Re: Which virtualization software should I choose
On Thursday 11 June 2009 14:58:14 David Gerard wrote: > VMWare is reputedly better - it's the oldest common VM software and > its emulation is very seasoned, well-tested and robust. If you discount that last I checked, VMware still couldn't do IPv6 in their virtual networks, and every single kernel upgrade your distro does is a pain. There's always something, I guess. Kai -- Kai Blin WorldForge developer http://www.worldforge.org/ Wine developerhttp://wiki.winehq.org/KaiBlin Samba team member http://www.samba.org/samba/team/ -- Will code for cotton. signature.asc Description: This is a digitally signed message part.
Re: OS X build broken
On Thu, Jun 11, 2009 at 11:39 PM, Austin English wrote: > AJ, looks like your 'configure: Detect the appropriate form for the > __ASM_GLOBAL_FUNC macro.' patch broke the compile on OS X. Could you > have a look? Bah you beat me reporting this by 13mins! :P -- Steven Edwards "There is one thing stronger than all the armies in the world, and that is an idea whose time has come." - Victor Hugo
Re: Alexandre Julliard : configure: Detect the appropriate form for the __ASM_GLOBAL_FUNC macro.
Hi, On Thu, Jun 11, 2009 at 11:54 AM, Alexandre Julliard wrote: > Module: wine > Branch: master > Commit: 857f1e0924f23865038a608ee6834ef524371f20 > URL: > http://source.winehq.org/git/wine.git/?a=commit;h=857f1e0924f23865038a608ee6834ef524371f20 > > Author: Alexandre Julliard > Date: Thu Jun 11 16:32:42 2009 +0200 > > configure: Detect the appropriate form for the __ASM_GLOBAL_FUNC macro. I believe this patch is causing a problem for me on OS X Leopard gcc version 4.0.1 (Apple Inc. build 5465). When I try to compile riched20 with a clean tree at HEAD I am getting this ld: absolute addressing (perhaps -mdynamic-no-pic) used in _itextHostStdcallVtbl from txthost.o not allowed in slidable image collect2: ld returned 1 exit status winegcc: gcc failed make: *** [riched20.dll.so] Error 2 Is the __ASM_GLOBAL_FUNC stdcall wrapper in txthost.c just borked? Thanks -- Steven Edwards "There is one thing stronger than all the armies in the world, and that is an idea whose time has come." - Victor Hugo
OS X build broken
AJ, looks like your 'configure: Detect the appropriate form for the __ASM_GLOBAL_FUNC macro.' patch broke the compile on OS X. Could you have a look? gcc -m32 -c -I. -I. -I../../include -I../../include -D__WINESRC__ -D_REENTRANT -fPIC -Wall -pipe -fno-strict-aliasing -Wdeclaration-after-statement -Wwrite-strings -Wpointer-arith -I/Users/austin/wine-deps/usr/include -I/Users/austin/wine-deps/usr/include/libxml2 -I/Users/austin/wine-deps/usr/include/libxslt -I/Users/austin/wine-deps/usr/include/libpng12 -I/Users/austin/wine-deps/usr/include/gphoto2 -I/Users/austin/wine-deps/usr/include/sane -I/Users/austin/wine-deps/usr/include -o writer.o writer.c ../../tools/wrc/wrc --nostdinc -I. -I. -I../../include -I../../include -D__WINESRC__ -foversion.res version.rc ../../tools/winegcc/winegcc -m32 -B../../tools/winebuild --sysroot=../.. -shared ./riched20.speccaret.o clipboard.o context.o editor.o list.o paint.o para.o reader.o richole.o row.o run.o string.o style.o table.o txthost.o txtsrv.o undo.o wrap.o writer.oversion.res -o riched20.dll.so -luuid -lole32 -loleaut32 -limm32 -luser32 -lgdi32 -lkernel32 ../../libs/port/libwine_port.a -L/Users/austin/wine-deps/usr/lib ld: absolute addressing (perhaps -mdynamic-no-pic) used in _itextHostStdcallVtbl from txthost.o not allowed in slidable image collect2: ld returned 1 exit status winegcc: gcc failed make[2]: *** [riched20.dll.so] Error 2 make[1]: *** [riched20] Error 2 make: *** [dlls] Error 2 -- -Austin
Re: RDo we still need the regression test for 1.1.23 ?
