tonight. As for DXVK, I first tested it by installing it manually so I could test the game in a Wine prefix that isn't touched by any external tools; later just let Lutris manage it for simplicity's sake. I've not yet tried without DXVK, so I can do that first before submitting the new results.

The install process is much faster and does not "seem to do nothing" on Windows, I can confirm.

Thanks again for the feedback!
RE: Mistake in my own report
by Michael D on Wednesday December 1st 2021, 10:57
Chiitoo is correct. I should have stripped those out of the report entirely. I edited them to remove certain references but... I should not process submissions when I have a sinus infection impairing my cognitive faculties.

I really appreciate that you are testing the client and was excited to see a new result posted. I look forward to seeing more from you!
RE: Mistake in my own report
by Aurora Robb Kristiansen on Wednesday December 1st 2021, 17:31
(BTW, hope you get well soon Michael!)
RE: Mistake in my own report
by Aurora Robb Kristiansen on Wednesday December 1st 2021, 17:27
Hi again Chiito & Michael.

Thanks for taking your time to let me know how to make my test data better, and for bringing up the third-party program issues.
Regarding private servers and modding, I agree with you and regret I didn't realize this from the get-go.

I've submitted new test data that should be much better. I also did all my testing in vanilla, untouched Wine and no modifications to the game client.
DXVK's performance benefit, if any at all, seems to be minimal just like you said. However, esync does help smooth out stuttering and brings the game's performance up to par with native Windows, at least during my testing. Had a stroll in areas such as Jeuno, Windurst and Maze of Shakhrami (lots of enemies) without issue.

Let me know if there's anything else I can do!
RE: Mistake in my own report
by Chiitoo on Thursday December 2nd 2021, 2:16
Excellent, thank you again! No worries.

Yeah, I forgot to mention that esync is more likely to show a difference with this one.

Thank you also for confirming that the delay in the installer is not present under Windows. I'll try to get to filing a bug about that one.
Windower 5?
by Steve Cossette on Sunday October 18th 2020, 13:07
Anyone got the Windower to work? I am using the latest Wine build with Windower 5 (Non-DEV) and FFXI launches, but half of my keyboard doesn't work. Enter key being the main one too, which is quite important for FFXI.
RE: Windower 5?
by Michael D on Sunday October 18th 2020, 15:21
Windower is a separate program that hooks ffxi. As such, the research and guides for using ffxi in wine would not be likely to work for windower.
Over the years, this subject has been rejected as it is against the terms of service for ffxi and would need to have it's own appdb entry separate from ffxi.

You may have more luck with the official windower forums.

I do find it appropriate though, that over the years, all of the complaints about ffxi in wine have become a part of the official client when run on a modern version of windows or modern video hardware.
Game crashing???
by Ben on Saturday March 28th 2020, 3:13
So I went to load up the game after a couple months maybe, and POL.exe kept crashing. Not sure if it's due to wine 5.4 (staging) or something else, but I went down to 5.0 (stable) and it doesn't work there either. I'm having flashbacks of the old days where I had to have a second computer just to run this game.
RE: Game crashing???
by Ben on Saturday March 28th 2020, 3:39
Upon further investigation, it only crashes when I have the "virtual desktop" enabled.
RE: Game crashing???
by Chiitoo on Saturday March 28th 2020, 5:48
Interesting. Haven't tried that in a while. Will give it a spin at some point.

Thanks for mentioning it!
RE: Game crashing???
by Ben on Saturday March 28th 2020, 6:03
So I can't get it to run on 5.4 at all. I tried installing on a fresh prefix, and can't get the installer to run on either 5.0 or 5.4. I've been moving the prefixes over between installs for many many years, I haven't installed fresh since the game first started working in wine. Kinda bummed about losing the virtual desktop, I think the game still runs smoother if it is "full screen" but maybe I'm thinking too far in the past.
RE: Game crashing???
by Chiitoo on Saturday March 28th 2020, 6:24
Just to be clear, you can't run it in a virtual desktop, or you can't run the game at all?

I can't really say if it runs smoother or not in a virtual desktop, as I never really did that with anything... but well, I guess I'll find out later when I have the time to try it out!
RE: Game crashing???
by Ben on Saturday March 28th 2020, 6:35
5.0
Can't install
Can't run in virtual desktop
Can run in windowed mode
5.4
Can't install
Can't run in virtual mode
Can't run in windowed mode

Just really weird, everything has been flawless for me until now
RE: Game crashing???
by Chiitoo on Saturday March 28th 2020, 7:08
What happens during install, and which installer are you using?

Note that the current installer will appear to hang at one point while extracting stuff, but it /is/ still working in the background. It will finish its job in due time.

I did have a recent issue myself, where my old prefix did not work any longer, not even with older Wine, but a fresh prefix worked. Did not look into it more.

If you have an updated install of FFXI, you can copy the FFXI files over to at least skip some of the downloading when doing a clean install.
RE: Game crashing???
by Ben on Saturday March 28th 2020, 23:21
I grabbed the installer from the official site, gives me an option to install DX, POL, FFXI. both POL and FFXI result in nothing happening, and the DX installer runs but I don't want to install it.
RE: Game crashing???
by Chiitoo on Monday March 30th 2020, 4:20
Indeed, no need to run the DirectX installer.

When installing PlayOnline or FFXI, please check your list of processes for a 'msiexec' in case it's there doing its thing.

If yes, give it some time to complete. If no, post the terminal output via a pastebin service such as 'dpaste.com'.
RE: Game crashing???
by Chiitoo on Saturday March 28th 2020, 8:09
Just tried with Wine 5.5, and it works for me (virtual desktop and FFXI in full-screen mode). Both, my current prefix installation and a new installation (just tried the installation part, not updating and so on).

Performance seems similar to windowed full-screen mode without virtual desktop to me (if not even slightly better /without/).

Had some funky stuff going on with the mouse though, while in virtual desktop mode. It would activate the mouse in menus messing up my menuing big time! I wonder if that's just me, or if there's a bug lurking there.
RE: Game crashing???
by Pandoraxero on Saturday March 28th 2020, 9:37
I have pointed this out elsewhere, but the new installer seems to have issues more so than older installers.
It hangs for at least long enough to make a pot of coffee, then picks back up like nothing went wrong.
Wine msiexec is actually stuck in a loop at that point, and it breaks out of that loop after about 5 minutes.
(...as if waiting on the initial FFXI update wasn't painful enough...)

Virtual Desktop does not seem to offer any performance gain/loss over Fullscreen.
Windowed mode remains untested on my end.
RE: Game crashing???
by Pandoraxero on Sunday March 29th 2020, 20:17
Based on the extent of this discussion, it may be wise to post it to the forums, then link to that post.
I'm sure Chiitoo agrees.
RE: Game crashing???
by Ben on Monday March 30th 2020, 4:52
Good idea. Made a post here, expected to put more errors in, but installing the new staging version has fixed the virtual desktop issue for me. Install still isn't working, but now I don't think I'll need to install anymore.

forum.winehq.org/viewtopic.php?f=2&t=33639
gamepad issues
by Pandoraxero on Wednesday March 25th 2020, 19:52
I cannot verify what causes this (possibly loss of connection to the gamepad?) but I seem to be having intermittent issue with the gamepad... it just drops out from time to time
I've had this happen twice before. it is random and unpredictable, but thankfully hasn't happened to me when I'm doing something important.
looks like the kernel threw a message that there was a USB hub disconnect (kernel notes this may be due to EMI, which is fair enough - it's running through an extension cable, after all
dmesg shows that the gamepad reconnects shortly after, but wine no longer takes input from it... I'll have to do further testing involving disconnecting and reconnecting the gamepad to figure this out.

...In other news, I have finally finished what I started 15 years ago on Windows XP: My WAR is at 99, so that's 4 jobs I have at 99, and this is the first one I've taken to 99 on Linux.
RE: gamepad issues
by Pandoraxero on Thursday March 26th 2020, 22:08
indeed it seems that when a gamepad is removed, then reconnected in LInux, it doesn't seem to track properly in wine.
In windows, if I disconnect and then reconnect the gamepad, as I did when I switched cables for my PS4 controller, FFXI picks it up just fine.
RE: gamepad issues
by Chiitoo on Friday March 27th 2020, 8:37
Yep, it's been like this for as long as I can remember, so it's probably not regression since recent times at least.
Symmetric hardware test
by Pandoraxero on Saturday March 21st 2020, 12:47
New FFXI installer seems to choke for a good few minutes.
It seems to be an issue with msidb that sorts itself after a few minutes - not much different from the old installer, but at a much less promising point of the install.

Amusingly - polcfg.exe uses the pc speaker to beep when it prompts for confirmation (don't think it ever did that in windows, but it's obviously not a problem.)
Of note - The New installer seems to pre-populate the FFXI config (in the registry) differently from how the old installer did. that, I think, is what caused the previous config issue I had.

5.4 - so obviously, opening movie doesn't play properly. no video output and it gets about as far on the audio as "It all began with a stone, or so-" then amusingly (or annoyingly? depends on who you talk to, I guess) gets cut off.

Screen tearing here is significantly less pronounced - perhaps due to the more advanced hardware?
Text box FPS Drops seem to be less pronounced, but still there. Seems to be contingent more upon background objects (Players, NPCs, etc.) moving.
FPS (without the window up) with shadows set to "Normal" seems to bottom out around 25 - a slight improvement.
with shadows at full, it bottoms out at 14

Settings: 1080p, matched foreground and background
System info - Dual GTX1080Ti GPUs, AMD Ryzen Threadripper, 128GB DDR4.
OS Info - Gentoo Linux AMD64 - kernel 4.19.97 - nvidia proprietary drivers.

I will reply to this post with my test results in Windows on the same hardware later today.
RE: Symmetric hardware test
by Pandoraxero on Saturday March 21st 2020, 15:28
Here's my test results in windows - same hardware, same settings

Fullscreen Textbox - FPS Bottoms out at 28
no shadows in front of LJeunoAH - FPS Bottoms out at 29
normal shadows, same location - FPS Still bottoms out at 29
high shadows, same location - FPS bottoms out at 28

We literally cannot give this game anything higher than Gold until the FPS issues are addressed.
In wine, Normal Play with default ingame settings is more than practical. As soon as you put shadows to full, though... That's another barrier to platinum status.

Now let's go over my testing methodology:
1. Obviously, the settings need to be the same between the two installs. So 1080p matched overlay and background in both.
2. You also need a way to measure FPS - wine can output FPS to console, or you can have the nvidia driver do it in an overlay. In windows, the nvidia driver also has an FPS overlay feature
3. Need an area with plenty of characters. LJeuno AH is fairly well-populated, so I tested here, but really anywhere with 30+ people in a single area should work
4. Chat window is tested with no character movement, so just pull up the fullscreen chat window and measure the FPS
5. Shadows are tested with character movement. I have third-person camera set to inverted on both axes.
6. I then proceed to hold the movement and camera stick in the same direction on the gamepad/controller I have connected, and measure the FPS with No shadows, Normal shadows, and High shadows - this causes VERY FAST camera movement... it also makes your character run in circles like a doofus for the period of the test.
7. Note only the LOWEST FPS you see in each test. You may need to run each test for a minute or two to get the most accurate results.
RE: Symmetric hardware test
by Pandoraxero on Saturday March 21st 2020, 16:01
I'd also like to note the Linux Test results were partially the result of an all-nighter. I didn't get to sleep until around 7AM, and even then I only got about 4 hours before the sun woke me up...
For the record, the install I tested used the new POL installer but the old FFXI installer - this makes little difference in the results.
I decided to give the new ffxi installer another try on a second wineprefix... with a little more patience. and that's how I found out the new installer DOES work, but chokes up for a few minutes.

Honestly, if we could find out what it is about the shadows and text boxes that's causing wine to choke so hard, we'd be a lot closer to platinum.
RE: Symmetric hardware test
by Chiitoo on Saturday March 21st 2020, 18:00
Yeah, I have mentioned the installer issue in the how-to, but never got to filing a bug about it still.

Thank you for the Windows testing! I have to admit I expected quite a bit more similarity there, but at least that gives some hope to there being a chance things can be improved upon.

