Date: Fri, 29 Oct 2004 00:04:20 EST-10EDT,10,-1,0,7200,3,-1,0,7200,3600 Subject: [UnixOS2_Archive] No. 431 ************************************************** Thursday 28 October 2004 Number 431 ************************************************** Subjects for today 1 Re: Posix/2 sys\select.h & NBBY : Stefan.Neis at t-online.de 2 Re: Subversion : Brian Havard" 3 Re: Eureka! : John Poltorak 4 Re: Posix/2 sys\select.h & NBBY : John Poltorak 5 Re: Posix/2 sys\select.h & NBBY : Stefan.Neis at t-online.de 6 Re: Eureka! : Dave Saville" 7 Re: Eureka! : John Poltorak 8 IMAPD : John Poltorak **= Email 1 ==========================** Date: Wed, 27 Oct 2004 15:45:57 +0100 From: Stefan.Neis at t-online.de Subject: Re: Posix/2 sys\select.h & NBBY Hi, > > include\sys\types.h:#define NBBY 8 /* number of bits per byte; CHAR_BIT */ > configure:6896: gcc -c -Zmt -D__ST_MT_ERRNO__ -D_XOPEN_SOURCE=500 > -D_POSIX_C_SOURCE -DNDEBUG conftest.c 1>&5 > In file included from configure:6888: > t:\ux2bs\posix2\include\sys/select.h:26: `NBBY' undeclared here (not in a > function) Well, that definition in sys\types.h is inside #if !defined (_POSIX_SOURCE) && !defined(_XOPEN_SOURCE) ... #endif Since gcc command like explicitly defines _XOPEN_SOURCE (and maybe implicitly also _POSIX_SOURCE (via _POSIX_C_SOURCE), it's not a big surprise that NBBY doesn't get defined in this instance... The interesting question is what the right fix would be... Specifically in view of line 152 (the same #if check nested into the condition starting at line 97), I'd guess it's an error in Posix/2's sys\types.h and the #endif (line 175) should be moved to line 108 (i.e. right after typedef int64_t int64;). Regards, Stefan **= Email 2 ==========================** Date: Wed, 27 Oct 2004 23:44:44 +1000 (EST) From: "Brian Havard" Subject: Re: Subversion On Wed, 27 Oct 2004 14:02:58 +0100, John Poltorak wrote: [...] >Doing a little reading around the place I've seen something about >Subversion using WebDAV as its network protocol. Does that mean I can use >Apache as the server for a version control system? I assume I would need to >install mod_dav... Is that available on OS/? > >Apologies for being hazy on this - I'm just trying to get to grips with >WebDAV. mod_dav is included with Apache 2.0. However you still need mod_dav_svn which is part of subversion. I haven't tried building that yet. svnserve (server for the svn: protocol) might also work but again, I haven't tried it. -- ______________________________________________________________________________ | Brian Havard | "He is not the messiah! | | brianh at kheldar.apana.org.au | He's a very naughty boy!" - Life of Brian | ------------------------------------------------------------------------------ **= Email 3 ==========================** Date: Wed, 27 Oct 2004 16:17:21 +0100 From: John Poltorak Subject: Re: Eureka! On Wed, Oct 27, 2004 at 09:00:46AM -0400, Henry Sobotka wrote: > John Poltorak wrote: > > > > It took around nine hours in total but I did end up with a mozilla.exe... > > unfortunately it just hangs after the yellow mozilla splash box comes up. > > Launching it with "mozilla > mozlog 2>&1" might provide a clue. No - nothing. The log just shows the full pathname of mozilla and then Process terminted by SIGTERM I killed the process after about 15 mins as it wasn't doing anything. > Make sure you don't have any other instances of moz running or envars > such as MOZILLA_HOME pointing elsewhere. > > And if your moz is the end-product of a stop/start build, wiping out > your objdir and rebuilding from scratch