2009/6/12 chris ahrendt : > > Ben Klein wrote: >> 2009/6/11 chris ahrendt : >> >>> to see what patch caused the failure on non NVIDIA cards? >>> >> >> Where's the rest of your message? Oh, I see, it's in the subject. >> >> Wasn't there already a thread about this? >> >> Back on topic, if it is definitely a regression then a regression test >> on bugzilla will be useful. Or at least, some information on how to >> reproduce it. >> >> >> > The test is easy... use a ATI card... and run 1.1.23 with or without > the Registry value to set it back to the old way of rendering. Not so easy if you don't have an ATI card, or if you don't know what apps cause the error. Wine Notepad seems to be working fine for me, no crashes.
Re: Which virtualization software should I choose
2009/6/11 Paul Vriens : > Just installed VirtualBox and it does have snaphots but not so extensive as > VMware. In VMware I have some W2K snapshots: > > - out-of-the-box > - SP1 > - SP2 > - SP3 > - SP4 + Windows Update > > I can freely choose which I want to go to. Not so with VirtualBox I'm afraid > as they are all on top of each other. > > (I guess for now I keep my other, F10, box around). You can get this behaviour, but not using snapshots (which as you noticed are cumulative, and rolling back one snapshot deletes its changes). What you do is install your "out-of-the-box" version in a new drive image, then mark that as "immutable". AFAIK, you still have to do that from the command-line, and it's not really intuitive. Something like: 1) De-register your "out-of-the-box" drive image from all VMs you have configured 2) $ vboxmanage closemedium disk 3) $ vboxmanage openmedium disk --type immutable 4) Re-associate your VMs to the "out-of-thebox" image. It will inform you now that it is immutable. You can then create as many VMs as you need with the "immutable" image as the base, and it will automagically create differencing images for you, which you can install the service packs in. The disadvantage of this is you have multiple VM configurations, all of which have independent settings, but this is what I do for my WinXP browser testing (so I have multiple different versions of IE/Opera/other browsers that don't like other versions existing on the same installation :) ).
Re: RDo we still need the regression test for 1.1.23 ?
2009/6/11 chris ahrendt : > > to see what patch caused the failure on non NVIDIA cards? Where's the rest of your message? Oh, I see, it's in the subject. Wasn't there already a thread about this? Back on topic, if it is definitely a regression then a regression test on bugzilla will be useful. Or at least, some information on how to reproduce it.
Re: Disable forum edits
On Mon, Jun 8, 2009 at 11:12 PM, Dmitry Timoshkov wrote: > "Igor Tarasov" wrote: > >> 2009/6/9 Austin English : >>> >>> Users are editing their posts when asked for terminal output/wine >>> version/etc. Those of us on the mailing list side don't get this >>> information. >> >> Maybe just add notice on edit page stating that edits are for minor >> changes only, and if one has some additional information he should add >> it as a separate post? > > There is a request to not paste logs and backtraces when adding a comment > to a bug, that doesn't prevent that very well. So disabling the feature > is always a preference. Jeremy, would you mind disabling edits please? -- -Austin
Re: advpack: Implement ExtractFilesW [try 2]
2009/6/11 Vladimir Pankratov : > Hello all. > > Implemented ExtractFilesW > > Changed files: > advpack/files.c > +if (GetFileAttributesA(ExpandDir) == INVALID_FILE_ATTRIBUTES) +return HRESULT_FROM_WIN32(ERROR_PATH_NOT_FOUND); Why are you testing this in ExtractFilesA? Almost all of the logic should be passed on to ExtractFilesW so we don't duplicate code. +if (FileList != NULL) Please don't change the style of the file. if (FileList) is fine. +WideCharToMultiByte( CP_ACP, 0, FileList, -1, szFileList, sizeof(szFileList), NULL, NULL ); +szConvertedList = convert_file_list(szFileList, &dwFileCount); I'm not sure this is the way to go. You're converting A -> W -> A. This happens because cabinet.Extract is (as far as we know) ansi-only. In this case, I think it's ok to make ExtractFilesW call into ExtractFilesA, with an appropriate comment as to why this is happening. -- James Hawkins
Fwd: RDo we still need the regression test for 1.1.23 ?