I'm not sure if there is a better way to go about finding he cause, than profiling (which I've already done) and looking deeper into the data with regards to what functions are being hit the most (not done much yet).
RE: Symmetric hardware test
by Pandoraxero on Sunday March 22nd 2020, 12:37
Well, what's even better about my setup... I set it up such that I can run my Gentoo install under windows using Hyper-V and manage updates while I'm running Windows.
That way, when I need to test in Linux, I can do so almost instantly. It's absolutely brilliant. Why didn't I think of this sooner?
RE: Symmetric hardware test
by Chiitoo on Friday March 27th 2020, 14:46
I know I recently said it's all mostly about the CPU, and I still do say that, but I certainly got an exciting improvement to performance when I switched from an NVIDIA GTX 960 to an AMD RX 5700 XT.

With the old, I would get around 5 FPS if not lower with shadows set to high, during something like Dominion Invasion at around day change (Earth time). Lower population servers will have around 100 people there at the time, with most of them having 1 to 5 creeper egos with them (trust magic). Would be interesting to know how things are on Windows there. It's probably the best stress test there is in the game right now. :]

With the new, I get around 10, with or without shadows. That's with a resolution of 2560 x 1440 and a background resolution of 5120 x 2880.

I did some quick testing at the Lower Jeuno auction house on Asura, and at 1080p I'd get around 16 to 20 FPS with shadows set to high, and 23 to 29 at normal.

A very happy unexpected boost all in all.

Which Ryzen CPU do you have exactly? Mine is from the first generation, a Ryzen 7 1700 (wanting a new one badly, but am going to wait 'til the next new stuff at the very least).
RE: Symmetric hardware test
by Pandoraxero on Friday March 27th 2020, 19:26
I have a First-gen Ryzen Treadripper 1950X - which is part of the reason why I'm not averse to running Gentoo in a VM for maintenance purposes: I've got cores to spare.

I'm waiting for the Gentoo team to put out a 5.5 ebuild so I can test that and see if it'll play the opening movie (I assume USE flag "gstreamer" is needed for that?) ... I'll test it once they put it up. I'm trying to keep my gentoo as far away from the -9999 ebuilds as possible, so I'm not going to use the WINE_COMMIT setting just to test 5.5 a couple days early. what I will to is adjust my package.acc*/wine to ~amd64 for
RE: Symmetric hardware test
by Chiitoo on Saturday March 28th 2020, 5:51
Ah, I see. Thought it was already from the 3000-series, which I'd expect a little to do better.

I don't know of anything blocking 5.5 inclusion, but I'll see if there's anything I can do to make it come out faster. ;]
RE: Symmetric hardware test
by Pandoraxero on Saturday March 28th 2020, 14:19
See, my primary catch is I don't want people to have to enable ** on ANY package.acc*.
Ideally, I wouldn't be advocating ~arch either, but that is the nature of testing these things.
There is, however, some irony in that I'm using ** on my my Raspberry Pi 4 to get snes9x running, but that's another story entirely.
Time was, I had an Athlon64 rig that was FULL ~arch. Maintenance was difficult, to say the least.

Basically, I just want people to have their Gentoo installs (assuming they are using Gentoo) as vanilla as possible.
I just keep the ~arch to a minimum and no package.un* or ** in package.acc*, that's my method.
RE: Symmetric hardware test
by Chiitoo on Saturday March 28th 2020, 14:54
Well, yes, users generally should not set '**'... unless they want to, of course.

For me, "stable" amd64 got boring real fast when I started with Gentoo back in 2010, and after I switched to ~amd64, I still get bored with nothing breaking, but I guess I might just be lucky. :]

Okay I don't really get bored. No time to get bored, unfortunately, but I think you get the idea.

We decided that for Wine, the only versions we'll mark stable are upstream stable, as it really doesn't make sense for the development releases, and as such, ~amd64 are needed for those.

I do admit that I started with a 6-core Phenom CPU, so having things to build often was not that big of a deal. I do have some single-core systems still laying around, where I also maintain Gentoo to some extent, but I rarely use them. Updating years and years old installs can be fun. :]

If vanilla "stable" is what a user wants, then yes, I want that for them too, but please don't wish that for everyone because then we have no one to test the "unstable". /grin
RE: Symmetric hardware test
by Pandoraxero on Saturday March 28th 2020, 20:37
Dunno if you had a hand in it, but the 5.5 ebuilds just got released sometime in the past 12 hours.
I'm gonna wear out this next echad ring charge then see how 5.5 does.
No install testing will be done this time. I think my latest installation test was sufficient.

I'll be symmetric testing it using the Threadripper build I have as well.

Off-hand - I decided we're testing it on a different OS with the same architecture, so I decided to see what would happen if I tried it on my Surface Pro X, which runs Windows 10 on ARM, and has x86-32 emulation.
Framerates are generally better on Windows 10 on the Surface Pro X than they are on my Linux installs. But if you have photosensitive epilepsy, it's not something I would advise, because, wow... the graphical glitches.
And if you don't know what to type where in POL, then good luck being able to get that working on a the Surface Pro X - all I had to work with was muscle memory.

But that's neither here nor there - just an indication that every "unsupported" platform has issues to work out, and wine will never be an exception to that particular rule.
I do have to wonder exactly how well the game runs with Valve's Proton, though.
And since I'm out of a job, and it's literally impossible to get one with the current situation, I have time (and bottles of whisk[e]y) to kill, so that'll be a fun test, I think.
RE: Symmetric hardware test
by Chiitoo on Saturday March 28th 2020, 7:46
Yes, you'll need USE="gstreamer".
RE: Symmetric hardware test
by Pandoraxero on Saturday March 28th 2020, 22:44
The movie seems to work - but after watching the whole thing (for the first time in over a decade, mind you), it basically terminated my POL connection.
I seem to recall this being an issue in Windows some 15 years ago... perhaps why the option to disable it is in the configs.
All the same, it brings back memories of good times - good times had under the influence of alcohol consumed underage. I regret nothing.

I'm having some issues updating my threadripper build (not wine-related, and nothing to do with the half-virtualized, half-baremetal nature of the install, I assure you).
It may be tomorrow or the day after before I am able to post test results.
Preliminary results on the i7-4771/2x780Ti system look promising, though. We shall see where this leads in due course.
RE: Symmetric hardware test
by Chiitoo on Monday March 30th 2020, 4:07
Yeah, the application terminates connection if 'accept' doesn't get triggered in time. Great design. :]
Fullscreen Text box
by Pandoraxero on Tuesday March 17th 2020, 20:45
Okay... I noticed something that is certainly unusual - when you select the text box and make it fullscreen - FPS plummets.
I''m literally seeing 11FPS With a random spike to 12 and a random dip to 10 now and then on my hardware... not sure what's causing this. It might be worth looking into
Steps to replicate: while ingame - Numpad+, Enter.
I don't know how much this would affect modern hardware.... but it seems to affect Intel i7-4771 w/nvidia GTX 780Ti accordingly.
RE: Fullscreen Text box
by Chiitoo on Tuesday March 17th 2020, 21:40
Yeah, the chat log has always, as far as I can tell, done that.

I never remembered to compare to how much it does that with Windows, and kind of expected it to be "normal", so didn't look more deeply into it still. I think I have profiled it a bit, but it was too long ago, and I can't remember the details (I'd guess most samples to be somewhere in wined3d in any case, but I'll have to re-check).
RE: Fullscreen Text box
by Pandoraxero on Wednesday March 18th 2020, 0:02
The unfortunate part is that I do not have a similar Windows system to compare this to. As one might expect, all my systems are drastically different in terms of hardware, and none of them dual-boot, since I consider that counterproductive, so I have no means by which to perform a direct comparison. Full disclosure, the system I'm playing on was meant as an emulation/couch gaming system and nothing more. The fact it runs FFXI does not surprise me, though. It does, after all, run PS2 emulation at 1080p and 4-5x the framerate.

I don't know specifically what causes this framerate drop... but it's probably worth looking into. ...even if it's not important to anyone who keeps their chat down to 8 lines max - it might affect other apps on wine in ways we do not fully comprehend.

Also - Jameson... Jameson is my alcohol of choice this St.Pat's day. If anyone was wondering!
RE: Fullscreen Text box
by Chiitoo on Wednesday March 18th 2020, 17:10
For sure it would nice to be able to alleviate the effect the chat log has, as it would definitely improve overall performance in general, at least for FFXI.

I've always felt the UI is probably built to be drawn over the game area very inefficiently, which is the main cause for this here slow-down we're seeing. That's just a feeling though. :]

A quick ten second profiling shows pretty much what I remembered.

UI visible, but chat log lines at 0:

samples| %|
------------------
397450 81.5612 wined3d.dll.so
38653 7.9320 kallsyms
14777 3.0324 libnvidia-glcore.so.440.64
8247 1.6924 anon (tgid:23910 range:0x1b11000-0x1e2efff)
7805 1.6017 dsound.dll.so
6100 1.2518 libasound.so.2.0.0
3243 0.6655 ntdll.dll.so
2303 0.4726 libc-2.30.so
2068 0.4244 [vdso] (tgid:23910 range:0xf7f67000-0xf7f67fff)
1495 0.3068 libpthread-2.30.so
1017 0.2087 d3d8.dll.so
911 0.1869 libX11.so.6.3.0
[...]

Chat log full-screen (well, half-screen as that's the width I happened to have configured at the time):

samples| %|
------------------
619083 70.9361 wined3d.dll.so
86606 9.9235 libnvidia-glcore.so.440.64
51648 5.9180 kallsyms
15377 1.7619 anon (tgid:23910 range:0x1b11000-0x1e2efff)
14145 1.6208 libX11.so.6.3.0
13837 1.5855 libc-2.30.so
12422 1.4233 ntdll.dll.so
10250 1.1745 libpthread-2.30.so
8602 0.9856 dsound.dll.so
8209 0.9406 [vdso] (tgid:23910 range:0xf7f67000-0xf7f67fff)
7681 0.8801 d3d8.dll.so
7370 0.8445 winex11.drv.so
[...]

I'd definitely need to test this out with Windows before I'd file a bug report about it, but I don't have any Windows installations around myself.
RE: Fullscreen Text box
by Pandoraxero on Friday March 20th 2020, 9:46
No Windows Install on which to test, you say?
I was hoping I wouldn't have to do this, but I have a spare hard drive or two, and a first-gen Threadripper build with Windows 10... and I suppose there is nothing technically preventing me from using one of those spares for purposes of a direct comparison.
in fact, I'd bet 9 million Asura gil that it'll compile WAY faster... and give me an idea of what cpu opts to compile for if I want to get a Ryzen3000-series build done.
...I'll get on it right away.
RE: Fullscreen Text box
by Chiitoo on Friday March 20th 2020, 10:03
That would be nice indeed!

Technically I might be able to pull it off as well, but I don't know when I'd have the time for it.
The Trees, Screen Tearing, and Shadows
by Pandoraxero on Monday March 16th 2020, 15:32
The tree flickering seems to happen in conjunction with screen tearing, which I think may be partially due to FFXI being locked to 30FPS, though I have no evidence to prove that belief.

Okay, I have the following input.
averaged between 1080p and 4k background resolution, at LJeuno AH on Asura (one of the busiest places):
Shadows disabled: Bottoms out at 29FPS (in other words - stable)
Shadows Normal: Bottoms out at 23FPS (playable)
Shadows High: Bottoms out at 11FPS (undesirable)

Notably - on a 780Ti GPU, background resolution set to 3840x2160 versus 1920x1080 seems to show negligible change in performance.

