AppDB: Rating / Patching
Hi, I started a discussion inside AppDB about the in my eyes "strange" Gold rating of Fallout 3 (it's actually just an example, other entries suffer the same problems). I know there was a discussion about the rating on this list last month, but as far as I could see my particular issue wasn't covered and I only just subscribed. So here goes a new thread. The appdb thread [1] is included below (prevent OT-cleanup deletion inside appdb). I don't expect anyone to read it completely, but I didn't want to repeat every argument again either. ;) The basic point is: Fallout 3 (a game) only works with a small -- but nevertheless -- patch, otherwise it will crash, no matter what dll-overrides/settings/3rd-party apps.
From how I understand the wording in [2], an app that requires a patch to run, can't get a gold rating. In fact, if there is no way to get it working in a vanilla wine release, then there is no other option than "Garbage" IMO. The various arguments for that assumption (possible breaking of other apps run by the same wine installation, regression tracking, etc.) are in the quote below.
The maintainer tried to convince me that a gold rating is valid, because Fallout 3 works great with the patch applied. Well obviously he failed, and the discussion went away from Fallout in particular to a more general interpretation of [2]. I'm neither saying I'm right and he's wrong, nor the other way around. IMO both interpretation can be valid, depending on how you read "some DLL overrides, other settings or third party software." My suggestion now would be: * Clarify the wording on what "other settings" really means (my interpretation is mainly registry modifications with winecfg and/or regedit). * Add an explicit statement about patched wine versions. Something like: - Any application that requires a patched wine to run MUST NOT be rated higher than "Garbage" (or whatever rating was intended for this situation). [1] http://appdb.winehq.org/objectManager.php?sClass=version&iId=14322#Comment-4... [2] http://appdb.winehq.org/help/?sTopic=maintainer_ratings Björn ==== Copy from AppDB discussion [1] below ==== Björn Krombholz on Monday January 5th 2009, 10:10: Hi, I don't understand the recent ratings. I've played the game a lot under Windows and with various versions of wine, using all kinds of different tweaks and settings the game itself and wine offer. It runs pretty well and it's become even better (meaning faster) since 1.11 (no evidence, could be caused by a driver update as well). Nevertheless, as far as I can see, noone has ever managed to get past menu + the loading screen without at least applying the D3D patch attached to [1], that allows setting the VideoDescription and VideoDriver in the registry. For flawless mouse interaction, you also need the mouse hack from [2]. Considering these 2 as a fact (please correct me, if there actually is a way to run it without patches), how can it ever be rated higher than "Garbage"? My understanding of the ratings are, that they offer a quick hint on what you are able to run (nearly) without problems, while the discussion and howto sections deal with "broken" stuff and tell a user, how he might be able to run an app/game despite the garbage state. [1] http://bugs.winehq.org/show_bug.cgi?id=15839 [2] http://bugs.winehq.org/show_bug.cgi?id=6971 ----- RE: Rating / Patching by Tymoteusz Paul on Monday January 5th 2009, 10:20 Let me explain you how ranking system works, because you seem to have it confused. As expalined in http://appdb.winehq.org/help/?sTopic=maintainer_ratings gold/silver rating is ok as long as you can use it for it's purpose. All you have to do is override couple dlls and settings and then the game is working pretty much flawlessy. Garbage rating would be then if it was impossible to play the game. Ratings as is provide good info about the game, if you see that its rated as gold - then you are safe to buy it for linux playig, it may take some work to get it work but it will work and will be playable. Only rating that fits your description "hint on what you are able to run (nearly) without problems" fits platinum rating, which this game didn't get so far. Hope this will be more clear for you now. ----- RE: Rating / Patching by Björn Krombholz on Monday January 5th 2009, 14:08 I was referring to this help text as well, although I read it differently. ;) In any vanilla/clean release version of wine, Fallout 3 won't run. You get the menu, can change settings, but when you start the game to play ("purpose it was designed") it crashes, even "with some DLL overrides, other settings or third party software" you install in wine. => Garbage rating. Let's imagine the small driver patch in bug 15839 was applied to the development