Re: [ITP] ec-fonts-mftraced
The ec-fonts-mftraced postinstall script will try to run mktexlsr. Perhaps we should check if /bin/mktexlsr exists before calling it? The other thing: post-texmf does a 'texconfig rehash' that is equivalent to mktexlsr, so after installing tex, it will find the fonts. So I suppose that this dependency is not needed. However, before I upload, I will check that it works along with lilypond. Bert You'll only have to check what happens (postinstall scripts) if you install ec-fonts-mftraced first, and install tetex later.
Re: [ITP] ec-fonts-mftraced
Bertalan Fodor writes: The ec-fonts-mftraced postinstall script will try to run mktexlsr. Perhaps we should check if /bin/mktexlsr exists before calling it? Yes. The other thing: post-texmf does a 'texconfig rehash' that is equivalent to mktexlsr, so after installing tex, it will find the fonts. Yes. So I suppose that this dependency is not needed. However, before I upload, I will check that it works along with lilypond. Very good. Jan. -- Jan Nieuwenhuizen [EMAIL PROTECTED] | GNU LilyPond - The music typesetter http://www.xs4all.nl/~jantien | http://www.lilypond.org
Re: [ITP] mathomatic-11.5-1
Reini schrieb: I've updated my not yet approved package of mathomatic to the latest upstream version. [EMAIL PROTECTED] ~]$ mv mathomatic/ /var/ftp/pub/cygwin/release/ Gerrit -- =^..^=
Do we still have a ghostscript maintainer?
I notice that there hasn't been a new release of ghostscript in some time. We need a new release to eliminate some of the lurking dependencies on the deprecated XFree86. Dario, are you still with us? If so, is there any chance that we could get a new version of ghostscript sometime soon? cgf
[ITP] gdk-pixbuf-0.22.0-1
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 I would like to contribute gdk-pixbuf to the Cygwin distribution. As this is included in Fedora Core and Debian testing, all it needs is a GTG review. http://cygwin-ports.sourceforge.net/install/release/gdk-pixbuf/gdk-pixbuf-0.22.0-1-src.tar.bz2 http://cygwin-ports.sourceforge.net/install/release/gdk-pixbuf/gdk-pixbuf-0.22.0-1.tar.bz2 http://cygwin-ports.sourceforge.net/install/release/gdk-pixbuf/setup.hint category: Graphics Libs X11 requires: cygwin glib gtk+ libiconv2 libintl3 libjpeg62 libpng12 libtiff5 zlib sdesc: Gdk and X11 image loading library ldesc: The gdk-pixbuf library provides a number of features; ~ - GdkPixbuf structure for representing images. ~ - Image loading facilities, both synchronous and progressive. ~ - Rendering of a GdkPixbuf into various formats; ~drawables (windows, pixmaps), GdkRGB buffers. ~ - Fast scaling and compositing of pixbufs. ~ - Simple animation loading (ie. animated gifs). ~ - Xlib port of gdk-pixbuf. This included library lets you use ~the gdk-pixbuf rendering functions on plain Xlib programs ~that do not use Gdk. Yaakov -BEGIN PGP SIGNATURE- Version: GnuPG v1.2.4 (Cygwin) Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org iD8DBQFBlDuJpiWmPGlmQSMRAmGnAKDTskos7KdmjGgzkky2UdMLm//RgACdHEIL 7Wx6OLeL1lQWS7qHyv40OUs= =H0Zw -END PGP SIGNATURE-
Re: Please upload: desktop-file-utils-0.10-1
Yaakov Selkowitz wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Please upload this at your earliest convienence. http://cygwin-ports.sourceforge.net/install/temp/desktop-file-utils/desktop-file-utils-0.10-1-src.tar.bz2 http://cygwin-ports.sourceforge.net/install/temp/desktop-file-utils/desktop-file-utils-0.10-1.tar.bz2 Uploaded. Gerrit -- =^..^=
Re: xorg installation fails, 99% complete only
the third setupfull3.txt.bz2 Description: Binary data
