October 2003 - Wine-devel - list.winehq.org

GetProcessWindowStation
by Mathias Berchtold 22 Oct '03

22 Oct '03

Hello ... I get the following error when trying to run SmartFTP on Wine. fixme:advapi:RevertToSelf (), stub fixme:advapi:SetThreadToken ((nil), 0xa0): stub (NT impl. only) fixme:win:GetProcessWindowStation (void): stub fixme:win32:GetUserObjectInformationA (0x1 1 0x4078e8e4 12 0x4078e8d8),stub! And afer that I get a Debug ASSERT from the Microsoft Visual C++ Library. File: Allocate.cpp Line: 108 Expression: bSuccess I think the Assert is caused because of the missing function implementations. Do you have any plans to implement them in the future? Wine Information ================ wine --version Wine 20031016 windows="win2k" ("win98" doesn't work because wine "ignores" the unicows.dll) [AppDefaults\\smartftp.exe\\DllOverrides] "ole32" = "native" "oleaut32" = "native" "rpcrt4" = "native" "urlmon" = "native" "comctl32" = "builtin" "wininet" = "builtin" "shdocvw" = "native" "shlwapi" = "native" "crypt32" = "native" "mshtml" = "native" "unicows" = "native" "crypt32" = "native" "cryptui" = "native" "wintrust" = "native" If you need the debug build of SmartFTP to test please contact me. Thanks a lot -Mat SmartFTP.com -----Original Message----- From: wine-devel-admin(a)winehq.org [mailto:wine-devel-admin(a)winehq.org] On Behalf Of Alexandre Julliard Sent: Mittwoch, 22. Oktober 2003 04:20 To: Vincent Béron Cc: Wine devel Subject: Re: Changes to kernel messages lang Vincent Béron <vberon(a)mecano.gme.usherb.ca> writes: > Attached program says so on W2K SP4. There's a message table in > kernel32 for 0x0000, for the others it returns NULL. Is that way to do > it different from yours? Yes, FindResource with language 0 is going to return a resource in all sorts of cases, so it's not a good test. And actually if the resource was really NEUTRAL, then all your FindResource calls should have succeeded. The best way to find out is to dump the resource table of the dll, using something like winedump -x. -- Alexandre Julliard julliard(a)winehq.com

2 1

Re: NTFS driver (was: ReactOS GPL vs. proprietary drivers)
by Fabian Cenedese 22 Oct '03

22 Oct '03

Hi I don't know much about NTFS (only using it) and even less about Linux-NTFS. But it was funny following this conversation. >> Filesystem must be the rock solid data storage structure. You must >> know the meaning of each byte (*) for such reliable and interoperable >> filesystem. > >Exactly. Every needed byte is known. Wow! Really? >> (*) You do not need to know the journalling metadata as long as you do not >> support journalling and/or its recovery. > >One of the unknown issues is journaling :) AHA! And not only "The unknown issue" but "One of the unknown issues". Reminds me of Asterix: Whole Europe is part of the Roman Empire. Whole...? Sorry, I couldn't resist, no offense :) bye Fabi

2 1

Xlib and wine and XGrabButton
by Gerhard W. Gruber 22 Oct '03

22 Oct '03

I currently write a small desktop app (in case you are curious http://sourceforge.net/projects/launchmenu) and for this I have to research on how to grab mouse events and eventually keyboard as well. Now I thought that this might be usefull for wine when playing games under the DirectX emulation, so I checked the sources for XGrabButton()/XGrabKey() and I couldn't find any references to it. When I tried to play Thief and Thief 2 I always had the problem that the keys are going to the shell where I started it from, instead of going into the game. I wrote a small test app for my project in order to get this feature straight and there I could capture all mouse events (haven't tried keyboard, but it should work equally well) without any interference with the remaining X system. Is this intenttionally that these functions are not used and if so what is the reason for it? Or is there a layer in between wine and X so that wine is not using Xlib API directly? I think that using this should work at least in programs where the entire screen is taken over, like in games, probably other applications well. -- Gerhard Gruber Für jedes menschliche Problem gibt es immer eine einfache Lösung: Klar, einleuchtend und falsch. (Henry Louis Mencken)

3 4

Re: Changes to kernel messages lang
by Alexandre Julliard 22 Oct '03

22 Oct '03

Vincent Béron <vberon(a)mecano.gme.usherb.ca> writes: > W2k kernel messagetable is LANG_NEUTRAL, SUBLANG_NEUTRAL only. > > Changelog: > Make the kernel messagetable LANG_NEUTRAL, as native. Are you sure about that one? All the dlls I have checked (Win95, NT4, XP) have the language set to 0x0409. In what case is that causing trouble? -- Alexandre Julliard julliard(a)winehq.com

2 3

Re: More dutch translations
by Alexandre Julliard 21 Oct '03

21 Oct '03

Hans Leidekker <hans(a)it.vu.nl> writes: > Changelog: > - More resources translated to dutch. > - Make an example consistent with it's description (wcmd). The wcmd change is wrong, the default prompt really is $p$g. Also your patch got wrapped, you'll have to resubmit. -- Alexandre Julliard julliard(a)winehq.com

1 0

S3 texture compression patent
by Jonathan Wilson 21 Oct '03

21 Oct '03

What specificly does this patent cover and what parts of WINE would it affect? (i.e. which bits of windows, if implemented in WINE, would fall under the patent?)