Having vulkan and vkd3d use flags enabled seems to have no measurable impact on performance or image quality - further testing may be needed to verify.
Reviewing debug output
by Pandoraxero on Monday March 16th 2020, 0:59
I set WINEDEBUG=warn+all wine pol.exe &> ~/wineout.log

In my experience (and granted my experience is from a dozen years ago, so... grain of salt?) the messages that show up the most are the most likely causes of slowdown. The two (ok, technically three) most prominent lines I'm seeing:
d3d:wined3d_texture_load_location Operatio requires 0x2 access, but texture only has 0xd.
d3d:wined3d_context_gl_bind_shader_resources No resource view bound at index 0, [0-1].

not sure what these are in reference to, but resolution might resolve the framerate issue.
RE: Reviewing debug output
by Michael D on Monday March 16th 2020, 11:01
I'm with you, most of my FFXI WINE experience is from WINE 0.7-0.9
I do not know if it is trying to use a different display or if there is simply a new d3d module that doesnt like the old methods that FFXI uses.
I will try to recreate it on another distro.
RE: Reviewing debug output
by Pandoraxero on Monday March 16th 2020, 21:04
Y'know, it's funny, but I cannot for the life of me remember who actually wrote the imm32 patch that got this game running a dozen years ago.
But I found it back then... I tested it. It worked. I had ganiman test it. It worked for him, too. I then appointed myself interim head hype-man and did my best to spread the word.
And FFXI has gone from being Garbage status to being a top voted app, displacing rival MMO World of Warcraft in both votes and rated compatibility in the process.
I've been debating whether or not I should have YouTuber Huntin 4 Games cover it, because honestly it speaks to the desire to play FFXI that it works on platforms the game was never intended to run.
We got people running this game on all manner of different flavors of Linux, I've tested it myself on FreeBSD, and we even got some people who just wanna play on their Mac.
It's crazy how the state of FFXI in Wine - and the state of Wine in general - has progressed over the past dozen years.
Immediate test results
by Pandoraxero on Sunday March 15th 2020, 0:26
I finally got my gentoo rig back up and running... I had to take the near-nuclear option... blow out everything
Here's my results:
Install for POL and FFXI ran as expected. I only ran pol setup and the setup in INST1.
Update of POL ran as expected.
Update of FFXI ran just as expected - horrifically slow.
FFXI booted up the first time with the most awful graphics because I didn't change settings

Apparently in the config, you need to SET EVERYTHING. the Resolution showing in the config won't actually set in the registry unless you select it. same with the resolution of the drawing precision. they will most likely be set to 800x600, with drawing precision set to garbage. THis can and will cause FFXI to fail to load.

The game itself appears to work on wine-vanilla-5.3 - more extended testing is needed.
I am noting tons of screen tearing (likely driver-related)
The issue I previously had with teleporting seems to have improved.
RE: Immediate test results
by Pandoraxero on Sunday March 15th 2020, 20:57
Update:

Okay, so I have leveled Monk from 25 to 50 on wine.
In this time I have noticed one - and only one - glaring issue: The framerates. Now, I don't have a FPS counter on my display, but it looked like in crowded areas (Jeuno, for example) framerates seemed to drop to 10FPS or less.

so, in summary:
1. The framerates are (still) an issue
2. If you run the config tool, you have to configure EVERYTHING. Don't take any shortcuts there - you'll regret it.
3. there seems to be an oddity with the textures drawing... the trees seem to flicker. May be due to mipmapping settings. I'll have to check into that.

Overall - it's stable. It runs out of the box with no more configuration than you would need for Windows. Minor texture and major framerate issues, though, So I cannot in good conscience rate this one platinum.
RE: Immediate test results
by Chiitoo on Monday March 16th 2020, 5:45
Do you have shadows set to 'High'? If so, I believe the FPS drop is normal. That is... it will happen on Windows just as well, although I admit I'm guessing here quite a bit as I haven't tried since several years ago. :]

By the by, you can use WINEDEBUG="fps" to have the value printed in a teminal window.
RE: Immediate test results
by Pandoraxero on Monday March 16th 2020, 10:55
Reducing the shadows seems to help... it doesn't completely eliminate the issue. I don't think my present setup would have this sort of issue at all on a windows install - though I don't have a spare drive just lying around to throw windows on in order to test that theory.
Also, the fact this system runs a 780Ti - which is, what, 3 and a half generations old? - might have some impact on performance.

Good call on the Shadows - That might cause me to downgrade the framerate issue from "Major" to "Minor" considering shadows aren't set to full by default... And likely earn a Gold ranking for this game.

Not platinum, though... there are still minor framerate issues, and the flickering trees that I keep seeing (which are minor issues, btw), and some issues with the ffxi config tool that don't seem to exist in windows. Speaking of which, do you think we should link to ffxiclopedia.fandom.com/wiki/Graphics in the howto? I mean, a lot of long time players don't even use the config tool, instead manually editing the config in the registry the way they want it.

It is good tat over time we have become more... stringent, i guess is the word? ... in our expectations as t how the game should function. I mean... Gani and I were giving this thing silver ratings in 2008 after manually applying the imm32 patch. I'd say the lower ratings the game has been getting lately aren't truly a sign of the game getting worse per se, but our expectations becoming more complete. We gave the game Platinum before - retrospect, I don't think it deserved it. It seems to just barely deserve gold right now, and it's running even better than when it WAS getting platinum.
RE: Immediate test results
by Chiitoo on Monday March 16th 2020, 12:29
Well, FFXI has always, at least during my 10 years with it, been more about the CPU than GPU. :]

I might recommend monitoring your CPU usage to see if you're hitting around 100% on two cores with your current settings to make sure that's where the neck of the bottle is.

Would it be possible to show the flickering issue in a video? I'm not sure if I've seen the same thing (it might also be normal, and go away if you use a higher background resolution).

I for one will definitely not link to anything under 'fandom.com'. They've done a great job at making the site horrible to use, and it's a bit sad that I can't recommend using FFXIclopedia any longer because of it. I spent quite a lot of time editing said wiki until the founders left (I left with them, and am quite happy I did seeing what wikia/fandom is doing nowadays).

I haven't edited the registry directly ever since they added more background resolutions and such while updating the config utility, so I don't see a reason to recommend poking at the registry. If there's anything specific, it would probably be better if it was added here, instead of off-site linking.
RE: Immediate test results
by Pandoraxero on Monday March 16th 2020, 14:42
Fair enough, it has always been more about the CPU with FFXI, even on Windows (Well, as long as you have practically any PCIe GPU). FFXI was also developed at a time where desktop CPUs had a single core. I feel like if the game were fully retooled to work with multi-core CPUs, it would work a lot better. This, unfortunately, is something only Square-Enix can do, and at this point, there is no incentive for them to do so - which sucks.

I don't know if I can get the flickering issue in a video. Using a higher background resolution might help... but a 780Ti isn't exactly a 4K GPU - yet more reasons to get a new GPU.

I was pretty heavily involved in the scene before gani and his crew left ffxiclopedia. Then BGWiki sprung up. It became obvious to me that the BGWiki/FFXIclopedia rivalry was a political dumpster fire with all the drama of an extramarital affair. So I mostly quit playing so as not to get involved. These days ffxiclopedia has some very egregious advertising - That's the worst thing about them at this point, and that's pretty much the same for everything hosted by fandom.

The problem with the new config utility is that on a new install, you have to set everything, otherwise certain fields in the registry will be populated with garbage.. And instead of dropdown boxes for a lot of the stuff, they now have sliders. I'd rather just use regedit and call it a day, but that's not how we test things here.

In any event, I have just recompiled wine-vanilla and staging with the relevant use flags for Vulkan set... we'll see if this affects performance one way or the other... or possibly not at all
RE: Immediate test results
by Chiitoo on Monday March 16th 2020, 17:15
I don't know when BGWiki sprung up, but the founders of FFXIclopedia went on to set up Gamer Escape. Sadly not too many followed, and these days I'm the only one maintaining their FFXI wiki, while the rest of them focus on the XIV side. I never felt like getting involved with BGWiki, as that community accepts behaviour I don't (mostly regarding third-party applications and such).

But yeah, the advertising over at fandom is probably the main reason I would not link to it. They don't deserve it (in my opinion).

All that aside, I wonder if that's something new, regarding the config utility. I have seen drop-downs there being filled with question marks, but so far I've figured it was just a display issue. I'll try to re-test soon, and check the registry too.
RE: Immediate test results
by Pandoraxero on Sunday March 15th 2020, 23:17
setting mipmapping (0000 in the registry) to 4 seems to reduce (not eliminate) the tree flickering issue.
I enabled the overlay frame counter (a function of the nvidia drivers) - I seem to be getting as low as 15FPS in L.Jeuno in front of AH, so not quite as bad as I thought, but still pretty bad
Screen Tearing seems to be an issue, but again, that's possibly a driver thing.
I do apologize for the massive uptick in notifications you may have been receiving, but as a former hype man for this effort, this is important to me...
I'll probably be testing with wine-staging sometime later, but I am NOT doing the install again. that update is MURDER.
I'm back... for now
by Pandoraxero on Sunday March 8th 2020, 17:56
Honestly, my FFXI career has been off-and-on. But I've been playing a bit recently, and decided I was going to give it another test run on Gentoo.
Unfortunately for me, something broke between the nvidia drivers and the hightime (4.20) kernel in Gentoo, and that apparently caused a large number of underlying problems (with my system, not wine/ffxi) to surface once I put 4.19.97 in place.
It'll be a while before I get these issues sorted, but as a tester and former (2007/2008) hype man, I look forward to being able to provide the community with new test results as soon as I can.
4.21 and CSMT
by JY on Tuesday December 24th 2019, 23:47
Long Time Listener, First Time Caller

I have been running on Debian jessie (wine32 1.6.2-20) and stretch (wine32 1.8.7-2) for a long, long time on a Haswell with integrated graphics without any problems. I finally upgraded to Debian buster last month, but the distro packaged wine32 4.0-2 crashes with a page fault on read access. I found in someone else's bug report that this was fixed in a 4.0.x patch release (sorry can't find browser history of it), and the 4.0.3~buster (and 4.0.2~buster) from wine-builds have been working for me (env WINEDEBUG=-all wine /path/to/pol.exe). However, these versions are using 200% CPU when the older wine versions only used about 80% CPU.

I built Wine 4.21 using Tk-Glitch/PKGBUILDS on Debian buster. This also runs the game, but still uses 200% CPU and I noticed kernel messages indicating thermal throttling events. i7z says the temperature would reach 80-100C, especially if I pin the two threads onto the same physical core. I tried some software optimizations like changing the InterlockedCompareExchange() in wined3d_resource_wait_idle() into a test-and-test-and-set so that it's not doing a lock cmpxchg to just do a poll on a reference count, which seemed to improve frame rate and responsiveness, but still caused throttling with 150% CPU. Finally, I ran winecfg and disabled CSMT. The temperature is back to 70-80C, frame rate feels a little slower, but no more thermal throttling.
Everything works fine...to a point. Help?
by Ryan on Monday November 25th 2019, 21:32
First of all, let me say that your guide has been a godsend in what I knew was going to be a lengthy process. It is easily readable and doesn't make any assumptions. I am not sure if anyone even checks this any more, but I'd love some help.

I was able get DirectX installed (as far as I know) and make the exception in WineCFG as you instructed. I was then able to get PlayOnline installed and even able to log into it as far as launching the game (which obviously I could not do because the game was not installed yet).

My issue comes at the game installation, in which I am presented with just a black window of the FFXI screen. The Square Enix site does say that while the installer is saying "computing space requirements" that this can take some time. Sometime when I open the installer I can actually see the screen and graphics where it says "computing space requirements" but there does not appear to be any rhyme or reason to this.

I am confident it is something silly I am missing. To make things easier, I have gone ahead and saved a copy of my terminal script for you to notice any errors or issues from my end and I can provide if needed.

This is the part that messed up I think:
00f9:fixme:ntdll:NtLockFile I/O completion on lock not implemented yet
00f9:fixme:msi:event_spawn_wait_dialog doing nothing
00f9:fixme:msi:internal_ui_handler internal UI not implemented for message 0x0b000000 (UI level = 5)
00f9:fixme:msi:internal_ui_handler internal UI not implemented for message 0x0b000000 (UI level = 5)
0101:fixme:ntdll:NtLockFile I/O completion on lock not implemented yet
RE: Everything works fine...to a point. Help?
by Chiitoo on Tuesday November 26th 2019, 3:09
Please not the warning: "The following information may be largely out of date since 2016-12-19 or earlier."

At this time, or at least last I tested, there was no need to "install DirectX" or make any overrides. Things may work for you just fine even if you already did do those things, so I'm not saying you should definitely get rid of them, but at least for future record, they're not needed (or should be reported if so).

I'll try to clean more of the old info out soon. Apologies for that.

So to reiterate, anything past the note I mentioned should not be required for installing and playing the game.

As for your issue, note the part before "Last updated until this point: 2019-06-14":

"During the installation process of Final Fantasy XI, a window with only borders, and no rendered content may be seen, and the 'msiexec' process can be using up to 100% of a single CPU core while possibly processing the 'FINAL_FANTASY_XI.msi' file probably. This can take up to at least four (4) minutes, but should eventually complete."

Could you try and see if this is the case for you, too?