so that you get an uninterrupted > build from start to finish should fix whatever's wrong. I had managed to build it previously in about four goes after filling up the disk space, but then disabled debug options and rebuilt it from the start using a different OBJDIR. Is there anything I should check the build log for? May another build, this time with debug enabled, is in order.. > h~ -- John **= Email 4 ==========================** Date: Wed, 27 Oct 2004 19:51:44 +0100 From: John Poltorak Subject: Re: Posix/2 sys\select.h & NBBY On Wed, Oct 27, 2004 at 03:45:57PM +0100, Stefan.Neis at t-online.de wrote: > The interesting question is what the right fix would be... > > Specifically in view of line 152 (the same #if check nested into > the condition starting at line 97), I'd guess it's an error in > Posix/2's sys\types.h and the #endif (line 175) should be moved to > line 108 (i.e. right after typedef int64_t int64;). I wish I could offer some informed opinion but.... Do you see this as a Posix/2 problem and if so will a fix become available? I'm in no big rush but it would be nice to keep up with apps like NCURSES that have been OS/2 friendly over the years. As for the longer term replacement of Posix/2 by Innotek's gcc, how do we ensure that any updates made by you will find their way into Innotek's headers? At some point we need to establish that Posix/2 is redundant and is completely superceded by Innotek's offering. It would be useful to try and agree on some sort of procedure... > Regards, > Stefan > -- John **= Email 5 ==========================** Date: Wed, 27 Oct 2004 22:37:46 +0100 From: Stefan.Neis at t-online.de Subject: Re: Posix/2 sys\select.h & NBBY Hi, > I wish I could offer some informed opinion but.... Well, your hint to look at Innotek's libc was a good one. ;-) > Do you see this as a Posix/2 problem and if so will a fix become > available? Yes, that #endif definetly should be moved. Unfortunately, CVS access seems to be broken currently ... :-( > As for the longer term replacement of Posix/2 by Innotek's gcc, how do > we ensure that any updates made by you will find their way into Innotek's > headers? Actually, Innotek is using a fresh copy of BSD headers, so they typically avoid the problems that Posix/2 introduced by relatively many changes needed to make everything work with the existing EMX libraries... > At some point we need to establish that Posix/2 is redundant and > is completely superceded by Innotek's offering. It would be useful > to try and agree on some sort of procedure... I have no idea, what that could possibly be except for compiling everything _requiring_ Posix/2 with Innotek and check if it works. Personally, I do think that Innotek's compiler (the current alpha version adding "fork" and other functions) is what I was waiting for... Now, if somebody manages to compile XFree86 and a couple of libraries with that compiler, I'm ready to forget about EMX+Posix/2, especially, since I don't typically need to debug my software on OS/2 (though it was often helpful to be able to do that in the past). Regards, Stefan **= Email 6 ==========================** Date: Thu, 28 Oct 2004 08:21:58 +0100 (BST) From: "Dave Saville" Subject: Re: Eureka! On Wed, 27 Oct 2004 16:17:21 +0100, John Poltorak wrote: >On Wed, Oct 27, 2004 at 09:00:46AM -0400, Henry Sobotka wrote: >> John Poltorak wrote: >> > >> > It took around nine hours in total but I did end up with a mozilla.exe... >> > unfortunately it just hangs after the yellow mozilla splash box comes up. Doubt if it helps but I had the same problem