May 2010 - Wine-Devel - list.winehq.org

Re: Assigning bugs, and how new developers can find bugs to work on

by Octavian Voicu

On Tue, Jun 1, 2010 at 1:56 AM, Alexandru Băluț <alexandru.balut(a)gmail.com> wrote: > The git log shows what you did. This would show how much you have to do > (rough approximation, of course). btw, how exactly do you keep track of what > you have to do? Having bugs assigned looks like a way of managing your > todo-s. One way is to add yourself to the bug's CC list (you'll probably want that anyway, since if you're working on that bug you're interested if anybody posts updates), then save a search that lists bugs which have you in the CC field: hit Search on the toolbar, set criteria, Search, then fill in the field next to "Remember search as" and click the button. A new item will appear on the left in "Saved Searches". Hope it helps, Octavian

May 31, 2010

Re: Assigning bugs, and how new developers can find bugs to work on

by Nikolay Sivov

On 6/1/2010 02:56, Alexandru Băluț wrote: > >> In the meanwhile, please assign the bugs, and spare others of >> spending time on them. As a bonus, it would be nice to see >> accurate numbers about how many bugs are on nobody's plate vs how >> many are on somebody's plate. Data! :) At the moment, there are >> 15 assigned bugs (ignored 11 assigned to wine-bugs@), vs 5580 >> unconfirmed/new/reopened bugs, for the Wine component. > Why do you need this? git log/blame shows all stats on every > person. It's a waste of time IMO to go with these ASSIGNED/VERIFIED. > > > The git log shows what you did. This would show how much you have to > do (rough approximation, of course). btw, how exactly do you keep > track of what you have to do? Having bugs assigned looks like a way of > managing your todo-s. You're overestimating a value of this tracking process. Currently a really long bug list is available, all you need is to choose a bug that fits your interests, put a note and start to fix it. > > Thanks, > Alex

May 31, 2010

Re: Assigning bugs, and how new developers can find bugs to work on

by Alexandru Băluț

Hi Nikolay, On Mon, May 31, 2010 at 18:27, Nikolay Sivov <nsivov(a)codeweavers.com> wrote: > On 5/31/2010 20:01, Alexandru Băluț wrote: > > Hi Paul, > > I was wondering what is the best way to pick bugs, for a developer > aspiring to fix nonspecific Wine bugs. Normally, our hero would look at the > unassigned bugs to find one to work on. For some bugs, a fix appeared with > no notification that somebody is working on them, for example: > > http://bugs.winehq.org/show_bug.cgi?id=22263 > - julliard@ unexpectedly announced that the bug should be fixed by a > specific patch. > > Unexpectedly is very questionable word here cause it was a regression. > The bissection did pinpoint the commit that caused the bug, and its author, but you still have to read the comments to figure that out, when it would have been easier to immediately see that the bug is assigned, and just leave it alone. (I'm talking from the point of view of a newbie Wine developer who is looking for a bug to fix.) http://bugs.winehq.org/show_bug.cgi?id=22845 > - arethusa26@ unexpectedly mentioned that the bug should be fixed by > somebody else's specific patch. > > Anyone could retest a make a note about that. > Yes, anyone who retests can make a note about that. > Notice both are marked as FIXED, but none of them have been assigned! > > ASSIGNED state isn't used too often and there's nothing wrong with that. > I'm not saying it's wrong if people don't use it, I'm saying it's useful when people use it. > Let's say our hero picks a bug; should he ask on the bug page whether > anybody is already working on it, or ask on wine-devel, or #wine-hackers? > > Pick whatever you want, read through comments to figure out does somebody > work on it or not and write a fix. > > Or is it fine to just announce, in the bug page, that you'll work on that > bug? See, for example, this comment: > http://bugs.winehq.org/show_bug.cgi?id=22264#c9 > > That's fine. > Ok. I personally don't like to pollute the bug log with "I'll work on this" messages. But I guess for newbie developers, who don't have bugzilla rights to change the Assigned field, it's the only solution. Any suggestions about improving this process are welcome. I think people > will show up, wanting to help, and it would be nice to help them find > something to work on fast. ;) > > There's nothing to improve. You could search by component or any way you > want. > I realized while writing this email.. the process of searching a bug nobody is working on can be improved. > In the meanwhile, please assign the bugs, and spare others of spending time > on them. As a bonus, it would be nice to see accurate numbers about how many > bugs are on nobody's plate vs how many are on somebody's plate. Data! :) At > the moment, there are 15 assigned bugs (ignored 11 assigned to wine-bugs@), > vs 5580 unconfirmed/new/reopened bugs, for the Wine component. > > Why do you need this? git log/blame shows all stats on every person. It's a > waste of time IMO to go with these ASSIGNED/VERIFIED. > The git log shows what you did. This would show how much you have to do (rough approximation, of course). btw, how exactly do you keep track of what you have to do? Having bugs assigned looks like a way of managing your todo-s. Thanks, Alex