Thanks!
RE: Everything works fine...to a point. Help?
by Michael D on Tuesday November 26th 2019, 6:46
As Chitoo has mentioned, there is indeed a good chance that you will encounter the blank window. The speed and power of your computer do not seem to hold sway over this happening or how long it happens. I have seen more powerful computers take 6 to 8 minuted to get past this part while lesser computers have finished in 3 or 4. This part will make you think that there is something seriously wrong; just hold on and wait for it to either crash or finish (it has never crashed for me)
I did not have to install anything other than the POL viewer and FFXI when I reinstalled it 2 months ago. There was no need to do the dxdiagn override.
Now that the game is effectively run the same way on both windows and WINE because of how windows and your graphics card likely no longer support directx8, I think a portion of the HowTo might need to reflect that the lighting bugs (flashing and clipping) are present in the Windows version of the game.
RE: Everything works fine...to a point. Help?
by Christopher Strom - Kbtaru on Tuesday November 26th 2019, 7:35
I agree with everything above. I am curious what version of wine you are running.

I been working on getting atm0s�s directx8 to directx9 proxy working. As I believe it may help with these challenges.


www.bluegartr.com/threads/129943-Direct3D8-to-Direct3D9-Proxy-Performance-Helper-(For-FFXI)
RE: Everything works fine...to a point. Help?
by Michael D on Wednesday November 27th 2019, 11:38
another option to use is dgvoodoo, While i will not get into the loyalties of the two projects, atm0s' project is specifically designed to work with ashita and some versions will absolutely check to see if you are using ashita and refuse to load and intentionally crash if they do not see ashita.
As far as I am aware, Wine is already handling these conversions and has been for at least nine years, making the use of these directx tools potentially redundant. This is why there was always a complaint that lights flickered in WINE but not in windows. Now windows has to do the same type of thing and there is a flicker in the official windows version run on windows 10. Please do not take my word as expert opinion, I am just some random person who fills out paperwork for this game in wine.
Be forewarned, the latest version of dgvoodoo2 WILL be flagged as a dangerous executable containing malware. You should absolutely be wary of this and do your own research. I will not tell someone to ignore their AV warnings because that leads to people ignoring their AV full time or thinking that it is useless. You can obtain previous versions that do not trigger windows device driver signing problems.
RE: Everything works fine...to a point. Help?
by Ryan on Tuesday November 26th 2019, 10:12
Thank you for getting back so quickly. I did miss the note about DirectX not being needed for installation now, so I apologize for that. I think I went right to the how to and missed a few comments. :P

I can confirm that even after doing the DirectX overrides, the game DID eventually start. It took about 9 minutes, so there appears to be no rhyme or reason on how long it takes, but it does eventually change from a black screen to the right one.

As of now, everything is running pretty good on WINE 4.02, albeit a little bit laggy. How do you launch the FFXI Config tool in Wine. So yeah, everything is working so far and thanks again for getting back to me so soon. :)
RE: Everything works fine...to a point. Help?
by Michael D on Tuesday November 26th 2019, 10:16
Normally, I used registry overrides instead of the ffxiconfig.exe. This is an artifact of the old days when you had to edit the registry in windows to get high resolution and such. When you say laggy, let us know what you mean. I'm going to call out some extreme cases here, but I have seen people call a full second input lag "a littel laggy" and others call 250ms "extreme and unplayable". I have also seen people claim that 48fps is "unplayable" when the original game wouldnt break 30 without overrides. Can you estimate the lag time and the type? low framerate, high input latency, etc.
RE: Everything works fine...to a point. Help?
by Ryan on Tuesday November 26th 2019, 10:36
Haha! Thanks again for the quick reply. For the purposes of your response, you can pretty much assume that I�m technologically inclined, but certainly not a hardcore PC gamer.

I�m not concerned with getting more out of the game than what is standard at a base level, so when I say laggy, I just mean that there is some slowdown every so often, certainly more than I see on FFXIV, even in the PS3 beta days.

If that is to be expected then I am certainly good to go and again, I very much appreciate this guide, your comments and the help getting this running! It�s the only FF game I�ve never played. :)
RE: Everything works fine...to a point. Help?
by Chiitoo on Tuesday November 26th 2019, 10:33
Glad to read it was just about that delay and the window not being drawn! (One of these days I'll have the time, energy, and memory to look into it more deeply...)

Regarding the configuration utility, you can find it under '/SquareEnix/FINAL FANTASY XI/Tools' (or/and 'ToolsEU' and 'ToolsUS)'.

Simply 'cd' there in a terminal window, and do 'wine FINAL\ FANTASY\ XI\ Config.exe' for example (can also quote the exe name, instead of using the backslashes).

It's also possible that you can just double-click the exe to run it via a file manager.
2019 Reassessment
by Pandoraxero on Tuesday January 8th 2019, 5:29
Just did a reassessment of the game on 4.0 rc3 (Highest Version I could find an overlay for in Gentoo)

Didn't exactly run as expected. I assessed it the same way a total noob would. ...or at least a total noob who somehow knows how to use portage overlays. No real modifications (except setting the res to 1600x900. Home Point Warping crashed the game.
I'm going to try the an ancient tech support trick, which is known to all but the noobest of noobs, I guess that would make them... n00bs (because double zeros. ...Amirite? ...Guys? *sigh*
Okay. Ill-conceived lame pun aside, I'm gonna turn my Linux rig off... stand outside in the apparently rainy weather smoking a cigar on a covered porch for about an hour... then turn it back on again... then RE-reassess the game. Who knows, maybe my dual 780Ti's are running too hot or something.
RE: 2019 Reassessment
by Chiitoo on Tuesday January 8th 2019, 7:29
Which overlay were you using? I don't think we have any official ones with release candidates on Gentoo.

We tend to skip those for a reason, but you can always use the '9999' to get them if you like. :]

I have not had the issue with home-points that you mention. Perhaps I just conveniently dodged that bullet by not following the tags too closely (currently at rc5).

The only real issue that I have had, (though only really an issue if you ever move the mouse around) is bugs.winehq.org/show_bug.cgi?id=46059 which is also now fixed.

That, and bugs.winehq.org/show_bug.cgi?id=42008 which is still around.

It would probably be a good idea to try with an official Gentoo ebuild too, though it may be unlikely to fix the issues you have been seeing.

Thanks!
RE: 2019 Reassessment
by Pandoraxero on Tuesday January 8th 2019, 8:09
overlay I'm using is unofficial, "hossie"
I ran a search for wine overlays for Gentoo and came up with this list:
gpo.zugaina.org/app-emulation/wine-staging

Which seems to be about the most comprehensive list I can find. I'm considering 9999, but like I said elsewhere, that's an unofficial version, and its results don't really count for testing purposes. As it stands, the game's only KINDA playable... IF you're lucky.
I do understand the rc releases being only about a week apart from one to the next being more than a minor hassle to the Gentoo maintainers, though.
What USE flags do you have enabled for wine? I have USE="X alsa cups fontconfig gecko jpeg lcms ldap mono mp3 ncurses nls opengl perl png pulseaudio realtime run-exes sdl ssl threads truetype udev xcomposite xml" for wine-staging
RE: 2019 Reassessment
by Chiitoo on Tuesday January 8th 2019, 8:47
If you use the environment variables I mentioned, you can get any officially released version you want with 9999. In a sense, I would say that it would count more as a test result, than one with an ebuild from an unofficial source does. :]

Which version of nvidia-drivers are you using (415.25 here at this time)?

Also, could you grab the terminal output from a session where you experience a crash, and throw it into a 'pastebin' of some kind (and link to it here)?

I haven't actually tried Wine-Staging for FFXI in a while. It causes my whole desktop go wonky (windows turn black and stuff like that). It may be caused by some setting in the prefix that I have set in the past, however, though I think this one should be relatively clean. I'll need to look into that more with time at some point.

Here are my usual flags:

app-emulation/wine-vanilla-9999::wine was built with the following:
USE="X alsa fontconfig gecko gstreamer jpeg mono mp3 ncurses nls openal opengl png ssl threads truetype vkd3d vulkan xinerama xml -capi -cups -custom-cflags -dos -gphoto2 -gsm -gssapi -kerberos -lcms -ldap -netapi -odbc -opencl -osmesa -oss -pcap -perl -prelink -pulseaudio -realtime -run-exes -samba -scanner -sdl (-selinux) -test -udev -udisks -v4l -xcomposite" ABI_X86="32 64 (-x32)"
CFLAGS="-march=znver1 -O2 -pipe"
CXXFLAGS="-march=znver1 -O2 -pipe"

app-emulation/wine-staging-9999::wine was built with the following:
USE="X alsa fontconfig gecko gstreamer jpeg mono mp3 ncurses nls openal opengl png ssl (staging) threads truetype vkd3d vulkan xinerama xml -capi -cups -custom-cflags -dos (-ffmpeg) -gphoto2 -gsm -gssapi -kerberos -lcms -ldap -netapi -odbc -opencl -osmesa -oss -pcap -perl -pipelight -prelink -pulseaudio -realtime -run-exes -samba -scanner -sdl (-selinux) -test -themes -udev -udisks -v4l -vaapi -xcomposite" ABI_X86="32 64 (-x32)"
CFLAGS="-march=znver1 -O2 -pipe"
CXXFLAGS="-march=znver1 -O2 -pipe"
RE: 2019 Reassessment
by Chiitoo on Tuesday January 8th 2019, 8:55
Just remembered that it's not too long from now, when I did experience some rather random-like crashes myself... but they were very rare, so I'm not too certain it could have been what you are seeing. It very well might have, however.

I haven't seen them in a while now though, so I'd definitely encourage you to at least try with rc5 via the 9999, or wait for the 4.0 release for a re-reassessment. :]

Also, if possible, try Wine-Vanilla as well (which I mainly use for FFXI, so there might be things happening with Staging that I know not about).
RE: 2019 Reassessment
by Pandoraxero on Tuesday January 8th 2019, 7:49
The problem appears to be less consistent, but more widespread, than my initial tests implied. The cutscene for leaving the castle in San d'Oria also seems to trigger the issue. what's worse, when trying to get more output from winedbg using the "Program Error" window, it brings up the window... but I can't interact with it. I can go for the slightly more risky proposition and run wine-9999 in Gentoo, but results from that don't count as actual test results, since it's not an actual version.
RE: 2019 Reassessment
by Chiitoo on Tuesday January 8th 2019, 7:54
You can usually use 'EGIT_COMMIT' to get a specific tag with the 9999 versions, or use the Wine-specific ones we have.

See for example: wiki.gentoo.org/wiki/Wine#Environment_variables
RE: 2019 Reassessment
by Chiitoo on Tuesday January 8th 2019, 7:57
Way to submit early, me.

Of course it would be best to use an officially released version from the official Gentoo ebuild repository, instead of something from an overlay or the 9999, but I just wanted to mention these tricks since we do indeed skip release candidates, and since there seem to be users wanting to test using them. :]
RE: 2019 Reassessment
by Pandoraxero on Monday January 14th 2019, 11:01
I used this method to get rc6. Or perhaps... git rc6? Because I'm using git.... right? ...geez, tough crowd...
I'm not going to review it because I still have the same problems.
Thought I saw one post where you asked what nvidia-drivers version I'm using... probably been deleted, since I don't see it anymore, but in any case, I'm using 415.23. looks like 415.25 is available, but I don't really think this is an issue with graphics, since the issue is intermittent, with no guaranteed means by which it can be made to present itself... aside, perhaps, from repeatedly HP warping. and it's not just HP Warping that causes it.

Now, I should note that I'm NOT using the WINEARCH variable... so I'm not sure whether or not that has anything to do with it. Debug info shows Platform: i386 (WOW64)
Should I try running it with the WINEARCH set, or is that just my, uh... suspicions from past testing talking? Because as I recall, the game didn't ALWAYS work without setting WINEARCH. ...And I've been on the FFXI/Wine effort for over a decade.
RE: 2019 Reassessment
by Chiitoo on Monday January 14th 2019, 19:30
The comment is still there ( appdb.winehq.org/commentview.php?iAppId=1992&iVersionId=2739&iThreadId=104127 ), a bit towards the top from this one. We've got multiple 'threads' of replies going on here now, which is probably why it got “lost”. :]

I'm not blaming the nvidia-drivers (yet), just comparing out set-ups so as to get a better idea of what might make the difference.

I have been mainly using a 32-bit prefix, but I'll create a WoW64 one to test this out. I should actually have one around already, as I was mainly using that before creating this to test something else... I don't remember having issues with it before, aside from something during the installation phase, way back.

With regards to Home Points, I tend to make use of them extensively while doing Ambuscade, every few minutes or so, so if it's triggered by something during changing areas for example, I definitely should be hitting it too.
RE: 2019 Reassessment
by Pandoraxero on Tuesday January 15th 2019, 1:23
Okay, I see it now. Yeah, musta got lost.