with the official builds through several versions. Never did suss it out. I just kept trying new versions and one day it worked. The only change was the moz version - nothing else had. -- Regards Dave Saville **= Email 7 ==========================** Date: Thu, 28 Oct 2004 10:05:30 +0100 From: John Poltorak Subject: Re: Eureka! --rAm5ImPkZ2FSQHOL Content-Type: text/plain; charset=us-ascii On Wed, Oct 27, 2004 at 09:00:46AM -0400, Henry Sobotka wrote: > John Poltorak wrote: > > > > It took around nine hours in total but I did end up with a mozilla.exe... > > unfortunately it just hangs after the yellow mozilla splash box comes up. > > Launching it with "mozilla > mozlog 2>&1" might provide a clue. After a rebuild with debug enabled I get the attached log... Among other things it suggests the file gfx_os2.dll is missing although I have found several instances of it throughout the subdirectory so maybe it has not been built correctly or hasn't been moved to the expected location. I did copy it to my libpath, but that made no difference. Is there anything else I should look at? > h~ -- John --rAm5ImPkZ2FSQHOL Content-Type: text/plain; charset=us-ascii Content-Disposition: attachment; filename="P:\\unixos2\\logs\\mozlog" nsNativeComponentLoader: autoregistering begins. nsNativeComponentLoader: SelfRegisterDll(gfx_os2.dll) Load FAILED with error: No such file or directory nsNativeComponentLoader: autoregistering succeeded nNCL: registering deferred (0) nsNativeComponentLoader: autoregistering begins. nsNativeComponentLoader: SelfRegisterDll(gfx_os2.dll) Load FAILED with error: No such file or directory nsNativeComponentLoader: autoregistering succeeded nNCL: registering deferred (0) Type Manifest File: G:\UX2BS\WORKDIR\MOZILLA\OBJ\DIST\BIN\components\xpti.dat +++ JavaScript debugging hooks installed. nsNativeComponentLoader: autoregistering begins. nsNativeComponentLoader: SelfRegisterDll(gfx_os2.dll) Load FAILED with error: No such file or directory nsNativeComponentLoader: autoregistering succeeded nNCL: registering deferred (0) ###!!! ASSERTION: running socket transport thread without a pollable event: 'mThreadEvent', file G:/ux2bs/workdir/mozilla/netwerk/base/src/nsSocketTransportService2.cpp, line 386 Couldn't find DC instance for nsWindow ++WEBSHELL == 1 ++DOMWINDOW == 1 WARNING: NS_ENSURE_TRUE(NS_SUCCEEDED(rv)) failed, file G:/ux2bs/workdir/mozilla/extensions/cookie/nsPermissionManager.cpp, line 635 Couldn't find DC instance for nsWindow ++WEBSHELL == 2 ++DOMWINDOW == 2 WARNING: NS_ENSURE_TRUE(mDeviceContext) failed, file G:/ux2bs/workdir/mozilla/docshell/base/nsDocShell.cpp, line 4398 WARNING: NS_ENSURE_TRUE(NS_SUCCEEDED(EnsureDeviceContext())) failed, file G:/ux2bs/workdir/mozilla/docshell/base/nsDocShell.cpp, line 4806 WARNING: NS_ENSURE_TRUE(global) failed, file G:/ux2bs/workdir/mozilla/content/base/src/nsDocumentViewer.cpp, line 896 WARNING: NS_ENSURE_TRUE(mDeviceContext) failed, file G:/ux2bs/workdir/mozilla/docshell/base/nsDocShell.cpp, line 4398 WARNING: NS_ENSURE_TRUE(NS_SUCCEEDED(EnsureDeviceContext())) failed, file G:/ux2bs/workdir/mozilla/docshell/base/nsDocShell.cpp, line 4806 ###!!! ASSERTION: no script global object: 'mScriptGlobalObject != nsnull', file G:/ux2bs/workdir/mozilla/content/xul/document/src/nsXULDocument.cpp, line 3304 ###!!! ASSERTION: no script global object: 'mScriptGlobalObject != nsnull', file G:/ux2bs/workdir/mozilla/content/xul/document/src/nsXULDocument.cpp, line 3304 ###!!! ASSERTION: no script global object: 'mScriptGlobalObject != nsnull', file