Forwarding to mailing list... -- Forwarded message -- From: Danila Sentiabov Date: Thu, Jun 11, 2009 at 23:02 Subject: Re: RDo we still need the regression test for 1.1.23 ? To: chris ahrendt I didn't follow the topic, but as far as I know, if there is a regression it's surely begging for regression test. On which application the failure can be reproduced? On Wed, Jun 10, 2009 at 22:47, chris ahrendt wrote: > > to see what patch caused the failure on non NVIDIA cards? > > > Chris > -- Best regards, Danila Sentiabov aka dsent -- Best regards, Danila Sentiabov aka dsent
Re: Which virtualization software should I choose
On Thu, Jun 11, 2009 at 11:22 AM, David Gerard wrote: > 2009/6/11 Austin English : >> On Thu, Jun 11, 2009 at 7:58 AM, David Gerard wrote: > >>> VirtualBox does okay for Windows and Linux, barely for FreeBSD with >>> lots of caveats and not really for anything else. Notably, OpenBSD >>> doesn't work and the VirtualBox developers admit it but consider the >>> bug beneath their attention. It's really not a very well-written >>> virtualisation app. > >> Virtualbox works great for OpenSolaris as well. > > > *cough* Yes, I expect it would :-) > > (explanation: Sun owns VirtualBox, you can be very sure their own OS > runs like a charm!) I'm well aware...though, interestingly, sound is broken out of the box and requires compiling your own oss driver :-/ -- -Austin
Re: Which virtualization software should I choose
2009/6/11 Austin English : > On Thu, Jun 11, 2009 at 7:58 AM, David Gerard wrote: >> VirtualBox does okay for Windows and Linux, barely for FreeBSD with >> lots of caveats and not really for anything else. Notably, OpenBSD >> doesn't work and the VirtualBox developers admit it but consider the >> bug beneath their attention. It's really not a very well-written >> virtualisation app. > Virtualbox works great for OpenSolaris as well. *cough* Yes, I expect it would :-) (explanation: Sun owns VirtualBox, you can be very sure their own OS runs like a charm!) - d.
Re: Which virtualization software should I choose
2009/6/11 Paul Vriens : > The problem with both VirtualBox and QEMU/KVM seems to be supporting older > Windows versions as guests (< NT4). Yeah. Try QEMU without KVM on a very fast host machine, your ancient Windows should run just as well as it would on a 486 ;-) - d.