tree and so released with 1.1.13. The game would become playable with one major "issue". No matter what settings you change/etc. the mouse is either always stuck in the middle of the screen or it gets stuck at the screen edges (bug 9671). => Bronze rating for wine >= 1.1.13. Maybe even Silver, as it might work very well with a different kind of controller like a gamepad. Now the mouse bug gets fixed in 1.1.x. Suddenly all you need to do is to override some (actually just one) dlls and tweak some settings to make it enjoyable. => Silver (in case you consider Live functionality an important part of the program) or Gold rating. Eventually someone adds the d3dx9_38.dll to wine and the DeviceDriver + DeviceDescription settings will be automatically detected by wine. That means the game works flawlessly without touching any wine setting. => Platinum rating. IMO this makes sense, otherwise a version 1.1.x gold app, wouldn't be downgradable in its rating for 1.1.x+1 (in case version x+1 introduced a regression that breaks this app); because all you had to do to make it work again is unapply the faulty patch. In addition to that, users wouldn't be encouraged to file a bug report if there is source-patch that upgrades it to gold anyway. Anyway, I don't care much about which interpretation should be used, I just want it to be more consistent. As you can see in the reports history Fallout 3 people rated in both ways. E.g., the newest one voting gold, because it works great after patching according to the HOWTO; http://appdb.winehq.org/objectManager.php?sClass=version&iId=14322&iTestingI... rating "Garbage" because "Mouse is borked and gameplay is impossible without patches. Which would be the best way to ask for a more indisputable explanation on the help page? Filing a bug report against appdb or using the "Email Us" link at the right, or ...? ----- RE: Rating / Patching by Radosław Ciechowski on Monday January 5th 2009, 15:15 Gold is for apps that "Works flawlessly with DLL overrides, third party software or other settings" so if Fallout 3 is working without bugs after applaying patches, dll-s and other settings it should be rated Gold. If there was no way to get it working or there were lots of bugs after all tweaks, then it could be rated Garbage. If game works flawlessly then no matter how mamy things you have to do, it can be rated Gold which makes sense for me. There is no sense in rating apps Garbage when they can work flawlessly ! Also there is already rating system which we are using ( http://appdb.winehq.org/help/?sTopic=maintainer_ratings ) and changing it now wouldnt be a good idea. ----- RE: Rating / Patching by Björn Krombholz on Monday January 5th 2009, 15:55
Gold is for apps that "Works flawlessly with DLL overrides, third party software or other settings" <
Indeed, that's what I'm saying. But applying a patch to the wine source and recompiling it -- which effectively makes it some _different_ version than the _release_ version you are supposed to rate -- is neither of those 3. - "DLL override": installing a custom DLL into $WINEPREFIX and adding the override) - "third party software": installing additional software into $WINEPREFIX that is needed and expected by the app you want to run - "other settings": tweak the wine registry and similar configuration None of these change the wine version, any such change would still be a "wine 1.1.x", while patches like the mouse patch might even break other software even if those are installed in a different $WINEPREFIX, because the binary changed and it's behaviour is different from the official release. You can't say "A is running well on wine 1.1.x", if in fact it is running well on wine 1.1.x+patch, while it actually breaks on wine 1.1.x. And yes, as I already said, there is a help page for the rating system. I don't ask to change, but to clarify it, as at the moment people are rating differently depending on their interpretation. This has become OT on this page - I will try to move it to the -devel list. :)
Björn Krombholz wrote:
Hi,
I started a discussion inside AppDB about the in my eyes "strange" Gold rating of Fallout 3 (it's actually just an example, other entries suffer the same problems). I know there was a discussion about the rating on this list last month, but as far as I could see my particular issue wasn't covered and I only just subscribed. So here goes a new thread.
The appdb thread [1] is included below (prevent OT-cleanup deletion inside appdb). I don't expect anyone to read it completely, but I didn't want to repeat every argument again either. ;)
The basic point is: Fallout 3 (a game) only works with a small -- but nevertheless -- patch, otherwise it will crash, no matter what dll-overrides/settings/3rd-party apps.