Re: xorg installation fails, 99% complete only
(As I had problems sending this mail, you'll receive it as the last one, excuse me, there is only the setup.log, before you have the three last compressed part of the setup full and after the first part) Good morning this time, Ok I'm here again. The situation. The problem occurs when I do a complete install of cygwin. It blocks every time on the same load (see before). Wen cancelling, no setup.log is created (I think it is normal as there is no post install made). If after this first failed install I do a setup with only the X11 package there, then it goes to the end of install and here is the setup.log and the first of four compressed part of a setup.log.full (because of the size as you know). Following, the last three part. BPatin 2004/11/11 10:39:55 Starting cygwin install, version 2.427 2004/11/11 10:39:55 Current Directory: C:\Documents and Settings\bruno\Bureau\08112004 2004/11/11 10:39:55 Changing gid to Users 2004/11/11 10:39:55 Could not open service McShield for query, start and stop. McAfee may not be installed, or we don't have access. 2004/11/11 10:40:01 source: from cwd 2004/11/11 10:40:03 root: C:\cygwin binary user 2004/11/11 10:40:05 Selected local directory: C:\Documents and Settings\bruno\Bureau\08112004 2004/11/11 10:41:12 Installing file://C:\Documents and Settings\bruno\Bureau\08112004/http%3a%2f%2fmirrors.sunsite.dk%2fcygwin/release/ash/ash-20040127-1.tar.bz2 2004/11/11 10:41:12 Installing file://C:\Documents and Settings\bruno\Bureau\08112004/http%3a%2f%2fmirrors.sunsite.dk%2fcygwin/release/base-files/base-files-3.0-3.tar.bz2 2004/11/11 10:41:12 Installing file://C:\Documents and Settings\bruno\Bureau\08112004/http%3a%2f%2fmirrors.sunsite.dk%2fcygwin/release/base-passwd/base-passwd-2.1-1.tar.bz2 2004/11/11 10:41:12 Installing file://C:\Documents and Settings\bruno\Bureau\08112004/http%3a%2f%2fmirrors.sunsite.dk%2fcygwin/release/bash/bash-2.05b-16.tar.bz2 2004/11/11 10:41:13 Installing file://C:\Documents and Settings\bruno\Bureau\08112004/http%3a%2f%2fmirrors.sunsite.dk%2fcygwin/release/bzip2/bzip2-1.0.2-6.tar.bz2 2004/11/11 10:41:14 Installing file://C:\Documents and Settings\bruno\Bureau\08112004/http%3a%2f%2fmirrors.sunsite.dk%2fcygwin/release/X11/cgoban/cgoban-1.9.14-1.tar.bz2 2004/11/11 10:41:15 Installing file://C:\Documents and Settings\bruno\Bureau\08112004/http%3a%2f%2fmirrors.sunsite.dk%2fcygwin/release/cygipc/cygipc-2.03-2.tar.bz2 2004/11/11 10:41:16 Installing file://C:\Documents and Settings\bruno\Bureau\08112004/http%3a%2f%2fmirrors.sunsite.dk%2fcygwin/release/cygutils/cygutils-1.2.5-1.tar.bz2 2004/11/11 10:41:19 Installing file://C:\Documents and Settings\bruno\Bureau\08112004/http%3a%2f%2fmirrors.sunsite.dk%2fcygwin/release/cygwin/cygwin-1.5.11-1.tar.bz2 2004/11/11 10:41:23 Installing file://C:\Documents and Settings\bruno\Bureau\08112004/http%3a%2f%2fmirrors.sunsite.dk%2fcygwin/release/cygwin-doc/cygwin-doc-1.3-7.tar.bz2 2004/11/11 10:41:28 Installing file://C:\Documents and Settings\bruno\Bureau\08112004/http%3a%2f%2fmirrors.sunsite.dk%2fcygwin/release/X11/cygwin-x-doc/cygwin-x-doc-1.0.4-1.tar.bz2 2004/11/11 10:41:30 Installing file://C:\Documents and Settings\bruno\Bureau\08112004/http%3a%2f%2fmirrors.sunsite.dk%2fcygwin/release/X11/ddd/ddd-3.3.9-1.tar.bz2 2004/11/11 10:41:31 Installing file://C:\Documents and Settings\bruno\Bureau\08112004/http%3a%2f%2fmirrors.sunsite.dk%2fcygwin/release/diffutils/diffutils-2.8.7-1.tar.bz2 2004/11/11 10:41:34 Installing file://C:\Documents and Settings\bruno\Bureau\08112004/http%3a%2f%2fmirrors.sunsite.dk%2fcygwin/release/editrights/editrights-1.01-1.tar.bz2 