Re: Which virtualization software should I choose
On 06/11/2009 02:58 PM, David Gerard wrote: 2009/6/11 Michael Stefaniuc: Paul Vriens wrote: I need to run several Windows versions (95 up to Vista for now) for our winetest and I really like the snapshot possibilities of VMware. Suggestions, recommendations? I'm using KVM/Qemu with libvirt aka virt-manager. I have problems with WinNT and NetBSD; those hang/crash during the install. Win2k3, FreeBSD and OpenSolaris work just fine and I didn't try anything else yet. But I'm on F9 and once I have a little spare time I'll move to F11. I'll test it there (never qemu) if that fixes WinNT. Since you're asking on the Wine list ... I'm using VMware only for running winetest on Windows, so there is some relevance. VirtualBox does okay for Windows and Linux, barely for FreeBSD with lots of caveats and not really for anything else. Notably, OpenBSD doesn't work and the VirtualBox developers admit it but consider the bug beneath their attention. It's really not a very well-written virtualisation app. VMWare is reputedly better - it's the oldest common VM software and its emulation is very seasoned, well-tested and robust. I've been a happy user of VMware Workstation from the very start but it always annoyed me that they don't support Fedora. QEMU *without* KVM works well, if slowly - it's almost a complete "red pill" for the guest OS. With KVM, OpenBSD is known not to work entirely properly. (OpenBSD is a bit of a torture test for virtual machines. It's very wary and cautious about what it runs on, and will happily segfault at a perceived hardware problem rather than risk letting a program access memory it shouldn't. Theo de Raadt says about a third of all problem traces come from VMs.) For testing Wine in other OSes, bugs found in a VM should always, always be confirmed on a physical machine, with the OS running on the bare metal - there are too many "glitches in the Matrix" in almost any VM software to be sure a bug is real. That's basically true for most applications. Most vendors have these caveats in there support contracts. Maybe I just revert to F10 for now and wait till VMware can properly install/run on F11 (It's mostly issues with compilation and dealing with newer kernel versions, 2.6.29/30). F11 doesn't do OSS (by default) anymore which means no out-of-the-box sound in VMware Workstation 6.5.2 The problem with both VirtualBox and QEMU/KVM seems to be supporting older Windows versions as guests (< NT4). -- Cheers, Paul.
Re: [Bugzilla]: Refuse to accept comments with big number of logs / back traces
On Thu, Jun 11, 2009 at 7:18 AM, Ken Sharp wrote: > Attachments can be deleted, can comments be deleted? No. They're archived all over the interwebs anyway, there is no point. I support this patch. -- -Austin
Re: Which virtualization software should I choose
On Thu, Jun 11, 2009 at 7:58 AM, David Gerard wrote: > VirtualBox does okay for Windows and Linux, barely for FreeBSD with > lots of caveats and not really for anything else. Notably, OpenBSD > doesn't work and the VirtualBox developers admit it but consider the > bug beneath their attention. It's really not a very well-written > virtualisation app. Virtualbox works great for OpenSolaris as well. -- -Austin
Re: [Bugzilla]: Refuse to accept comments with big number of logs / back traces
Rein Klazes wrote: > On Thu, 11 Jun 2009 16:30:57 +0400, you wrote: > >>> http://bugs.winehq.org/show_bug.cgi?id=5 >> There's nothing about why it's so important. These 25, 35 or 20 lines to >> allow means nothing. >> Actually I don't see any problem here, and, yes, I'm subscribed on >> wine-bugs list and this full log pasting >> occurs for 1 in 30 (maybe less a bit) new bugs reported. > > Agreeing 100% with that. > Reading each individual post via e-mail is not a problem. Going to bugzilla and trying to read through the entire bug is. Yes the stop sign does help with new posts, but does absolutely nothing for people who post 100+ lines of their terminal output into old bugs and make them totally impossible to follow. Again this patch does not completely stop all logs from being posted, only standard output of fixmes, errs, warns, etc (without thread IDs). This is on purpose as Anastasius almost always adds +tid to the mix. And even allows some number of standard log lines, quite enough for quotation. The fact that we get 1 in 30 doesn't mean we'll always see that much. It's much more common among first time reporters then anyone else. And number of new users will always go up. Vitaliy.