I'll try with a new WINEPREFIX and WINEARCH=win32
This may or may not help depending on the actual nature of the issue, but it's worth a shot. If it does help, I won't be able to rate the game as platinum because of the workaround.
RE: 2019 Reassessment
by Pandoraxero on Tuesday January 15th 2019, 6:01
I was almost convinced setting win32 did the trick. then I talked to the NPC to zone out of the castle in sandy. I don't know what it is, but SOMETHING is causing issues here. WINEARCH=win32 seems to make the game slightly more stable... but when things like this happen, it raises a question.
I get the following output:
0031:err:d3d8:d3d8_device_upload_sysmem_vertex_buffers Failed to update buffer (I don't have timestamps enabled so I'm not sure if that has anything to do with it)
wine: Unhandled page fault on read access to 0x00000120 at address 0x79ee23f0 (thread 0135), starting debugger... (and the debugger doesn't start. It must have crashed PRETTY HARD.)
RE: 2019 Reassessment
by Jeffrey K Berhow on Sunday January 27th 2019, 2:14
This is what I got when the first portion of the Lower Delfutt's Chains of Promathia cutscene ended.

0009:err:d3d8:d3d8_device_upload_sysmem_vertex_buffers Failed to update buffer.
wine: Unhandled page fault on read access to 0x00000084 at address 0x7cbfeb39 (thread 0009), starting debugger...
00cb:fixme:dbghelp:elf_search_auxv can't find symbol in module

In the Staging tab of winecfg, I unticked "Enable CMST for better graphics performance (deprecated)" and was able to get through the cutscene.

Everything seems to be working now.
Re-assessment
by Pandoraxero on Saturday December 3rd 2016, 9:17
Just so everyone knows, I am one of the people who pushed for the imm32 sanity check patch (which saw this game go from Garbage to Gold) 8 years ago. I see this as a good time to re-assess the situation, and now I'm running Debian, with the intent of getting a clearer picture of the present state of this project on a base binary distro (which has forked more than any other distro, and most notably is the parent distro of the *buntu distros, and mint, among others)

After completing my assessment of the status on Debian, I intend to touch other base or near-base distros. My agenda currently includes Fedora (near-base), Slackware (base), FreeBSD (base, non-linux), and finally, I will be re-assessing the situation with Gentoo (base, though I suspect Chiitoo may also be working this part of the operation right now).

At each step, I'll be providing test results, and writing a How-to. The test results will be published, but the How-to will not be finalized, let alone published until I find an appropriate version of wine upon which to base it, and wade through the intricacies of each distro.

This will be a slow process, and could easily take 3 months, or possibly longer (I suspect FreeBSD and Slackware will be the hardest for me to work with, considering my lack of experience with them). I will be finishing up with Gentoo, since it's what I've been using for a decade. Essentially, these different distros will be taking me out of my element, for the sole purpose of achieving a fresh perspective on the status of this particular game on wine as a whole, rather than its status on just one distro.

If I'm not covering your distro of choice, well... I'm sorry, but I don't have time to touch on every single *buntu, *BSD, and all the numerous forks of Debian, Red Hat, and Slack. I'd be working this piece for another 8 years if I did that. My goal is to evaluate distros which have demonstrated, and continue to demonstrate, longevity. This essentially means base and near-base only.
RE: Re-assessment
by Michael D on Wednesday December 7th 2016, 8:05
Your work is much appreciated. I will also be submitting tests again. Hopefully I can contribute, rather than just create headaches for you.
RE: Re-assessment
by Pandoraxero on Wednesday December 7th 2016, 9:59
Well, so far, fedora's giving me nothing but a truckload of derp on install, complaining about missing mono and stuff. I got what I believe to be an effective-enough workaround. I've been tried and tested, but I won't accept defeat. (Bonus points if you can name where that line's from, without googling it.)
RE: Re-assessment
by Michael D on Wednesday December 7th 2016, 10:16
I had to do a winetricks dotnet20 in a 32 but prefix to get it installed on kubuntu 16.10
I'll be reactivating my account to test fully, but pol updated fine.
Sorry I can't place your reference. However, for when the day is long, forever carry on.
RE: Re-assessment
by Chiitoo on Thursday December 15th 2016, 3:49
Which installer were you using?

I only needed to set the reported Windows version to 7, when using the installer from the PlayOnine site (to get into the title screen, after the licence agreement). This is the only thing keeping it from being Platinum, but that should soon change, with the default becoming Windows 7 some time after 2.0 release. \ö/
RE: Re-assessment
by Pandoraxero on Monday December 12th 2016, 6:44
Well, I've stumbled upon the beast that is the reason I hold such disdain for binary distros... I'm co-posting this to the forums, but I figured I'd mention it here, too. This is on Fedora25. I think Fedora might have forgotten a dep or two...

bash-4.3$ wine32 C:\\Program\ Files\\PlayOnline\\SquareEnix\\FINAL\ FANTASY\ XI\\ToolsUS\\FINAL\ FANTASY\ XI\ Config.exe
fixme:winediag:start_process Wine Staging 1.9.23 is a testing version containing experimental patches.
fixme:winediag:start_process Please mention your exact version when filing bug reports on winehq.org.
fixme:htmlhelp:HtmlHelpW HH case HH_INITIALIZE not handled.
err:d3d:match_fbo_tex_update FBO status 0x8cdd
err:d3d:match_broken_arb_fog FBO status 0x8cdd
^Cfixme:console:CONSOLE_DefaultHandler Terminating process 8 on event 0
bash-4.3$ wine32 C:\\Program\ Files\\PlayOnline\\SquareEnix\\PlayOnlineViewer\\pol.exe
fixme:winediag:start_process Wine Staging 1.9.23 is a testing version containing experimental patches.
fixme:winediag:start_process Please mention your exact version when filing bug reports on winehq.org.
fixme:wbemprox:wbem_services_CreateInstanceEnum unsupported flags 0x00000030
fixme:wbemprox:enum_class_object_Next timeout not supported
err:d3d:match_fbo_tex_update FBO status 0x8cdd
err:d3d:match_broken_arb_fog FBO status 0x8cdd
^Cfixme:console:CONSOLE_DefaultHandler Terminating process 8 on event 0
RE: Re-assessment
by Pandoraxero on Tuesday December 27th 2016, 1:56
I have recently come to the conclusion that this game is, in fact, easier to install and run in wine... than it is on Windows 10. There's been enough people complaining about Windows 10, and this most certainly confirms that there is good reason behind the near-unilateral hatred Windows 10 has been getting... and continues to get. Wine has been getting consistently better over time, but catching up to windows has been a cat-and-mouse game for over a decade. As resentment towards the newsances (see what I did there?) of higher and higher Windows Versions grows, I think people will eventually start looking for a new OS... but then I said the same thing a decade ago when Vista had its RTM, and the net change so far has been negligible, although the non-GNU, yet Linux-based, Android OS is showing a substantial market success.
Japanese input?
by GF on Thursday September 8th 2016, 13:08
Is anyone aware of a way to get Japanese input working? My locale, keyboard, and FFXI client are in Japanese, and I can read what others write, but cannot type back in Japanese. When I write in Japanese the usual prompt appears below the FFXI window, but the characters themselves never appear in the FFXI chat log. I'm on wine 1.8.4 and have also tried this with the English client.
RE: Japanese input?
by Chiitoo on Thursday September 8th 2016, 13:24
As far as I'm aware, it should be possible only in the Japanese client (without modifying the client).

That said, I've never even tried to input Japanese into an application running via Wine, only native Linux applications (via uim), and as such, I don't have any immediate ideas on what the issue could be.

Are you able to input Japanese into other applications running via Wine?
RE: Japanese input?
by GF on Thursday September 8th 2016, 13:38
I hadn't actually tried before.. but just did, and can enter Japanese text in fine with WINE's Wordpad.
RE: Japanese input?
by Chiitoo on Thursday September 8th 2016, 13:53
OK that's very good to know. If it works in other applications, it would seem to me that it might be time to open up a bug!

I don't have the Japanese client, so I can't help by testing this myself... although I guess I should be able to get it from the official website nowadays, and there we go, downloading it now.

I'm not sure when I'll have the time to do it, but I'll give it a go.
RE: Japanese input?
by Pandoraxero on Monday December 12th 2016, 9:20
The last time I tried it, which, as you probably guessed, was 8 years ago, it DID NOT work. The game would run, but switching to Japanese input did not work, presumably because the IMM/IME was not 100% compatible with what the JP version was expecting. That may have changed, though. It was a real bummer for me back then, as, in addition to college classes, I was using FFXI as a platform to get a better grasp on the Japanese Language. The download from the JP site is, of course 100% legal... though 8 years ago, it did not exist, so I had to pirate the Japanese version, if getting a disc for an MMO that requires a user account and monthly payments can be described as "piracy"...

Operating off my knowledge from Windows: notably, Final Fantasy XI itself fully supports Japanese input, regardless of region, it's just disabled by detecting what PlayOnline version is being used (which explains how windower made it work on US and EU versions. Its IME plugin alters that part of the detection, but leaves it so that the ingame text is still in... whatever native language. I think the only options were EN, DE, and FR). So you don't need to reinstall everything, just PlayOnline. You can use registry hax to migrate your US, EU, or JP settings, verbatim, to any other region, that way you're not wasting drivespace (Which, 8 years ago, was NOT CHEAP)
RE: Japanese input?
by Chiitoo on Thursday December 15th 2016, 0:10
Apologies for the delay!

I finally put some time into trying this out, and while it took a while for me to figure out, it seems to work at least in the PlayOnline Viewer.

In short, have 'app-i18n/uim' (Gentoo package naming) and its dependencies installed, have 'uim-xim' running, and start the application with XMODIFIERS="@im=uim" set (either on the command-line, or exported for the shell).

I can then switch the input method via the global hotkeys set via uim settings.

The IME controls on the lower right part of the screen don't seem to be usable for me though. It did curiously say 'ON' before I got it working. Now when I have it working, it says 'OFF'. :]

The fact that it does work in Wine notepad there, but not in-game is curious as well. Regardless, I hope this will be of help. If you can confirm that it does work for you in PlayOnline Viewer as well, but not in-game, let me know and I'll test that out as well (may test it anyways at some point, as it shouldn't be too big of a deal to update the client with my already up-to-date files).
RE: Japanese input?
by Chiitoo on Thursday December 15th 2016, 5:08
Bäh! I knew it was too easy.

Does not work in-game, at least not yet. Now that I think about it, it does make sense, because the Japanese Input Method is more or less built-in to the game client, as far as I understand it, and it's not even possible to paste text into it. This may explain why 'uim' works for PlayOnline Viewer, since it's possible to paste text into that.

I feel it may require considerable work to have it working in-game, via Wine, but I will hope I'm wrong... :]
RE: Japanese input?
by Pandoraxero on Thursday December 15th 2016, 8:56
You are correct that it is built in to the game itself. My best guess is that such a method was required for cross-compatibility between PS2, PC, and XB360, but the console versions have been decommissioned, so those are hardly relevant. I think Ganiman and I were trying to get it to work 8 years ago, with some assistance from Aikar (makes sense, as Aikar was partly responsible for the Windower project). A fruitless endeavor, but things may have changed in 8 years.

If it weren't for Aikar, I would have likely never figured out that the JP version of PlayOnline could be used with a NA FFXI install. I'm all but done testing on Fedora. I still have Slack and FreeBSD to go, then I get Gentoo back in its place for testing, then test with Knoppix, then back to Gentoo to see if I can give JP input the old college try.
WARNING: New fishing system (March 2014 Version Update) will occasionally crash the game
by Megas of Vecanti on Friday March 21st 2014, 14:54
About 0.5% of the time, correctly pressing your stick in the direction of an arrow will cause the game to crash. (Both instances resulted from pressing a silver arrow; unsure if there's any correlation.)

Tested at the moat North of the entrance to Port Windurst in West Sarutabaruta while using the Carp Route for skillups. (One RoE (Reel In Small Fish I) was active, though it probably had nothing to do with the crash.)

I've crashed twice. Both stack traces ended up being identical to each other and both resulted in the culprit being:
"0x012f1cf2: divl %esi,%eax"
Since the last few times we got a crash that involved ASM code, it was on the Test Server with unfinished stuff, _IT MAY BE THAT THIS CRASH IS IN THE WINDOWS VERSION AS WELL_. Further investigation is going to be required and if borne out, the bug should be reported to SE. Chitoo? You out there?