6 7

Re: Summary of translations
by Vincent Béron 21 Oct '03

21 Oct '03

Le mar 21/10/2003 à 14:45, Hans Leidekker a écrit : > This is certainly useful for me, as I was just about to start translating > some dlls to dutch. I checked the page you mention above and it showed > there are some parts missing from dlls that already haven been translated > to dutch. I don't think I would have found those otherwise (I wouldn't have > gone looking for them). So I think this tool can certainly prove valuable in > keeping resources up to date. That's why I did it (after I saw Dimi's output). Although now, all French resources seem to be translated... :) Glad to know it helped you too. > > And now for some criticisms. I can understand that you only check languages > for which at least one resource exists because otherwise the table would grow > large and show lots of empty cells for years to come, but the first thing I need > to set out translating is a list of dlls that have translatable resources, which I > now obtain by doing: > > [wine-src] $ find dlls -name "*En.rc" > > Which is crude and depends on a file naming convention. Maybe your tool > could help here too. If there's an html file, the dll contains some resources. Now, not all resources are translatable (icons, cursors, bitmaps we probably don't care, not sure about versioninfo). That can be integrated in the tool too. For now, if there's only a LANG_NEUTRAL version of a resource, it doesn't complain, else it creates the file. > > Second, those filenames for the html output really aren't fit to be served > by apache. They're truncated in a directory listing page so I have to hover > the links to be able to find a particular file. I think you need to either split > them into directories or provide an index page. As Dimi pointed out, all on a single page would be better. Or a couple of pages, with AAFRIKANS-GERMAN on one, HUNGARIAN-SLOVAK, etc. to prevent the table from being too wide, and marking only the relevant languages (ie, those for which at least one translated resource exists). Another thing is languages for which more than one sublanguage is used, such as Chinese. > > Third, I think you should sort the table columns alphabetically by language > name. But they're sorted! Ok, by language ID :) Good idea. > > Lastly, Konqueror doesn't render those popup/layer thingies well, and in > Mozilla I can't can rid of a popup once I clicked a link (well, I can press > the Back button, but that's non-intuitive to say the least). To be honest I > think they're overkill. Make them a regular link. Remember, you said all > comments were welcome ;-) No problem with that. I tried to get the same feeling as Mozilla's Tinderbox, but I can make them regular links instead. Or "tooltips". Vincent

1 0

Re: NTFS driver (was: ReactOS GPL vs. proprietary drivers)
by Szakacsits Szabolcs 21 Oct '03

21 Oct '03

> > Let me tell an example, the (Linux) NTFS driver supports transparent > > compression. Today CPU's are very fast and the disk bandwidth is very > > slow compared. So doing bulk data transfers, the disk bandwidth is the > > bootleneck. Would you be faster using the filesystem's transparent > > compression? In theory yes, CPU can [de]compress when transfering data > > and you must also transfer less data. No main Linux fs supports this, > > hopefully Reiser4 will in the future ... (there was hope for e2compr as > > well for many years but it didn't work out after all). > > ext2 have supported this for ages, check out chattr(1),lsattr(1) This feature is reserved for about 8 years. But it never worked reliable. Google for the story. I wrote above the keyword (e2compr). > In what way are the compression features in NTFS different? It works reliable since about 8 years. Something what none of the main Linux filesystem can do today (ext3, reiserfs, JFS, XFS). Szaka

1 0

Re: [Rebar] Don't store band text if "" is passed into SetBandInfo
by Huw D M Davies 21 Oct '03

21 Oct '03

On Tue, Oct 21, 2003 at 07:24:31PM +0100, Robert Shearman wrote: > Hi, > > This makes IE look a bit neater. > > Rob > > Changelog: > - Don't store band text if "" is passed into SetBandInfo > @@ -3447,8 +3450,11 @@ > } > if (lprbbi->lpText) { > INT len = lstrlenW (lprbbi->lpText); > - lpBand->lpText = (LPWSTR)Alloc ((len + 1)*sizeof(WCHAR)); > - strcpyW (lpBand->lpText, lprbbi->lpText); > + if (len > 1) Shouldn't this be len >= 1 ? > + { > + lpBand->lpText = (LPWSTR)Alloc ((len + 1)*sizeof(WCHAR)); > + strcpyW (lpBand->lpText, lprbbi->lpText); > + } > } > } > Huw.

1 0

Re: Fix comctl32 conformance test
by Rolf Kalbermatter 21 Oct '03

21 Oct '03

Alexandre Julliard <julliard(a)winehq.org> wrote: <Hans Leidekker <hans(a)it.vu.nl> writes: > >> It appears that the comctl32 conformance test calls functions by name >> that are exported by ordinal only on Windows. Still the test runs (and >> thus run-time links) on Wine. Is this wanted behavior? > >Yes it is, the functions are converted to ordinal imports by the >import library. The existing code works just fine, except when >cross-compiling on Mingw because the import libraries are broken >there. This should be fixed in Mingw not in Wine. Mmh! I have downloaded the PSDK from the MSDN site, and although it says Microsoft Platform SDK February 2003 in the Start menu it really was just downloaded one month ago and my comctl32.lib does not contain symbols for at least DPA_Create() and DPA_GetPtr() while it does contain most of the other DPA symbols. What am I missing here? Rolf Kalbermatter

3 2