May 31, 2010

Re: RFC: msvcrt: support dates after year 2038 in mktime64

by Robert Wilhelm

On Sun, 2010-05-30 at 13:02 -0700, James McKenzie wrote: > I think I know your confusion here. Does UNIX and Windows implement > these functions for the same date range or should we be rejecting a set > of dates as being invalid? > Windows has 1970-2038 for mktime32 and 1970-3000 for _mktime64. The lower border is ok: Some Unix versions support dates before 1970. We already check for negative numbers in wine and return -1 in this case. This is fine. For upper border there is check for dates >3000 missing. Some Unix systems support dates up to 9999. There may be also corner cases at the borders (depending on timezone etc.) which are different. But I doubt anybody will care.

May 31, 2010

Re: New wisotool 20100530: new verbs galore

by wylda＠volny.cz

> > Does bioshock install ok manually from .mds for you > with cdemu? Yes, without any problem. I always double check each image, i.e. first i test the image under winxp and when the image is ok, i take it under wine. Bioshock passed both these tests successfully, i.e. installs and runs without originals (just the mounted image, no cracks of course). BTW: if you have a non-working image under winxp (have to work - necessary condition) check your available drive speeds. For SecuROM is recommended: CD medium: Use lowest available speed for DPM DVD medium: use highest available speed for DPM But my personal experience is different. For example DVD F.E.A.R. Extraction Point i succeeded after 4th try in case of highest DPM speed. Then i tried lowest DPM speed even for DVD and succeeded in every test. So since this i use in alcohol120: * 8x (CD-ROM read) * 4x (DVD-ROM read) Those are lowest speeds for my Samsung SH-S223F. It _should_ work setting 1x and the drive should choose lowest possible, but it happened sometimes to me, that max DPM speed was chosen instead. So i set the speeds explicitly. Unfortunately i also have DVD drives which will _never_ be capable of making working image (to name one: Sony/Optiarc AD-5200S). So if you can't succeed with image creation for longer time, send your drive to trash ;) What makes the image successful is .mds file (mdf should be always the same for different tries). So in case you have non-working image, you can ask someone just for .mds (but .mdf must have same size - i.e. same language DVD or gamebox). > > That's a familiar error. We can and should fix that > before you try again. OK :) > > Also how do you handle multi CD games whit this tool? > > I have 8 virtual drives... > The script can handle unmounting the 1st disc and mounting > the 2nd, but maybe we should mount multiple as you suggest. > (That's a bit more complicated to script because we want the > script to run on windows, too). > - Dan I met games, which are hard to persuade even with "wine eject" and also then you usually need to confirm "OK CD2 is in drive". So to me it looks easier to mount all CDs/DVDs before installation begins. (I'm not a programmer so maybe i'm wrong here;)