Re: Which virtualization software should I choose
2009/6/11 Michael Stefaniuc : > Paul Vriens wrote: >> I need to run several Windows versions (95 up to Vista for now) for our >> winetest and I really like the snapshot possibilities of VMware. >> Suggestions, recommendations? > I'm using KVM/Qemu with libvirt aka virt-manager. I have problems with WinNT > and NetBSD; those hang/crash during the install. Win2k3, FreeBSD and > OpenSolaris work just fine and I didn't try anything else yet. But I'm on F9 > and once I have a little spare time I'll move to F11. I'll test it there > (never qemu) if that fixes WinNT. Since you're asking on the Wine list ... VirtualBox does okay for Windows and Linux, barely for FreeBSD with lots of caveats and not really for anything else. Notably, OpenBSD doesn't work and the VirtualBox developers admit it but consider the bug beneath their attention. It's really not a very well-written virtualisation app. VMWare is reputedly better - it's the oldest common VM software and its emulation is very seasoned, well-tested and robust. QEMU *without* KVM works well, if slowly - it's almost a complete "red pill" for the guest OS. With KVM, OpenBSD is known not to work entirely properly. (OpenBSD is a bit of a torture test for virtual machines. It's very wary and cautious about what it runs on, and will happily segfault at a perceived hardware problem rather than risk letting a program access memory it shouldn't. Theo de Raadt says about a third of all problem traces come from VMs.) For testing Wine in other OSes, bugs found in a VM should always, always be confirmed on a physical machine, with the OS running on the bare metal - there are too many "glitches in the Matrix" in almost any VM software to be sure a bug is real. - d.
Re: Which virtualization software should I choose
2009/6/11 Paul Vriens : > On 06/11/2009 11:30 AM, Nicolas Le Cam wrote: >> >> 2009/6/11 Paul Vriens: >>> >>> Hi, >>> >>> Just upgraded to Fedora 11 and (yet again) having issues with VMware >>> Workstation. >>> >>> VMware still doesn't support Fedora (as a host) and I'm tired of >>> fixing/patching things, again and again. >>> >>> I need to run several Windows versions (95 up to Vista for now) for our >>> winetest and I really like the snapshot possibilities of VMware. >>> >>> Suggestions, recommendations? >>> >>> -- >>> Cheers, >>> >>> Paul. >>> >>> >>> >> >> I'm using VirtualBox. Apart from some problems around d3d/opengl and >> interruptions, it works quiet well. It also manages snapshots and have >> a CLI. >> See w2k-sp4-fr results on winetest. It has passed the test suite once >> (I didn't run latest builds but I will). >> >> Unfortunately, I don't know VMWare Workstation at all, so I can't tell >> you what are the differences. >> > Just installed VirtualBox and it does have snaphots but not so extensive as > VMware. In VMware I have some W2K snapshots: > > - out-of-the-box > - SP1 > - SP2 > - SP3 > - SP4 + Windows Update > > I can freely choose which I want to go to. Not so with VirtualBox I'm afraid > as they are all on top of each other. > > (I guess for now I keep my other, F10, box around). > > -- > Cheers, > > Paul. > I thought it was possible but never tried as I'm just using an up-to-date clean install, it suffices for my needs. If I have some time I'll try more things with those snapshots. -- Nicolas Le Cam
Re: [Bugzilla]: Refuse to accept comments with big number of logs / back traces
On Thu, 11 Jun 2009 16:30:57 +0400, you wrote: >> http://bugs.winehq.org/show_bug.cgi?id=5 >There's nothing about why it's so important. These 25, 35 or 20 lines to >allow means nothing. >Actually I don't see any problem here, and, yes, I'm subscribed on >wine-bugs list and this full log pasting >occurs for 1 in 30 (maybe less a bit) new bugs reported. >> Agreeing 100% with that. Rein.