2004/11/11 10:41:34 Installing file://C:\Documents and Settings\bruno\Bureau\08112004/http%3a%2f%2fmirrors.sunsite.dk%2fcygwin/release/expat/expat-1.95.8-1.tar.bz2 2004/11/11 10:41:34 Installing file://C:\Documents and Settings\bruno\Bureau\08112004/http%3a%2f%2fmirrors.sunsite.dk%2fcygwin/release/fileutils/fileutils-4.1-2.tar.bz2 2004/11/11 10:41:36 Installing file://C:\Documents and Settings\bruno\Bureau\08112004/http%3a%2f%2fmirrors.sunsite.dk%2fcygwin/release/findutils/findutils-4.1.7-4.tar.bz2 2004/11/11 10:41:37 Installing file://C:\Documents and Settings\bruno\Bureau\08112004/http%3a%2f%2fmirrors.sunsite.dk%2fcygwin/release/X11/fontconfig/fontconfig-2.2.2-1.tar.bz2 2004/11/11 10:41:39 Installing file://C:\Documents and Settings\bruno\Bureau\08112004/http%3a%2f%2fmirrors.sunsite.dk%2fcygwin/release/X11/freetype2/freetype2-2.1.5-1.tar.bz2 2004/11/11 10:41:39 Installing file://C:\Documents and Settings\bruno\Bureau\08112004/http%3a%2f%2fmirrors.sunsite.dk%2fcygwin/release/X11/fvwm/fvwm-2.4.7-3.tar.bz2 2004/11/11 10:41:44 Installing file://C:\Documents and Settings\bruno\Bureau\08112004/http%3a%2f%2fmirrors.sunsite.dk%2fcygwin/release/gawk/gawk-3.1.4-3.tar.bz2 2004/11/11 10:41:45 Installing file://C:\Documents and Settings\bruno\Bureau\08112004/http%3a%2f%2fmirrors.sunsite.dk%2fcygwin/release/gdbm/gdbm-1.8.3-7.tar.bz2 2004/11/11
Re: Why Xming.exe?
On Wed, 10 Nov 2004, Christopher Faylor wrote: On Wed, Nov 10, 2004 at 09:02:53PM +0100, Alexander Gottwald wrote: On Wed, 10 Nov 2004, ?yvind Harboe wrote: * There are two many options and questions with CygWin. I imagine that Xming.exe can come with a ~zero questions installer. This is the main reason. There is still some work to do but I plan on having all defaults in Xming depend on the installation dir of Xming. Er, I guess it's time I stepped in, then. Since the name of this mailing list is cygwin-xfree and the site you are using is cygwin.com, I would not expect too much traffic here regarding non-cygwin solutions. This *has* come up before and I think I'm being pretty consistent here. Yes, this is right. If the port evolves and is more usable I'll search a new home for it (most likely freedesktop.org). Currently only Øyvind and I were discussing things and this can be easily done by personal mail. Alexander, if you are going to be devoting a lot of time to this project from now on do we need to find a Cygwin/X maintainer? My #1 priority i clearly Cygwin/X. The mingw port is just a sideeffect of the development for Cygwin/X. bye ago -- [EMAIL PROTECTED] http://www.gotti.org ICQ: 126018723
Keyboard unkown
Hello, My keyboard is not recognize. Here are the reference. (--) winConfigKeyboard - Layout: 100C (100c) (EE) Keyboardlayout Swiss French (100C) is unknown. With the following version. Release: 6.8.1.0-5 _ Express yourself instantly with MSN Messenger! Download today it's FREE! http://messenger.msn.com/
RE: xorg installation fails, 99% complete only
Freezing back under SP2. Here is the ps and cygcheck. cygcheck.out Description: Binary data ps.out Description: Binary data
RE: xorg installation fails, 99% complete only
It should be noted I'm using a Celeron 1.2 GB. Windows Professional. Service pack #2 has security features in it, and this could be one of them (not allowing Xorg server run) -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Bobby McNulty Sent: Thursday, November 11, 2004 5:53 AM To: [EMAIL PROTECTED] Subject: RE: xorg installation fails, 99% complete only Freezing back under SP2. Here is the ps and cygcheck.