May 31, 2010

Re: Last Minute Submission: Winecfg>About

by Austin English

On Mon, May 31, 2010 at 9:28 AM, Alexandru Băluț <alexandru.balut(a)gmail.com> wrote: > Please file a bug. > Thanks, > Alex > > On Mon, May 31, 2010 at 11:17, <Joerg-Cyril.Hoehle(a)t-systems.com> wrote: >> >> Hi, >> >> Joel Holdsworth wrote: >> >http://www.airwebreathe.org.uk/space/new-about-screen.png >> winecfg's window seems very tall. I'd like it if somebody >> took care that it displays correctly in a 800x600 screen or virtual >> desktop. >> Currently, the ok & cancel buttons at the bottom are cut off :-( Alexandru, please bottom post on wine mailing lists. There already is a bug, see http://bugs.winehq.org/show_bug.cgi?id=15818. -- -Austin

May 31, 2010

Re: Assigning bugs, and how new developers can find bugs to work on

by Nikolay Sivov

On 5/31/2010 20:01, Alexandru Băluț wrote: > Hi Paul, > > I was wondering what is the best way to pick bugs, for a developer > aspiring to fix nonspecific Wine bugs. Normally, our hero would look > at the unassigned bugs to find one to work on. For some bugs, a fix > appeared with no notification that somebody is working on them, for > example: > > http://bugs.winehq.org/show_bug.cgi?id=22263 > - julliard@ unexpectedly announced that the bug should be fixed by a > specific patch. Unexpectedly is very questionable word here cause it was a regression. > > http://bugs.winehq.org/show_bug.cgi?id=22845 > - arethusa26@ unexpectedly mentioned that the bug should be fixed by > somebody else's specific patch. Anyone could retest a make a note about that. > > Notice both are marked as FIXED, but none of them have been assigned! ASSIGNED state isn't used too often and there's nothing wrong with that. > > Let's say our hero picks a bug; should he ask on the bug page whether > anybody is already working on it, or ask on wine-devel, or #wine-hackers? > Pick whatever you want, read through comments to figure out does somebody work on it or not and write a fix. > Or is it fine to just announce, in the bug page, that you'll work on > that bug? See, for example, this comment: > http://bugs.winehq.org/show_bug.cgi?id=22264#c9 > That's fine. > > Any suggestions about improving this process are welcome. I think > people will show up, wanting to help, and it would be nice to help > them find something to work on fast. ;) There's nothing to improve. You could search by component or any way you want. > > In the meanwhile, please assign the bugs, and spare others of spending > time on them. As a bonus, it would be nice to see accurate numbers > about how many bugs are on nobody's plate vs how many are on > somebody's plate. Data! :) At the moment, there are 15 assigned bugs > (ignored 11 assigned to wine-bugs@), vs 5580 unconfirmed/new/reopened > bugs, for the Wine component. Why do you need this? git log/blame shows all stats on every person. It's a waste of time IMO to go with these ASSIGNED/VERIFIED. P.S. also I think our bugzilla setup allows to much currently. Along with practically unused states I mentioned already there's a difficulty field that needs to be removed; it's nice to tweak CC list a bit too to block reporter from being CCed. > > Thanks, > Alex

May 31, 2010

Assigning bugs, and how new developers can find bugs to work on

by Alexandru Băluț

Hi Paul, I was wondering what is the best way to pick bugs, for a developer aspiring to fix nonspecific Wine bugs. Normally, our hero would look at the unassigned bugs to find one to work on. For some bugs, a fix appeared with no notification that somebody is working on them, for example: http://bugs.winehq.org/show_bug.cgi?id=22263 - julliard@ unexpectedly announced that the bug should be fixed by a specific patch. http://bugs.winehq.org/show_bug.cgi?id=22845 - arethusa26@ unexpectedly mentioned that the bug should be fixed by somebody else's specific patch. Notice both are marked as FIXED, but none of them have been assigned! Let's say our hero picks a bug; should he ask on the bug page whether anybody is already working on it, or ask on wine-devel, or #wine-hackers? Or is it fine to just announce, in the bug page, that you'll work on that bug? See, for example, this comment: http://bugs.winehq.org/show_bug.cgi?id=22264#c9 Any suggestions about improving this process are welcome. I think people will show up, wanting to help, and it would be nice to help them find something to work on fast. ;) In the meanwhile, please assign the bugs, and spare others of spending time on them. As a bonus, it would be nice to see accurate numbers about how many bugs are on nobody's plate vs how many are on somebody's plate. Data! :) At the moment, there are 15 assigned bugs (ignored 11 assigned to wine-bugs@), vs 5580 unconfirmed/new/reopened bugs, for the Wine component. Thanks, Alex On Thu, May 27, 2010 at 20:00, Paul Vriens <paul.vriens.wine(a)gmail.com>wrote: > On 05/27/2010 07:48 PM, belnac wrote: > >> What about dev work, is any available? >> > > Did you ever have a look at our bugzilla ;) > > There is plenty of stuff to do besides fixing bugs of course. What you want > to do depends on many things like skills, interest, the amount of time you > have etc.. > > Remember that we are in code freeze now so looking at bugs and trying to > figure out what could be wrong or even providing information that could lead > to fixing the bugs is welcome. > > -- > Cheers, > > Paul. > > >