Re: Which virtualization software should I choose
On 06/11/2009 11:30 AM, Nicolas Le Cam wrote: 2009/6/11 Paul Vriens: Hi, Just upgraded to Fedora 11 and (yet again) having issues with VMware Workstation. VMware still doesn't support Fedora (as a host) and I'm tired of fixing/patching things, again and again. I need to run several Windows versions (95 up to Vista for now) for our winetest and I really like the snapshot possibilities of VMware. Suggestions, recommendations? -- Cheers, Paul. I'm using VirtualBox. Apart from some problems around d3d/opengl and interruptions, it works quiet well. It also manages snapshots and have a CLI. See w2k-sp4-fr results on winetest. It has passed the test suite once (I didn't run latest builds but I will). Unfortunately, I don't know VMWare Workstation at all, so I can't tell you what are the differences. Just installed VirtualBox and it does have snaphots but not so extensive as VMware. In VMware I have some W2K snapshots: - out-of-the-box - SP1 - SP2 - SP3 - SP4 + Windows Update I can freely choose which I want to go to. Not so with VirtualBox I'm afraid as they are all on top of each other. (I guess for now I keep my other, F10, box around). -- Cheers, Paul.
Re: [Bugzilla]: Refuse to accept comments with big number of logs / back traces
Ken Sharp wrote: Nikolay Sivov wrote: What is the purpose of that change? To reduce mailed amount or what? To improve readability of reports? Log quotation are very useful sometimes and this blocking only makes things harder for people who knows what they're pasting. http://bugs.winehq.org/show_bug.cgi?id=5 There's nothing about why it's so important. These 25, 35 or 20 lines to allow means nothing. Actually I don't see any problem here, and, yes, I'm subscribed on wine-bugs list and this full log pasting occurs for 1 in 30 (maybe less a bit) new bugs reported. On new bug page we already have a rather big stop sign, if it isn't enough could we just add permissions to some bugzilla guys to wipe comments with logs pasted or it's out of bugzilla functionality? Attachments can be deleted, can comments be deleted? I don't know, that's why I've asked.
Re: [Bugzilla]: Refuse to accept comments with big number of logs / back traces
Nikolay Sivov wrote: What is the purpose of that change? To reduce mailed amount or what? To improve readability of reports? Log quotation are very useful sometimes and this blocking only makes things harder for people who knows what they're pasting. http://bugs.winehq.org/show_bug.cgi?id=5 On new bug page we already have a rather big stop sign, if it isn't enough could we just add permissions to some bugzilla guys to wipe comments with logs pasted or it's out of bugzilla functionality? Attachments can be deleted, can comments be deleted?
Re: [Bugzilla]: Refuse to accept comments with big number of logs / back traces
Ken Sharp wrote: I must have missed this. This would be incredibly useful. Does this affect all users though? Anastasius Focht would probably go over these limits but his comments are helpful. Vitaliy Margolen wrote: Vitaliy Margolen wrote: The limits are: 20 lines for logs (fixme, trace, err, warn) 10 lines for back traces --- Bugzilla/Bug.pm |7 +++ template/en/default/global/user-error.html.tmpl |4 2 files changed, 11 insertions(+), 0 deletions(-) So who is the one to commit this? Alexandre? Jeremy? Vitaliy. What is the purpose of that change? To reduce mailed amount or what? To improve readability of reports? Log quotation are very useful sometimes and this blocking only makes things harder for people who knows what they're pasting. On new bug page we already have a rather big stop sign, if it isn't enough could we just add permissions to some bugzilla guys to wipe comments with logs pasted or it's out of bugzilla functionality?
Re: [Bugzilla]: Refuse to accept comments with big number of logs / back traces
I must have missed this. This would be incredibly useful. Does this affect all users though? Anastasius Focht would probably go over these limits but his comments are helpful. Vitaliy Margolen wrote: Vitaliy Margolen wrote: The limits are: 20 lines for logs (fixme, trace, err, warn) 10 lines for back traces --- Bugzilla/Bug.pm |7 +++ template/en/default/global/user-error.html.tmpl |4 2 files changed, 11 insertions(+), 0 deletions(-) So who is the one to commit this? Alexandre? Jeremy? Vitaliy.
