June 2009 - Wine-Devel - list.winehq.org

Re: Regular users right on bugzilla

by Ben Klein

2009/6/1 Ken Sharp <kennybobs(a)o2.co.uk>: > > > André Hentschel wrote: >> >> Jerome Leclanche schrieb: >>> >>> Since there's been some discussion about bugzilla recently... >>> >>> Is there any reason regular users aren't allowed to change keywords, >>> component and dependencies on bugs they don't own on Bugzilla? >>> Maybe marking as duplicate as well. I don't think it'd do any harm to >>> allow those rights to every registered user. >>> >>> Cheers >>> >>> Jerome >>> >> Maybe not to all users, there are always spammers, etc. >> But the idea is great, maybe we can give these rights just to the >> developers(everyone, who sent more than X patches). >> That makes more sense for me. >> Technical that means, every developer sends a mail somewhere with a >> keyword "bugzilla-mod" or something and the e-mail-adresses are compared >> with the registrations of wine-dev and wine-patches or with the commits in >> git. >> >> >> > > Most developers already have that access, and if they don't, they can just > ask for it. > > I think the reason all users can't do it, is because most users don't know > what they're doing. It's bad enough correcting all the mistakes they make > now, the last thing we need is every user doing whatever they like whenever > they like. Ask the poor triage guys who have to change nearly every bug. You beat me to it. Devs can ask for bugzilla "admin" rights by asking an appropriate admin or asking on -devel (I saw one just recently). I also agree with not giving extra control to all registered users (btw, I'm just a regular user on bugzilla; I've had to request bugs be marked DUP rather than set it myself :) ). It's bad enough they can change the status of their own bugs from INVALID to UNCONFIRMED (which I have seen all too often). "This is not a bug in Wine. Talk to <insert project here>. Closing invalid." "No, this IS a bug in Wine because Windows behaves differently! Reopening."

June 1, 2009

Re: Regular users right on bugzilla

by Ken Sharp

André Hentschel wrote: > Jerome Leclanche schrieb: >> Since there's been some discussion about bugzilla recently... >> >> Is there any reason regular users aren't allowed to change keywords, >> component and dependencies on bugs they don't own on Bugzilla? >> Maybe marking as duplicate as well. I don't think it'd do any harm to >> allow those rights to every registered user. >> >> Cheers >> >> Jerome >> > Maybe not to all users, there are always spammers, etc. > But the idea is great, maybe we can give these rights just to the > developers(everyone, who sent more than X patches). > That makes more sense for me. > Technical that means, every developer sends a mail somewhere with a > keyword "bugzilla-mod" or something and the e-mail-adresses are compared > with the registrations of wine-dev and wine-patches or with the commits > in git. > > > Most developers already have that access, and if they don't, they can just ask for it. I think the reason all users can't do it, is because most users don't know what they're doing. It's bad enough correcting all the mistakes they make now, the last thing we need is every user doing whatever they like whenever they like. Ask the poor triage guys who have to change nearly every bug.

June 1, 2009

Re: Regular users right on bugzilla

by André Hentschel

Jerome Leclanche schrieb: > Since there's been some discussion about bugzilla recently... > > Is there any reason regular users aren't allowed to change keywords, > component and dependencies on bugs they don't own on Bugzilla? > Maybe marking as duplicate as well. I don't think it'd do any harm to > allow those rights to every registered user. > > Cheers > > Jerome > Maybe not to all users, there are always spammers, etc. But the idea is great, maybe we can give these rights just to the developers(everyone, who sent more than X patches). That makes more sense for me. Technical that means, every developer sends a mail somewhere with a keyword "bugzilla-mod" or something and the e-mail-adresses are compared with the registrations of wine-dev and wine-patches or with the commits in git.

June 1, 2009

Patchwatcher not reporting failures?

by Austin English

Testing patchwatcher with a patch that simply removes two todo_wine's should've caused some failures. Indeed, looking at the log: cred.c:315: Test failed: expected CredentialBlobSize of 0 but got 20 cred.c:315: Test failed: expected NULL credentials but got 0x11d4ec cred.c:328: Test failed: expected CredentialBlobSize of 0 but got 20 cred.c:328: Test failed: expected NULL credentials but got 0x11d4d0 it did. But patchwatcher shows an 'OK'. Can you take a look? -- -Austin

June 1, 2009

Re: Its Alive! Patchwatcher is reborn.