To warrant further investigation, the full stack trace: pastebin.com/5reLTFE3

Good luck. We'll need it. XH
RE: WARNING: New fishing system (March 2014 Version Update) will occasionally crash the game
by Chiitoo on Friday March 21st 2014, 16:11
Interesting.

I had this weird feeling, after seeing the huge arrows in a picture at the Japanese forums before the implementation, that they might be trouble.

I have not had any crash happen yet, but I'll see if I can't lure it out.

I've actually been fishing quite a bit during the past few months myself! Got the two best rods and skill from 45 to 93, so the adjustments were something I was looking forward to quite a bit.


As a side-note, with regards to “moving the mouse cursor off any side of the FINAL FANTASY XI window can often cause the mouse input to endlessly move in that one direction”, I'm wondering if that's not the 'screen-edge panning'? (Config, Mouse/Cam, Screen Edge Panning...)

I can't remember at this time if it's possible to have it 'stuck' like it oft does with Wine, but I would guess so (would need to confirm, however).
RE: WARNING: New fishing system (March 2014 Version Update) will occasionally crash the game
by Megas of Vecanti on Friday March 21st 2014, 16:55
For best replication, try to fish in that moat I talked about and start "Reel In Multiple Small Fish I"; it seems you're almost guaranteed to get that crash by at least midway into the second repeated objective.

Good luck! (Still 20-something and I have to get skill on Dark Bass now...Which isn't happening so I'm considering moving on to the next location. I need me that Shaper's Shawl. /sigh)

Oh wow...I didn't even know that setting existed. I'm ashamed. XD I turned it off; will monitor what happens over the weekend. Thanks!
RE: WARNING: New fishing system (March 2014 Version Update) will occasionally crash the game
by Chiitoo on Thursday March 27th 2014, 7:58
I've still not had this happen to me. I've not yet tried to go after your steps, however. I would need to use another character for it, since my 'main' one has “too high” of a skill so that the fish below level 60-70 take their stamina down by themselves pretty quickly. ^^;

I have, however, been fishing cave cherax and liks a bit, and have yet to experience the crash. Matsya seem to be impossible with my skill level after the change (was easy before, just more rare to catch one than now, it seemed). Perhaps the mooching will help...


Anyblue!

Someone posted this at the official forums:

forum.square-enix.com/ffxi/threads/40909-Fishing-is-broken-and-causing-a-crash

If the post does not exist at some point, it was about a divide-by-zero crash during fishing. I suspect it may well be removed due to the more or less obvious violation of the user/software agreement, and/or terms of service, that may or may not have been occurred to get that data.

Either way, I probably guess that it may indicate the crash indeed happening within the natural habitat, too.

I'll still try to put some time into trying to get it to happen here, and possibly see if a temporary fail-safe could be created.
RE: WARNING: New fishing system (March 2014 Version Update) will occasionally crash the game
by Megas of Vecanti on Thursday March 27th 2014, 11:11
Hmm...SE may actually overlook the means of obtaining the data given the nature of the bug. We'll see.

Either way I suppose this'll be fixed in April at the latest. Good catch! It really does seem WINE is getting to emulate FFXI bug-for-bug now. XD
RE: WARNING: New fishing system (March 2014 Version Update) will occasionally crash the game
by Chiitoo on Thursday March 27th 2014, 11:24
I forgot to do this earlier, but I also checked the Japanese side of the forums. I believe the issue has been reported there as well, and has been moved to 'Confirmed/Planning to Address'.

forum.square-enix.com/ffxi/threads/40814-釣りをしているとエラーで落ちる事がある
RE: WARNING: New fishing system (March 2014 Version Update) will occasionally crash the game
by Megas of Vecanti on Thursday March 27th 2014, 11:31
That conclusively settles it, then. We'll probably have an Emergency Maintenance in the next couple of days. XD

[Close Ticket]
WARNING: New UI on Test Server Client CRASHES FFXI on disclaimer screen!
by Megas of Vecanti on Tuesday August 20th 2013, 23:13
As you may or may not know, the 08/20 Test Server update enabled a preliminary version of the new, revamped PC UI talked about at VanaFest last year. You can enable it by going into the Test Client's ToolsUS/FINAL FANTASY XI Config.exe, scrolling over to the "Misc." tab, and checking the box next to "Enable New UI" at the bottom.

Currently, though, it's unusable on WINE 1.7.0: the game crashes with an Unhandled Exception ("divide by zero in 32-bit code (0x022dd4ba)"; "0x022dd4ba: divl %ecx,%eax" according to the dump below) right after it draws the first new UI element (the right-facing arrow which would be next to the "Accept" option) on the "rules of conduct" disclaimer screen.

Any people in a position to fix the bug will want the full crash dump with symbols enabled, which is here:
pastebin.com/v399Y87G

This obviously needs to get working: good luck to all of you willing to make it work ASAP!
RE: WARNING: New UI on Test Server Client CRASHES FFXI on disclaimer screen!
by Chiitoo on Thursday August 22nd 2013, 6:17
Indeed. I put about 12 hours of staring at relay and other traces as well as the Wine code, and while I am quite confident I know why the crash happens, I haven't been lucky enough to stumble upon a fix. I also don't know enough to fix it myself without some help, and wasn't lucky on that at IRC.

My content ID will become inactive in a day (if it's not already), so I wont be able to test it more. :/

In short, it looks like being pretty much what bug 20847 describes. It looks like it's looking for the Alps Condensed font, uses AR PL UKai CN instead, and there are being zerœs returned (I tried different fonts, even the Alps Condensed, which it wouldn't use even.


Just some quick thoughts before thunderstorm breaks things.

Good luck!
RE: WARNING: New UI on Test Server Client CRASHES FFXI on disclaimer screen!
by Chiitoo on Thursday August 29th 2013, 17:53
Heh, I guess staring at code and backtraces for a few days sometimes yield some sort of results:

bugs.winehq.org/show_bug.cgi?id=20847#c17
RE: WARNING: New UI on Test Server Client CRASHES FFXI on disclaimer screen!
by Megas of Vecanti on Sunday September 1st 2013, 20:33
Wow, nice detective work. That fix should be pushed upstream ASAP; bug the IRC channel. XD
Can I get that fix in the format of a diff/.patch file so we have something to link others to?
RE: WARNING: New UI on Test Server Client CRASHES FFXI on disclaimer screen!
by Pandoraxero on Monday September 2nd 2013, 17:57
Agree. We... or rather... the wine devs... need a patch file to work with.

Megas, Chitoo? the two of you are starting to remind me of myself and Ganiman about 5 years ago... Those were good times... in either case, enjoy your cheesecake.
RE: WARNING: New UI on Test Server Client CRASHES FFXI on disclaimer screen!
by Megas of Vecanti on Tuesday September 3rd 2013, 23:40
Hey, stop acting like this is over. XD It won't be until 20847 is legitimately fixed and the new UI can actually work out-of-the-box. If we can get past this hurdle, it _should_ be the last major change to the FFXI client until such time as SE cuts the PS2 cord and goes full DX9. (And we can concentrate on those remaining graphical flaws which may or may not be WINE's or SE's fault.)

Stay frosty.
RE: WARNING: New UI on Test Server Client CRASHES FFXI on disclaimer screen!
by Pandoraxero on Wednesday September 4th 2013, 9:07
In my experience, it's about the closest thing to nailing a bug. We have the solution, now we need a diff/patch, and need to popularize the diff/patch such that enough people KNOW about it, and it starts gathering attention from the wine devs. That's what Gani and I did... that's the reason you see the ratings on this game jump from garbage/bronze to silver at around this time 5 years ago. I called it cheesecake soon's the third person came in rating it silver.

Unfortunate part with this one, though, is that we don't have a sudden, otherwise inexplicable jump in the game's ratings to draw attention to it. The IRC method may be the most viable.
RE: WARNING: New UI on Test Server Client CRASHES FFXI on disclaimer screen!
by Chiitoo on Monday September 9th 2013, 11:35
Sorry for the late reply.


Making patches is trivial, making a real fix is another story. I don't believe this is a proper way to fix the issue(s), as it's simply forcing a returned 0 into a 1. It doesn't cure the cause, and I highly doubt it would ever go into the Wine tree (that is, I don't think Julliard would approve, and I don't think I would either!).

I guess the actual issue lies within the code of FINAL (or FAINAL, heh) FANTASY XI Online, and I'm not too sure they will ever fix it, as it's likely not affecting Windows. We can find a graceful way to deal with it, something like replicating a bug, or insane behaviour, rather.

I made sure with Austin at IRC that they, too, think that it's probably not worth a new bug as it's very similar indeed to 20847, and they will probably be handled in the same way,.. eventually.

Anyblue, since I stated the patches would be trivial to make, I guess I should deliver some, if you really really want me to! Had to do a bit more since there seems to be regression in git since few days ago, namely the commit:

source.winehq.org/git/wine.git/commit/b28e9f6482bc7baca684d74c9f23bf56d2c6b8e9


I also found an additional crash when opening the job-selection window. Quick trace and just as I guessed, it uses the glyph 3000 as well (added it to the patch).

Furthermore, one additional patch for Gentoo, as the latest versions (including 1.7.1) do not build before this bug is handled:

bugs.gentoo.org/show_bug.cgi?id=482038


Patch to work around the division by zero for wine-1.7.0:

bpaste.net/show/130851/


Patch to revert the msxml changes for wine-1.7.1-158-g88c2a18:

bpaste.net/show/130852/


Patch to revert regression inducing changes for wine-1.7.1-158-g88c2a18:

bpaste.net/show/130853/


Patch to work around the division by zero for wine-1.7.1-158-g88c2a18:

bpaste.net/show/130855/


And finally, an all-in-one for wine-1.7.1-158-g88c2a18:

bpaste.net/show/130854/


Note that I'm not aware what sorts of issues these patches could possibly cause, so do use them at your own peril. ^^;

I did try to make sure to test them all individually, so they should be OK concerning their functionality, but I seem to have a habit of making weird mistakes like forgetting parenthesis somewhere or something like that, So if something doesn't click, apologies for that in advance from Dragoy (of Fenrir)!
RE: WARNING: New UI on Test Server Client CRASHES FFXI on disclaimer screen!
by Chiitoo on Friday September 20th 2013, 11:46
I brought up the “regression in git” matter with the Fellow with Strings and as of:

source.winehq.org/git/wine.git/commit/3fb53e21fb2cccf249ed65a4641eac21422f6609

the:

Patch to revert regression inducing changes for wine-1.7.1-158-g88c2a18:

bpaste.net/show/130853/

is no longer needed.
RE: WARNING: New UI on Test Server Client CRASHES FFXI on disclaimer screen!
by Megas of Vecanti on Sunday October 6th 2013, 19:52
Tested on vanilla 1.7.3-144-g6d03ce8 off of git. Awesome job! It's buggy and slightly crash-prone (in ways that make me suspect that the UI code itself is at fault) but it does work as advertised.

I need to know how we can get the UI font working, though--It defaults to Times New Roman. Is there any way to get WINE to parse that Alps Condensed font internally?
RE: WARNING: New UI on Test Server Client CRASHES FFXI on disclaimer screen!
by Chiitoo on Monday December 30th 2013, 7:53
Toadally forgot I didn't reply to this yet. Sorry!

Firstly, not that I have tested it too much (very few times have I logged on the test-server after the divide by zarrro bug), but the while I did, it didn't crash for me after the fix.

Are there any specific things you experienced that with? I feel as if the new UI might not be working too much different from the old one in ways, as there's quite the performance drop when it's visible compared to how the game is running while it's hidden.

I should like to test it on Windows, but I've been too lazy fer that.


As for the font, I'm not entirely sure what it would require for Wine to use it. I merely tried copying it in, and even removing the ones it looked like it was substituting it with, but it would simply use another one instead (perhaps it didn't seem like a valid candidate for some reason).

I didn't look into that any more than that, and I haven't been hanging out in the IRC lately, but I'll definitely make a note of it here if I ever get to it!
RE: WARNING: New UI on Test Server Client CRASHES FFXI on disclaimer screen!
by Megas of Vecanti on Monday December 30th 2013, 10:04
Yeah, the only issue for me is the font--the internal Alps Condensed is completely ignored and the new UI defaults to 8px Times New Roman. It seems to work fine except for that.