From how I understand the wording in [2], an app that requires a patch to run, can't get a gold rating. In fact, if there is no way to get it working in a vanilla wine release, then there is no other option than "Garbage" IMO. The various arguments for that assumption (possible breaking of other apps run by the same wine installation, regression tracking, etc.) are in the quote below.
The maintainer tried to convince me that a gold rating is valid, because Fallout 3 works great with the patch applied. Well obviously he failed, and the discussion went away from Fallout in particular to a more general interpretation of [2]. I'm neither saying I'm right and he's wrong, nor the other way around. IMO both interpretation can be valid, depending on how you read "some DLL overrides, other settings or third party software."
My suggestion now would be: * Clarify the wording on what "other settings" really means (my interpretation is mainly registry modifications with winecfg and/or regedit).
* Add an explicit statement about patched wine versions. Something like: - Any application that requires a patched wine to run MUST NOT be rated higher than "Garbage" (or whatever rating was intended for this situation).
[1] http://appdb.winehq.org/objectManager.php?sClass=version&iId=14322#Comment-4... [2] http://appdb.winehq.org/help/?sTopic=maintainer_ratings
Björn
Here is the rating system as I see it and most others: Platinum: -- ALL functions are as they are in Windows with maybe minor usage difficulties WITHOUT change to Wine. Gold -- ALL functions work as they do in Windows with only replacement of known broken dynamically linked library files. No code changes are acceptable. Patches, unless accepted by AJ, are not a reason to rate a program with this status. Silver -- MAJOR functions work as they do in Windows with only replacement of known broken dlls. Garbage -- Program does not function with Wine even if broken dlls are replaced. Needs major patching and/or repair work. Sounds like you have the right opinion, the program should be rated Garbage as it does not work with Wine, even upon replacement of broken dlls. James McKenzie
2009/1/6 Björn Krombholz <fox.box(a)gmail.com>:
Hi,
I started a discussion inside AppDB about the in my eyes "strange" Gold rating of Fallout 3
--snip--
The basic point is: Fallout 3 (a game) only works with a small -- but nevertheless -- patch, otherwise it will crash, no matter what dll-overrides/settings/3rd-party apps. -- snip -- From how I understand the wording in [2], an app that requires a patch to run, can't get a gold rating. In fact, if there is no way to get it working in a vanilla wine release, then there is no other option than "Garbage" IMO. The various arguments for that assumption (possible breaking of other apps run by the same wine installation, regression tracking, etc.) are in the quote below.
That's simply not true... Gold means you're either using native dlls, have modified the program by patching it with nocd or you've modified Wine to make it work, ie there is a work around that makes the application work flawlessly. There's no reason to exclude modifying Wine, you are empowered to change it as you see fit since it's free software. -Jeff
On Mon, Jan 5, 2009 at 8:09 PM, Jeff Zaroyko <jeffzaroyko(a)gmail.com> wrote:
2009/1/6 Björn Krombholz <fox.box(a)gmail.com>: Gold means [...] you've modified Wine to make it work, ie there is a work around that makes the application work flawlessly. There's no reason to exclude modifying Wine, you are empowered to change it as you see fit since it's free software.
This point is where the disagreement lies. Let me be very specific. The argument here is not that patches should disqualify Gold ratings. It is that patches change the version of wine that you are running. Specifically, the people giving "Fallout 3 on wine 1.1.12" a Gold rating are not actually running wine 1.1.12, they are running wine [snapshot/git]+patch. "Fallout 3 on wine [snapshot/git]+patch" deserves a Gold rating. "Fallout 3 on wine 1.1.12" deserves a Garbage rating, because it Does Not Work no matter what you change within the confines of wine 1.1.12.