G:/ux2bs/workdir/mozilla/content/xul/document/src/nsXULDocument.cpp, line 3304 ###!!! ASSERTION: no script global object: 'mScriptGlobalObject != nsnull', file G:/ux2bs/workdir/mozilla/content/xul/document/src/nsXULDocument.cpp, line 3304 ###!!! ASSERTION: no script global object: 'mScriptGlobalObject != nsnull', file G:/ux2bs/workdir/mozilla/content/xul/document/src/nsXULDocument.cpp, line 3304 ###!!! ASSERTION: no script global object: 'mScriptGlobalObject != nsnull', file G:/ux2bs/workdir/mozilla/content/xul/document/src/nsXULDocument.cpp, line 3304 ###!!! ASSERTION: no script global object: 'mScriptGlobalObject != nsnull', file G:/ux2bs/workdir/mozilla/content/xul/document/src/nsXULDocument.cpp, line 3304 ###!!! ASSERTION: no script global object: 'mScriptGlobalObject != nsnull', file G:/ux2bs/workdir/mozilla/content/xul/document/src/nsXULDocument.cpp, line 3304 ###!!! ASSERTION: no script global object: 'mScriptGlobalObject != nsnull', file G:/ux2bs/workdir/mozilla/content/xul/document/src/nsXULDocument.cpp, line 3304 ###!!! ASSERTION: no script global object: 'mScriptGlobalObject != nsnull', file G:/ux2bs/workdir/mozilla/content/xul/document/src/nsXULDocument.cpp, line 3304 ###!!! ASSERTION: no script global object: 'mScriptGlobalObject != nsnull', file G:/ux2bs/workdir/mozilla/content/xul/document/src/nsXULDocument.cpp, line 3304 ###!!! ASSERTION: no script global object: 'mScriptGlobalObject != nsnull', file G:/ux2bs/workdir/mozilla/content/xul/document/src/nsXULDocument.cpp, line 3304 ###!!! ASSERTION: no script global object: 'mScriptGlobalObject != nsnull', file G:/ux2bs/workdir/mozilla/content/xul/document/src/nsXULDocument.cpp, line 3304 ###!!! ASSERTION: no script global object: 'mScriptGlobalObject != nsnull', file G:/ux2bs/workdir/mozilla/content/xul/document/src/nsXULDocument.cpp, line 3304 ###!!! ASSERTION: no script global object: 'mScriptGlobalObject != nsnull', file G:/ux2bs/workdir/mozilla/content/xul/document/src/nsXULDocument.cpp, line 3304 ###!!! ASSERTION: no script global object: 'mScriptGlobalObject != nsnull', file G:/ux2bs/workdir/mozilla/content/xul/document/src/nsXULDocument.cpp, line 3304 ###!!! ASSERTION: no script global object: 'mScriptGlobalObject != nsnull', file G:/ux2bs/workdir/mozilla/content/xul/document/src/nsXULDocument.cpp, line 3304 ###!!! ASSERTION: no script global object: 'mScriptGlobalObject != nsnull', file G:/ux2bs/workdir/mozilla/content/xul/document/src/nsXULDocument.cpp, line 3304 ###!!! ASSERTION: no script global object: 'mScriptGlobalObject != nsnull', file G:/ux2bs/workdir/mozilla/content/xul/document/src/nsXULDocument.cpp, line 3304 ###!!! ASSERTION: no script global object: 'mScriptGlobalObject != nsnull', file G:/ux2bs/workdir/mozilla/content/xul/document/src/nsXULDocument.cpp, line 3304 ###!!! ASSERTION: no script global object: 'mScriptGlobalObject != nsnull', file G:/ux2bs/workdir/mozilla/content/xul/document/src/nsXULDocument.cpp, line 3304 mozilla.exe: Process terminated by SIGINT nsStringStats => mAllocCount: 5325 => mReallocCount: 82 => mFreeCount: 4131 => mShareCount: 7038 => mAdoptCount: 2026 => mAdoptFreeCount: 2010 --rAm5ImPkZ2FSQHOL-- **= Email 8 ==========================** Date: Thu, 28 Oct 2004 11:34:35 +0100 From: John Poltorak Subject: IMAPD Has anyone tried setting up IMAPD from the latest PINE (v4.61) release? I've set up IPOP3D and IMAPD to be run from the ported version of INETD (not IBM's) but can't get a good login. Is there any debug flag I can use with either ipop3d or imapd to output error msgs, possibly to syslog to show what is taking place? -- John