by Francois Gouget

On Sun, 31 May 2009, Lei Zhang wrote: [...] > Speaking of pre-commit checking, the Wine wiki page for patchwatcher > page [1] currently list the following todo items: > > Check for C++ comments. Currently we don't have a good way of detecting these. winapi_check attempts to detect them but it gets fooled by stuff like foo("http://www.winehq.org") or /* http://www.winehq.org */. With a better C parser it may work. > Check for nameless unions (GCC 2.95). The way I check for these is to compile with gcc 2.95. However nowadays that requires using a Debian 4.0 chroot because gcc 2.95 is not supported anymore by the recent glibc and other libraries. So this is not a good way to detect these issues in a commit hook. I'm not sure detecting such issues would be easy in a script. > Check for missing "\n"s in traces/fixmes/etc. These have a bunch of exceptions. There are really a bunch of cases where there is intentionally no trailing '\n'. Typically because we're dumping a bunch of flags or the contents of an array in hexadecimal. So the pre commit hook would have to be overridable and would need a regularly updated list of exceptions so people committing stuff on these files don't get warned again and again. > Make sure sizeof is not used in traces. This looks like it could be reasonably easy to check for. Actually it's not just traces it's all forms of printf. -- Francois Gouget <fgouget(a)free.fr> http://fgouget.free.fr/ "Only wimps use tape backup: _real_ men just upload their important stuff on ftp, and let the rest of the world mirror it ;)" -- Linus Torvalds

June 1, 2009

Regular users right on bugzilla

by Jerome Leclanche

Since there's been some discussion about bugzilla recently... Is there any reason regular users aren't allowed to change keywords, component and dependencies on bugs they don't own on Bugzilla? Maybe marking as duplicate as well. I don't think it'd do any harm to allow those rights to every registered user. Cheers Jerome -- J. Leclanche / Adys

June 1, 2009

Re: Installing ActivePython screws up PATH in Wine?

by Aurimas Fišeras

On 2009.06.01 08:40, Dan Kegel wrote: > In Wine, I can happily install mingw and add it to the > global PATH by editing the registry key PATH in > [System\\CurrentControlSet\\Control\\Session Manager\\Environment] > > However, once I install ActivePython, something > gets screwed up, and from then on, 'PATH' in CMD > says > PATH="C:\Python26\";;C:\Python26\; > even if I remove that bit from that registry key > and add other directories. > Anyone know what the heck is going on? Maybe it sets PATH as local variable in HKEY_CURRENT_USER/Environment? http://www.winehq.org/docs/wineusr-guide/environment-variables

June 1, 2009

Re: Its Alive! Patchwatcher is reborn.

by Lei Zhang

On Sun, May 31, 2009 at 12:05 AM, Zachary Goldberg <zgold(a)bluesata.com> wrote: > All, > > As you may have noticed in the last WWN I have been working on getting > patchwatcher back up on a rather nice server whose cpu time is > graciously donated by STWing (stwing.org) I've got it running at the > moment at > > http://winepatch.stwing.upenn.edu/results2/ > > There are still many, many kinks to be worked out before I would use > this as reliable and/or before I even think about enabling sending to > patch authors. But if you're curious thats where the results are > going and will likely be in the future. > > -Zach > > p.s. Kudos to Dan Kegel for a great walkthrough in the PW readme.txt, > and of course for putting patchwatcher together in the first place! > > > Speaking of pre-commit checking, the Wine wiki page for patchwatcher page [1] currently list the following todo items: Check for C++ comments. Check for nameless unions (GCC 2.95). Check for missing "\n"s in traces/fixmes/etc. Make sure sizeof is not used in traces. These sound like things we can catch with a git pre-commit hook, so developers who use git can fix these even before it hits patchwatcher. [1] http://wiki.winehq.org/PatchWatcher

June 1, 2009

Installing ActivePython screws up PATH in Wine?

by Dan Kegel

In Wine, I can happily install mingw and add it to the global PATH by editing the registry key PATH in [System\\CurrentControlSet\\Control\\Session Manager\\Environment] However, once I install ActivePython, something gets screwed up, and from then on, 'PATH' in CMD says PATH="C:\Python26\";;C:\Python26\; even if I remove that bit from that registry key and add other directories. Anyone know what the heck is going on?

June 1, 2009

Remove FIXME keyword from bugzilla

by Austin English

It serves no purpose, after all, 'FIXME's are not bugs'. -- -Austin

June 1, 2009