May 31, 2010

Re: comdlg32: Fix Italian translation

by Matteo Bruni

2010/5/31 Luca Bennati <lucak3(a)gmail.com>: > > > 2010/5/31 Matteo Bruni <matteo.mystral(a)gmail.com> >> >> > - LTEXT "File di &Tipo:", 1089, 6, 104, 90, 9 >> > + LTEXT "&Tippo del file:", 1089, 6, 104, 90, 9 >> >> One 'p' is enough :). More importantly, this and the shell32 patch >> have mangled accented letters (maybe the conversion to UTF8 didn't >> work out well?). No problems in the taskmgr patch though. >> >> By the way, thank you for your translations! > > The mangled acennted letters were the one i was removing or the ones i was > adding? > They show up good in my WINE copy... > BTW i need help for crypt and ole related translation because i dont > understand well the meanings > Yes, I referred to the ones you added. I didn't try to apply the patches, I just saw them from Firefox, and with the browser set to utf-8 encoding some letters were wrong, like when you look to an utf-8 file with an ISO-8859-1 setup. FWIW, changing the browser encoding to ISO-8859-1 breaks these characters more. Maybe it's simply my browser messing up when finding different characters encoding in the same file... Looking again at your comdlg32 patch I catched this typo: IDS_INVALID_FILENAME "Il nome di un file non può contenere i seguenrti caratteri:\n / : < > |" Regarding crypt and ole translations, feel free to write me privately for help (assuming I understand the meanings, which is not so sure :)).

May 31, 2010

Re: New wisotool 20100530: new verbs galore

by wylda＠volny.cz

> wisotool is a handy winetricks-like script for automatically > installing games from .iso or .mds files copied from > your own dvds I wanted to try it, but i wasn't successful probably because of my faults. So i prepared Bioshock .mds image, which i held on different HDD because of install speed, drive space etc. It works, if i manually mount it into virtual DVD drive (by cdemu) and then manually install by wine setup.exe etc... Can i instruct wisotool to read the image from my folder? Now i get: Executing wine vcredist_x86.exe /q 2690 cdemud is already running /mnt/isos/game_isos/bioshock/bioshock_my-private-game-key-here.mds is mounted in cdemu, forcing an eject eject: unable to find or open device for: `/media/BIOSHOCK' ERROR: Failed to load image: Image file cannot be opened or read. Mounting as drive i: ln: creating symbolic link `/home/user/.wine/dosdevices/i:' -> `': No such file or directory Setting DPM emulation flag of device 0 to 1. Setting transfer rate emulation flag of device 0 to 0. cat: /home/user/.wisotool/cache/bioshock/391042ea19b4e_some_hash.txt: No such file or directory BTW: looks like you are setting things like DPM (Data Position Measurement) explicitly? I have around 150 games transferred to .mds images and i've never need to do that. Just as a user: $ cdemu load 0 bioshock_my-private-game-key-here.mds Also how do you handle multi CD games whit this tool? I have 8 virtual drives and for example NFS Underground has two CD, so i mount both of them and let the installations begin. The advantage is, that the games do not asks "please put 2nd CD" etc.

May 31, 2010