Sometimes to make an app work, you need to copy over some native dlls. To get these dlls don't you need to own a copy of windows? Could this be a criterion in the rating system? Wether or not you need to own a copy of windows? nick *********************************************************************** WARNING: This e-mail (including any attachments) may contain legally privileged, confidential or private information and may be protected by copyright. You may only use it if you are the person(s) it was intended to be sent to and if you use it in an authorised way. No one is allowed to use, review, alter, transmit, disclose, distribute, print or copy this e-mail without appropriate authority. If this e-mail was not intended for you and was sent to you by mistake, please telephone or e-mail me immediately, destroy any hardcopies of this e-mail and delete it and any copies of it from your computer system. Any right which the sender may have under copyright law, and any legal privilege and confidentiality attached to this e-mail is not waived or destroyed by that mistake. It is your responsibility to ensure that this e-mail does not contain and is not affected by computer viruses, defects or interference by third parties or replication problems (including incompatibility with your computer system). Opinions contained in this e-mail do not necessarily reflect the opinions of the Queensland Department of Main Roads, Queensland Transport or Maritime Safety Queensland, or endorsed organisations utilising the same infrastructure. ***********************************************************************
Jeff Zaroyko wrote:
Gold means you're either using native dlls, have modified the program by patching it with nocd or you've modified Wine to make it work, ie there is a work around that makes the application work flawlessly. There's no reason to exclude modifying Wine, you are empowered to change it as you see fit since it's free software.
I'm siding with Bjoern on this one: If the patch is outside of the Wine tree, it is not Gold, it's Garbage. If you can 'fix it' with a native dll(s) then it is Gold IF and only IF all functions work. For instance, e-Sword, which is the only real reason I'm using Wine, has to have riched20.dll and riched32.dll in order to display text. Graphics do not display. It is NOT Gold, but Silver. Yes, you can display text without it, but you cannot input text for Notes or Topic Notes, which is what I consider a major function. The ability to display maps is not, as you can view them with a different, native, application. Now, the story changes if the patch is conforming and has been accepted by AJ and is pending the next development release. James McKenzie
Now, the story changes if the patch is conforming and has been accepted by AJ and is pending the next development release.
Then the next development release can get the gold, but previous ones still shouldn't. AppDB test ratings are tied to specific releases, and intended to tell normal users how different versions of Wine will work with their app. Patching Wine is not something normal users can or want to do, and allowing ratings based on patched versions of Wine is misleading, even if the patch does eventually make it in to a later release. -- Rosanne DiMesio <dimesio(a)earthlink.net>
On Tue, Jan 6, 2009 at 6:48 AM, Rosanne DiMesio <dimesio(a)earthlink.net> wrote:
Now, the story changes if the patch is conforming and has been accepted by AJ and is pending the next development release.
Then the next development release can get the gold, but previous ones still shouldn't. AppDB test ratings are tied to specific releases, and intended to tell normal users how different versions of Wine will work with their app. Patching Wine is not something normal users can or want to do, and allowing ratings based on patched versions of Wine is misleading, even if the patch does eventually make it in to a later release.
It sounds like the problem is that the version string in appdb isn't descriptive enough. It's perfectly reasonable to wonder if a given program can be made to work with a patched version of wine, and wonder how well it will work. It's also reasonable to wonder how it will work with a vanilla version. Both types of reports are useful to have in the appdb. Having a version "x.x.x (patched)" available to reporters would allow both types of reports to be clearly separated. Cheers, -n8 -- Nathan Gray http://www.n8gray.org/
On Tue, Jan 6, 2009 at 11:42 AM, Nathaniel Gray <n8gray(a)gmail.com> wrote:
On Tue, Jan 6, 2009 at 6:48 AM, Rosanne DiMesio <dimesio(a)earthlink.net> wrote:
Now, the story changes if the patch is conforming and has been accepted by AJ and is pending the next development release.
Then the next development release can get the gold, but previous ones still shouldn't. AppDB test ratings are tied to specific releases, and intended to tell normal users how different versions of Wine will work with their app. Patching Wine is not something normal users can or want to do, and allowing ratings based on patched versions of Wine is misleading, even if the patch does eventually make it in to a later release.