The font that WINE saddles the UI with is barely readable. It really warrants investigation. Haven't tested the UI extensively after the recent Test Server update, though, so at the moment I'm unaware of any new issues.

(If all goes well, BTW, I'll be submitting a new general test result in a few days.)
RE: WARNING: New UI on Test Server Client CRASHES FFXI on disclaimer screen!
by Chiitoo on Thursday January 2nd 2014, 16:13
Well, against my beliefs, I've solved the mystery. Or so I maybe think.

I'll try not to go on for pages about what all I did during the few days I looked into it, as it was very much a learning experience; new grounds for me, so to speak.

So... after I confirmed that no matter what I did, 'AR PL UKai CN' was chosen for me, as was the first time I looked into this a bit, and if that was not available, then another one from the arphic fonts was used. It was chosen even if I forced substitution via the registry (more on this later on).

When I noticed that, I started to look more into the 'Alps Condensed' font itself. Fontforge immediately reported various errors, ones that which might cause Windows to ignore the font altogether, so I fixed things here and there to no avail.

As I was testing the font in Notepad, where it worked(!), I noticed that the charset that was requested for it was 0 there, while FFXI requests 128. As one might suspect, this seems to refer to SHIFT-JIS (if I understand things right).

See where I am going with this yet?

Earlier on, I was looking at the images at the official forums ( forum.square-enix.com/ffxi/threads/36908-dev1156-New-User-Interface ), and comparing the font in them to 'Alps'. It was quite clear that in those images, the font is not being used either! The 'G' and 'R' for example give it away very quickly.

I suspect Windows rejects the font as well, since it has no support for the 128 charset (or it's broken; can't tell for the time being). It sort of looks like a mix of 'Arial' (Arial Unicode?) and 'Times New Roman'... or something like that.

I believe I confirmed this all by hacking on the Wine 'freetype.c' a little bit, making it so that if it sees the particular font, and codeset 128, the DEFAULT_CHARSET would be forced.
This allowed 'Alps Condensed' to be /finally/ used, but any Japanese that still is around in some menus would be seen as 'blocks', as I more or less expected.

It's possible to substitute 'Alps Condensed' without altering any actual code, as long as the font used has support for the 128 charset, such as 'MS Gothic'. Earlier I had only tried 'DejaVu Sans', which was the first and pretty much only font that came to my mind without thinking about it further.

For example:

[HKEY_LOCAL_MACHINE\Software\Microsoft\Windows NT\CurrentVersion\FontSubstitutes]
"Alps Condensed"="MS Gothic"


So all in all, it would seem to me as if Wine is actually doing things correctly here. Which font it chooses varies with what is available (arphic fonts likely being on the very top). I guess they haven't decided on the font at the SqEX side yet.

Japanese character display is of course desirable due to the fact that there be Japanese players around on the same servers.
I guess they should look into another font... though who knows; perhaps one will be able to select from several, but I guess I might even report this at the official forums. Not that they'll think much of it, I'm sure.


Meh!

That's all fer now, methinks.
RE: WARNING: New UI on Test Server Client CRASHES FFXI on disclaimer screen!
by Megas of Vecanti on Thursday January 2nd 2014, 17:13
Interesting stuff. MS Gothic seems to be close to what SE was using in the screenshots, though the antialiasing makes it a tiny bit difficult to read (regardless, a massive step up from AR PL UKai CN!). Otherwise I think you've nailed it. Thanks!

BTW: for people who don't want to or can't download Windows corefonts, monafont ( monafont.sourceforge.net/index-e.html ) seems like a suitable alternative as it's a spitting image of MS Gothic (probably by necessity, as monafont's targeted specifically at recreating 2ch ASCII art).

Also: is it just me or did the new UI reduce the framerate by about 10 fps this update...?
RE: WARNING: New UI on Test Server Client CRASHES FFXI on disclaimer screen!
by Chiitoo on Thursday January 2nd 2014, 19:30
Aye, there should be quite a lot of options out there for CJK fonts.

I was going through what I happened to have, and I must say 'mikachan' looks pretty fun, and rather fitting!

I can't say if the FPS is worse than before. It has been a while since I visited after all, and I didn't spend much time with it. I can say, however, that similar to the old, current UI, there's a big impact on the CPU use and FPS coming from the UI in general.

With this new one it's rather big, too.

I'm wondering if it can actually be font-related'ish. I noticed that the font debug channel alone is able to bring the game to a crawl. One time I counted about 24 lines being repeated over and over so fast that the file was over 1 million lines long in... 30-60 seconds I believe it was.

I tried the old UI and some Counter-Strike: Source to compare, and while they do similar things, they do it quite a bit less. I do wonder if there might be a place for a performance improvement there, but I can't tell. It could be something d3d-related just as well.


By the by, opening key-items that haven't been checked before produces a segmentation fault here! I might guess it could have something to do with nvidia-drivers, as I seem to be having such an issue with Terraria as well with new enough drivers, but a quick relay debug look-see shows something about 'want to examine the key item?' and then speak of a 'data/ffxi_menu_window.png', which I can't seem to find...

A missing file altogether?

Hmmm...
RE: WARNING: New UI on Test Server Client CRASHES FFXI on disclaimer screen!
by Megas of Vecanti on Thursday January 2nd 2014, 19:54
It's altogether possible that the framerate reduction is the result of font issues. I can't personally tell at the moment.

And yes, I just recreated the crash and the backtrace points to "0x040a0000: movb 0x1(%ebp),%dl" as the culprit. If you have something similar then I can assume it's the same thing.

(And BTW: unless you have something to decompress and search the .dat files, most likely you won't find that image. I doubt SE would leave anything out in the open.)
RE: WARNING: New UI on Test Server Client CRASHES FFXI on disclaimer screen!
by Chiitoo on Thursday January 2nd 2014, 20:21
I don't actually get a backtrace as you seem to. It just very quickly segmentation faults away before anything like that is produced.

As for the files, yeah, I don't tend to do /that/ sort of mining, but the references seemed to be to an actual png-file, possibly under one of the data paths under PlayOnlineViewer installation path where one can find some menu files in png form, though they don't quite seem readable.

You might be correct though.

I also did a file-check to be safe on having all the required files, but no problems were found.
RE: WARNING: New UI on Test Server Client CRASHES FFXI on disclaimer screen!
by Chiitoo on Friday January 3rd 2014, 10:43
Confirmed key item crash on Windows, as well as Alps font not being used even if it is installed.

Also noticed that, in my mog house, FPS would go downwards to 20 or even slightly below (it's literally full (80 items) of event items from holiday trees to the spinet and nearly all of the statues etc., which is nice to see as I had imagined it to be a Wine thing mostly.

When playing with Wine though, I tend to force the CPU to do things a bit more so I actually might get better performance when using Linux, it seems. ^^
It's been about 3-4 years since I last played FFXI on Windows, so I wouldn't remember (I believe last I booted Windows at all was on 15th of June; but it was useful today, although it failed updating itself, but I digress).


So, sum-summarum: Windows does not use the Alps font, 'key item crash' is not Wine related (working “as intended”), and performance drop induced by the UI is also very noticeable when running natively on Windows.
RE: WARNING: New UI on Test Server Client CRASHES FFXI on disclaimer screen!
by Megas of Vecanti on Friday January 3rd 2014, 20:34
Oh wow. Yeah, I guess both the crash _and_ the font issue need to be reported to SE on the official forums if they haven't already.

It's also gratifying to see WINE maintain a kind of performance edge. I'd long ago noted the difference in login speed between the few times that I'd logged in on native Windows and WINE--there's like a second-and-a-half reduction on WINE; barely noticeable but nevertheless there.

I guess that wraps things up until the next Test Server update. Kudos all around, man! Thanks for taking the time to boot into native Windows. XD

(BTW, there's actually _one_ issue that needs explanation: the Opening Movie. Have you ever gotten it to work? Supposedly it works if you use native d3dx9. Tried it and no dice.)
RE: WARNING: New UI on Test Server Client CRASHES FFXI on disclaimer screen!
by Chiitoo on Friday January 3rd 2014, 20:56
I didn't see it reported yet; not even on the Japanese side, and I only looked at some of the more recent posts the last time I checked. I remembered the recent update doing something with the key items, but I can't remember why, or what (probably nothing).

Will look more into it sooner or later.

(As for the Opening Movie, I can't remember my experiences playing it via Wine at all... Will definitely look into it, hopefully sooner than later, and will post a comment if I can find anything interesting at all.)
RE: WARNING: New UI on Test Server Client CRASHES FFXI on disclaimer screen!
by Chiitoo on Wednesday October 2nd 2013, 17:41
As of

source.winehq.org/git/wine.git/commit/db86409db66362c56031429ec124e65064e4f0bd

this issue (division by zero) should be (properly) fixed, and from a quick test, it looks like it is.

Fer now, at least. ^^;
Gentoo use flags?
by Claquesous on Thursday June 27th 2013, 0:24
Can someone who is successfully using Gentoo amd64 please post their emerge output?

This is my output, but PlayOnline crashes before it even loads the GUI. The error seems to be out of dsound.

[ebuild U ~] app-emulation/wine-1.6_rc3 [1.5.10-r1] USE="X alsa fontconfig* gecko jpeg mp3 ncurses opengl perl png prelink run-exes%* samba ssl threads truetype udisks* xml* -capi -cups -custom-cflags -gphoto2 -gsm -gstreamer -lcms -ldap -mono* -nls -odbc -openal -opencl -osmesa -oss -pulseaudio -scanner (-selinux) {-test} -v4l -xcomposite -xinerama (-gnutls%) (-win32%*) (-win64%)" ABI_X86="32%* 64%* (-x32)" LINGUAS="en%* -ar% -bg% -ca% -cs% -da% -de% -el% -en_US% -eo% -es% -fa% -fi% -fr% -he% -hi% -hr% -hu% -it% -ja% -ko% -lt% -ml% -nb_NO% -nl% -or% -pa% -pl% -pt_BR% -pt_PT% -rm% -ro% -ru% -sk% -sl% -sr_RS@cyrillic% -sr_RS@latin% -sv% -te% -th% -tr% -uk% -wa% -zh_CN% -zh_TW%" 0 kB
RE: Gentoo use flags?
by Megas of Vecanti on Thursday June 27th 2013, 16:48
The emerge output would typically only help you in the event of a build failure. That isn't the case here--the issue here is most likely with the emulation, which means you need debug output from WINE itself.

Please enable debug symbols by rebuilding your current version of WINE with:

FEATURES="nostrip" emerge wine

Once it's rebuilt, run POL, request the debug output from the "pol.exe has failed and has to close" window as soon as it crashes. Copy the _full text_ of that output into the "New Paste" text field on pastebin.com and click Submit. Copy the link to the file you just made and paste it here. (Copying the entire debug output into a reply can flood out other messages and is advised against!)

The Backtrace section will most likely tell us where things went wrong, but the other sections may also hold clues.

Good luck!
RE: Gentoo use flags?
by Claquesous on Tuesday August 13th 2013, 19:36
I finally found some time to do this. Here's the link:

pastebin.com/eFps8wjx
RE: Gentoo use flags?
by Megas of Vecanti on Saturday August 17th 2013, 14:54
The error that causes the unhandled exception is:

>0x7d7f1545 DSOUND_ReopenDevice+0xaa5 in dsound: divl %ecx,%eax

...So you're correct. The problem seems to be occurring when dsound tries to initially open the device and fails.

There are several ways you can try to fix this:

-Run winecfg _from the console_. From the Default Settings selection, confirm you can play the Test Sound from your current device. _When pressing the Test Sound button, note any errors in the console._ If the sound doesn't play it may be the Output Device you're using. Check your audio drivers, or try another device if available. Post any command line errors here if you get any.

Do the same from the pol.exe application settings, just to make sure. (You shouldn't typically have to change anything here--chances are that if the default audio settings aren't working, they probably aren't working with other WINE applications either.)

-In winecfg, check the Windows Version set for pol.exe. It should be "Windows 7." Set it to that if it isn't. (This may not have anything to do with dsound, but it might help!)

-In winecfg, Check Libraries from the Default Settings selection. If there's an existing override for dsound, check to see whether it's set to native or builtin--FFXI seems to play nicer with WINE's own builtin .dlls. If it is _not_ set to builtin in Default Settings, go into the pol.exe application settings and, if one doesn't exist, create an override for dsound. Set it to "builtin." (If one exists, set that to "builtin" instead.)