Re: Which virtualization software should I choose
Paul Vriens wrote: Just upgraded to Fedora 11 and (yet again) having issues with VMware Workstation. VMware still doesn't support Fedora (as a host) and I'm tired of fixing/patching things, again and again. I need to run several Windows versions (95 up to Vista for now) for our winetest and I really like the snapshot possibilities of VMware. Suggestions, recommendations? I'm using KVM/Qemu with libvirt aka virt-manager. I have problems with WinNT and NetBSD; those hang/crash during the install. Win2k3, FreeBSD and OpenSolaris work just fine and I didn't try anything else yet. But I'm on F9 and once I have a little spare time I'll move to F11. I'll test it there (never qemu) if that fixes WinNT. bye michael
Re: winemenubuilder: generate new menus in the right directory
Damjan Jovanovic writes:
> @@ -882,7 +882,11 @@ static BOOL write_menu_file(const char *filename)
>
> while (1)
> {
> +char *tmpdir = getenv("TMPDIR");
> +unsetenv("TMPDIR");
> tempfilename = tempnam(xdg_config_dir, "_wine");
> +if (tmpdir)
> +setenv("TMPDIR", tmpdir, 1);
That's ugly. You should use mkstemps() or something like that.
--
Alexandre Julliard
julli...@winehq.org
Re: Fix incorrect check in WCMD_HandleTildaModifiers().
Gerald Pfeifer writes: >>= '0' || <= '9' always is true. :-) This patch fixes it, and > avoids two just hand coded checks altogether. > > Gerald > > ChangeLog: > Fix incorrect check in WCMD_HandleTildaModifiers(). > Use isdigit() instead of hand-coding this check. Using isdigit() on a Unicode char is not a good idea (and in this case isdigitW() wouldn't be a good idea either). -- Alexandre Julliard julli...@winehq.org
Re: Which virtualization software should I choose
2009/6/11 Paul Vriens : > Hi, > > Just upgraded to Fedora 11 and (yet again) having issues with VMware > Workstation. > > VMware still doesn't support Fedora (as a host) and I'm tired of > fixing/patching things, again and again. > > I need to run several Windows versions (95 up to Vista for now) for our > winetest and I really like the snapshot possibilities of VMware. > > Suggestions, recommendations? > > -- > Cheers, > > Paul. > > > I'm using VirtualBox. Apart from some problems around d3d/opengl and interruptions, it works quiet well. It also manages snapshots and have a CLI. See w2k-sp4-fr results on winetest. It has passed the test suite once (I didn't run latest builds but I will). Unfortunately, I don't know VMWare Workstation at all, so I can't tell you what are the differences. -- Nicolas Le Cam
Re: opengl32: get rid of glext.h
On Thu, Jun 11, 2009 at 11:09 AM, Detlef Riekenberg wrote: > On Mi, 2009-06-10 at 21:56 +0200, Roderick Colenbrander wrote: > >> -#ifdef HAVE_GL_GLEXT_H >> -# include >> -#endif > > The same code is used in winex11.drv/opengl.c > Can this also be removed? > > When that's the case, then "GL/glext.h" can be removed from > AC_CHECK_HEADERS in configure.ac > > -- > > By by ... Detlef > > At a quick glance we don't seem to need it there. We only use a small number of GL calls which are all 1.1 and for extensions we use in there we already define our own function pointers anyway as the functions need to be loaded dynamically. I think we can get rid of it. Roderick
Which virtualization software should I choose
Hi, Just upgraded to Fedora 11 and (yet again) having issues with VMware Workstation. VMware still doesn't support Fedora (as a host) and I'm tired of fixing/patching things, again and again. I need to run several Windows versions (95 up to Vista for now) for our winetest and I really like the snapshot possibilities of VMware. Suggestions, recommendations? -- Cheers, Paul.
Re: opengl32: get rid of glext.h
On Mi, 2009-06-10 at 21:56 +0200, Roderick Colenbrander wrote: > -#ifdef HAVE_GL_GLEXT_H > -# include > -#endif The same code is used in winex11.drv/opengl.c Can this also be removed? When that's the case, then "GL/glext.h" can be removed from AC_CHECK_HEADERS in configure.ac -- By by ... Detlef