It sounds like the problem is that the version string in appdb isn't descriptive enough. It's perfectly reasonable to wonder if a given program can be made to work with a patched version of wine, and wonder how well it will work. It's also reasonable to wonder how it will work with a vanilla version. Both types of reports are useful to have in the appdb. Having a version "x.x.x (patched)" available to reporters would allow both types of reports to be clearly separated.
Cheers, -n8
-- Nathan Gray http://www.n8gray.org/
No. Because that allows for all sorts of dirty hacks, and is confusing to users. Ratings should specify default wine. They can list patches, etc., in the comments, with a note of how well it works. -- -Austin
On Tue, Jan 6, 2009 at 10:12 AM, Austin English <austinenglish(a)gmail.com> wrote:
On Tue, Jan 6, 2009 at 11:42 AM, Nathaniel Gray <n8gray(a)gmail.com> wrote:
It sounds like the problem is that the version string in appdb isn't descriptive enough. It's perfectly reasonable to wonder if a given program can be made to work with a patched version of wine, and wonder how well it will work. It's also reasonable to wonder how it will work with a vanilla version. Both types of reports are useful to have in the appdb. Having a version "x.x.x (patched)" available to reporters would allow both types of reports to be clearly separated.
No. Because that allows for all sorts of dirty hacks, and is confusing to users. Ratings should specify default wine. They can list patches, etc., in the comments, with a note of how well it works.
It seems to me that digging through comments to find out if a report refers to a version that was patched is more confusing than having it advertised right up front in the version string. And it makes sense -- a patched 1.1.11 is not the same *version* as 1.1.11. Cheers, -n8 -- Nathan Gray http://www.n8gray.org/
It sounds like the problem is that the version string in appdb isn't descriptive enough. It's perfectly reasonable to wonder if a given program can be made to work with a patched version of wine, and wonder how well it will work. It's also reasonable to wonder how it will work with a vanilla version. Both types of reports are useful to have in the appdb. Having a version "x.x.x (patched)" available to reporters would allow both types of reports to be clearly separated.
Patched with what? Lumping all the different possibilities into one "version" is also misleading. IMO, the appropriate rating for apps that can only be made to work by patching Wine is bronze: "Application works, but it has some issues, even for normal use..." Perhaps the wording can be changed to explicitly mention patching as one of the possible "issues" warranting a bronze, but I think the basic definition already fits. -- Rosanne DiMesio <dimesio(a)earthlink.net>
On Tue, Jan 6, 2009 at 10:15 AM, Rosanne DiMesio <dimesio(a)earthlink.net> wrote:
It sounds like the problem is that the version string in appdb isn't descriptive enough. It's perfectly reasonable to wonder if a given program can be made to work with a patched version of wine, and wonder how well it will work. It's also reasonable to wonder how it will work with a vanilla version. Both types of reports are useful to have in the appdb. Having a version "x.x.x (patched)" available to reporters would allow both types of reports to be clearly separated.
Patched with what? Lumping all the different possibilities into one "version" is also misleading.
Perfect information isn't necessary here. It's just useful to know what kind of performance you can get if you decide to venture into the world of patching and recompiling. Experts will know if it's worth their time, newbies will know to stay away. Cheers, -n8 -- Nathan Gray http://www.n8gray.org/
On Tue, Jan 6, 2009 at 8:48 AM, Rosanne DiMesio <dimesio(a)earthlink.net> wrote:
Now, the story changes if the patch is conforming and has been accepted by AJ and is pending the next development release.
Then the next development release can get the gold, but previous ones still shouldn't. AppDB test ratings are tied to specific releases, and intended to tell normal users how different versions of Wine will work with their app. Patching Wine is not something normal users can or want to do, and allowing ratings based on patched versions of Wine is misleading, even if the patch does eventually make it in to a later release.
+1 -- -Austin
participants (8)
-
Austin English -
Björn Krombholz -
James McKenzie -
Jeff Zaroyko -
Nathaniel Gray -
nicholas.g.lawrence＠mainroads.qld.gov.au -
Rosanne DiMesio -
Sparr