-If all else fails (NOTE: YOU SHOULD NORMALLY NOT _EVER_ HAVE TO DO THIS!): ensure all dsound overrides on pol.exe are removed, install the latest winetricks script and issue the following command:

winetricks d3dx9

This will install DirectX9 and make that version of DirectSound the default across all applications without overrides. See if that works.

Good luck!
RE: Gentoo use flags?
by Claquesous on Sunday August 25th 2013, 0:36
I took your suggestion and played with my audio settings in winecfg.

My audio settings seem to be a bit unique since I connect my PC directly to my TV via HDMI. This was something that took me a while to configure under Gentoo originally.

Anyway, I tried setting it I believe to HDMI 0, and lo and behold the game just loaded instead of my customary immediate dump. I then turned it back to System Default and it croaked again. I closed winecfg and loaded the game and logged in to make sure everything works (which it does).

The funny thing is after logging in I went back to winecfg and my audio is set to System Default and the option I chose isn't there anymore. As long as the game loads I don't mind.

I do notice at least one controller bug has been fixed, the audio seems to have less hiccups, and I'm sure there are many other fixes too, so I greatly appreciate the help finally upgrading to the latest and greatest wine.
Present Status - Glitches and Workarounds
by Pandoraxero on Thursday May 30th 2013, 10:34
The logout/shutdown crash glitch
-I've managed to isolate the cause of this glitch to something easily user-controllable... namely, the "show opening movie" option on the ffxi config. I have no idea why, but somehow, disabling this option prevents the crash from happening.

Crash after Disclaimer
-If FFXI loads, but won't get past the little disclaimer thing without crashing... you MIGHT be in WinXP mode. Try setting Win7 mode. It should work then

Installation
-If you're having problems with installation, specifically, if the installer is giving you a "1628: Failed to complete installation" error message, your WINEARCH is likely set to win64. This seems to be the default for 64-bit systems. To fix this, backup or delete your current ~/.wine and run WINEARCH=win32 winecfg. This will setup a new prefix that will run the 32-bit installers without flaw.

-In addition, the installers all seem to throw the same error messages to console towards the end, namely:

err:ole:ClientRpcChannelBuffer_SendReceive called from wrong apartment, should have been 0x80000002a
err:ole:dispatch_rpc no apartment found for ipid {ffffffff-ffff-ffff-2800-000008000000}
err:rpc:I_RpcReceive we got fault packet with status 0x80010108
err:ole:dispatch_rpc no apartment found for ipid {ffffffff-ffff-ffff-2800-000008000000}
err:rpc:I_RpcReceive we got fault packet with status 0x80010108
err:ole:dispatch_rpc no apartment found for ipid {ffffffff-ffff-ffff-2800-000008000000}
err:rpc:I_RpcReceive we got fault packet with status 0x80010108
err:ole:dispatch_rpc no apartment found for ipid {ffffffff-ffff-ffff-2800-000008000000}
err:rpc:I_RpcReceive we got fault packet with status 0x80010108
err:ole:dispatch_rpc no apartment found for ipid {ffffffff-ffff-ffff-2800-000008000000}
err:rpc:I_RpcReceive we got fault packet with status 0x80010108

...these messages do not seem to affect the install itself, as the game runs as well as it would on a similar windows system

Graphical Glitches:
-Specifically, the only graphical glitches that have stuck out to me are things like glow boxes (like fireplaces, or the entrances to the Bastok MH at night) extending beyond their normal visible bounds (does not affect gameplay), and other little things, like that little fountain kind of across from the L.Jeuno AH. Its "Water" displays outside of its normal bounds, too... but has no effect on gameplay.

-----
Please, if you are testing this game, bear all this in mind... especially the win32 thing. This game does not deserve the garbage ratings it's been getting, just like the number 2 game on the top 25 doesn't deserve the "Gold" rating it's got. I seriously can't eat my cheesecake in peace knowing this game's fallen so far.
RE: Present Status - Glitches and Workarounds
by Megas of Vecanti on Wednesday June 5th 2013, 11:00
I've mentioned before that setting a native msctf.dll resolves the shutdown/logout crash for me. I've had "Display Opening Movie" enabled at the same time without any issues.

(Though the opening movie still doesn't display for some reason; this isn't something I can pin down since I'm not certain _when_ the opening movie is supposed to display on the PC version. A while back someone mentioned that some version of native DirectX through winetricks makes the opening movie actually work, but I've never seen that happen.)

The main issues that stand out to me now are the glow containers and, more importantly, the unfiltered textures: it's annoying to have to be in Adoulin at night and all of the street lamp glow decals look like they're 10x10. That _can't_ be right...
RE: Present Status - Glitches and Workarounds
by Pandoraxero on Wednesday June 5th 2013, 12:42
I tend to go for minimum overrides. It's better, in my opinion to have Wine run with NO overrides at all. And that's what I've managed to get. That's really the intent of Wine: to pretty much invalidate Windows' reason for existing as a desktop system. clearly, msctf and the dx subsystems need work. ...converting calls from DX to GL is not exactly an easy task to accomplish, either, and, frankly, I'm surprised wine is able to do it as accurately as it does in the first place.

Odds are, though, if there's a problem, the solution is probably something GLARINGLY SIMPLE. We solved the imm/ime bug in 2008 with a relatively simple sanity patch, which got committed sometime in 2009, I believe. Irony of that situation was, the cause of the problem back then, and the course of action needed to fix it were all outlined in simple terms on MSDN.

Clearly, a number of things have improved with the Wine project in general since 2008 when I came on the scene... or at least, the project in regards to FFXI.

The opening movie is unimportant to gameplay, anyways, but I'm pretty sure when I first played (on windows), it came on BEFORE the little disclaimer screen thing.

I'll look into the glow decals in Adoulin. I've already submitted two images showing box flaws (one glow box and one fountain/water box)

Really, the more images we can get that highlight the graphical flaws, the better. So if you can get a screencap of it, and submit that, it'd help a little.
RE: Present Status - Glitches and Workarounds
by Megas of Vecanti on Thursday June 6th 2013, 19:33
Hey, I just made this imgur album with a decent selection of the most prominent texture filtering issues. Hope This Helps:

imgur.com/a/ID9kb

I can totally understand why it's important to have an app running without DLL overrides (it's one of the official requirements for Platinum, after all) but from a realistic standpoint, I also want the game to work in the meantime--and it tends to be a very long time before WINE devs get around to sorting most builtin issues out. =\

And I understand that getting the Opening Movie isn't necessary to make the game playable, but it _is_ a feature of the application and therefore one that should be factored into proper emulation in the long run--plus, whatever's blocking the Opening Movie may be blocking other apps as well. "A rising tide favors all boats" etc..
RE: Present Status - Glitches and Workarounds
by Chiitoo on Thursday June 6th 2013, 20:46
I can't say for sure, but I think at least the

i.imgur.com/3xvLkEP.jpg

is actually how it is supposed to be (as in, you see the same while running the game in its native environment).
With that in mind, I think there is a good chance that many, most, if not all of that can be re-produced in Windows.

Have you tried that?

At least I seem to remember seeing the block-pattern way before moving on to Linux some 3 years ago.
My memory might well fail me though!


I'll try to look into the opening cinematic some time soon (I hope), and see if I can dig something up.
RE: Present Status - Glitches and Workarounds
by Megas of Vecanti on Friday June 7th 2013, 14:26
Well, I don't run Windows natively, but if these filtering issues happen on Windows as well...I suppose it just can't be helped, then. =\ Really not good.
RE: Present Status - Glitches and Workarounds
by Pandoraxero on Friday June 7th 2013, 19:36
Well. I know the thing with the clouds is in windows. As far as the lamps/lighting ... or the Vunkerl cloud textures... Never really paid attention to them on windows.
RE: Present Status - Glitches and Workarounds
by Pandoraxero on Thursday June 6th 2013, 20:56
Hm. I see your point on the filtering... The clouds, though, I've noticed that same thing looking UP at the clouds in Konschtat back in 2005... on Windows. So that's more of a game issue and less of a wine issue. I'll have to get access to Adoulin and swap out my drives so I can see if the same thing happens in Windows with the light decals, though.

And I see where you're coming from on the movie thing, too. There's two important factors regarding that particular issue. The first, and most glaring, is the crash if no DLL overrides are present and "Show Opening Movie" is ticked. The second, while not important to gameplay for this game in particular, is that the video doesn't display without very specific settings... and you're right in saying that it could negatively impact playability of other games... or worse, usability of other apps in general.

The imm/ime issue in 2008 was a simple fix. This one, though... totally different story. It's clear we're dealing with an issue across multiple dlls.

I'd hesitate to call it a cascading failure, too, though. ...But the only way to determine if it IS a cascading failure is to somehow get the video to play, WITHOUT the msctf override. If the video can be played WITHOUT that override, but the game still crashes on /shutdown, then there are two independent, connected, but non-cascading failures in the builtins. On the other hand, if the video plays AND /shutdown or logout works properly, then it's a localized problem with the DLLs responsible for playing the video - and a cascading failure.

I'm a network engineer by trade, so I understand all about cascading failures. wine is notably unstable... and I don't mean that in a BAD way, I just mean everything is under constant development. It's possible this could be an issue with BOTH system libs. on the other hand, it could be an issue with one, that's causing an issue with another, that's causing the game/POL to crash on exit - a cascading failure... I'll have to look into it sometime.
A few odd items and a thought
by Astrit Prower on Sunday January 18th 2009, 1:42
Amongst my latest test I found a few interesting items. The first was the graphics. I don't know how but FFXI looks crisper in WINE than a real windows install, not to mention faster loading. The POL viewer doesn't crash when I try to access my mail in WINE.

Again I don't know what it is, but some windows apps are running better in WINE the windows. This also applies for other apps I maintain.
Multiple instances of FFXI
by Peter Andersson on Thursday October 23rd 2008, 9:05
Hello everyone.

Today I added a small guide on how to run more than one instance of FFXI at the same time on the same computer.

If you are one of all the people that like to play more than one character at a time I believe you will find this most useful.

Right now there are several draw-backs with the guide and it's methods. They are, at best, crude. I will keep on improving them.

Main problem right now is that it seems there is no way around the fact that you need a separate copy of the new instance. This is because wine seems to put a virtual lock on the files of the first instance.

An optimized way to export the registry data would be to know exactly what data to export and then just process those registry paths. I will see what I can dig up. Maybe someone here knows exactly what needs to be exported.

All the best.

Peter
RE: Multiple instances of FFXI
by Peter Andersson on Thursday October 23rd 2008, 9:12
I should also add that I pruned the Install/Config guide to reflect recent progress.

Also performed some tests on a Intel C2D @3GHz 6GB of RAM and a GeForce 8600 GT GPU on Ubuntu Hardy. Ran two instances of FFXI on a dual head system. The result is very pleasing. I am happy to report that running one or two instances made more or less no difference in performance.


Take care,

Peter
Necessary hacks.
by Pandoraxero on Saturday August 30th 2008, 0:11
I've been working somewhat in collaboration with ganiman today. i posted my data on ffxiclopedia forums, and allakhazam. Here's a condensed version:

- POL and FFXI installations should work perfectly.
- Set POL to run in a window and edit the registry so that the window dimensions are 640x480 (C:\\Program\ Files\\PlayOnline\\SquareEnix\\PlayOnlineViewer\\polcfg.exe, HKLM\\Software\\PlayOnline[US|EU]\\SquareEnix\\PlayOnlineViewer\\Settings)
- when launching POL before or after the update, you may need to register a copy of dxdiagn.dll from a windows xp install and set dlloverride to native
- after the ffxi update, it will most likely fail to launch, what you need to do is open the FFXI config utility and set ffxi to windowed mode, then go to the FFXI section of the registry and edit 0001 and 0003 to your desired ffxi window width, and 0002 and 0004 to height. (C:\\Program\ Files\\PlayOnline\\SquareEnix\\FINAL\ FANTASY\ XI\\Tools[US|EU]\\FINAL\ FANTASY\ XI\ Config.exe, HKLM\\Software\\PlayOnline[US|EU]\\SquareEnix\\FinalFantasyXI)
- if there's a problem connecting to lobby server, and the problem's persistent, you need to set your hostname to localhost
RE: Necessary hacks.
by Pandoraxero on Saturday August 30th 2008, 0:14
Oh yes. forgot to mention, you can get it to run at full speed by selecting yourself (F1) and opening a menu (Enter) and entering a chat dialog with yourself (Enter)

After that, you can get regular keyboard input by activating a macro with a chat line. the default TP macro should do.
Back