Re: xorg installation fails, 99% complete only
Last experiment, I use the ollydbg tool in order to begin to see what is the problem inside setup. It traps also the debug messae of setup and one ofit consists of saying (as it is given by this tool) : failed too open cygfile:///usr/X11R6/share/doc/Cygwin/xorg-x11-fscl-6.8.1.0.README for writing I'm examining the tool in order to know how to associate the src files to the debug tool. As it is a little long to do any experiment it will tae me a little time to gie more infos. I also joined again the cygcheck. I verified it is exactly the same running the setup or not (only the time of it) BPatin Cygwin Configuration Diagnostics Current System Time: Thu Nov 11 13:10:10 2004 Windows XP Home Edition Ver 5.1 Build 2600 Service Pack 2 Path: c:\program files\imagemagick-5.5.7-q16 C:\WINDOWS\system32 C:\WINDOWS C:\WINDOWS\System32\Wbem C:\Program Files\ATI Technologies\ATI Control Panel Output from C:\cygwin\bin\id.exe (nontsec) UID: 400(bruno) GID: 401(mkpasswd) 401(mkpasswd) Output from C:\cygwin\bin\id.exe (ntsec) UID: 400(bruno) GID: 401(mkpasswd) 401(mkpasswd) SysDir: C:\WINDOWS\system32 WinDir: C:\WINDOWS Path = `c:\program files\imagemagick-5.5.7-q16;C:\WINDOWS\system32;C:\WINDOWS;C:\WINDOWS\System32\Wbem;C:\Program Files\ATI Technologies\ATI Control Panel' ALLUSERSPROFILE = `C:\Documents and Settings\All Users' APPDATA = `C:\Documents and Settings\bruno\Application Data' CommonProgramFiles = `C:\Program Files\Fichiers communs' COMPUTERNAME = `BRUNOFONTENAY' ComSpec = `C:\WINDOWS\system32\cmd.exe' FP_NO_HOST_CHECK = `NO' HOMEDRIVE = `C:' HOMEPATH = `\Documents and Settings\bruno' LOGONSERVER = `\\BRUNOFONTENAY' NUMBER_OF_PROCESSORS = `1' OS = `Windows_NT' PATHEXT = `.COM;.EXE;.BAT;.CMD;.VBS;.VBE;.JS;.JSE;.WSF;.WSH' PROCESSOR_ARCHITECTURE = `x86' PROCESSOR_IDENTIFIER = `x86 Family 6 Model 13 Stepping 6, GenuineIntel' PROCESSOR_LEVEL = `6' PROCESSOR_REVISION = `0d06' ProgramFiles = `C:\Program Files' PROMPT = `$P$G' SESSIONNAME = `Console' SystemDrive = `C:' SystemRoot = `C:\WINDOWS' TEMP = `C:\DOCUME~1\bruno\LOCALS~1\Temp' TMP = `C:\DOCUME~1\bruno\LOCALS~1\Temp' USERDOMAIN = `BRUNOFONTENAY' USERNAME = `bruno' USERPROFILE = `C:\Documents and Settings\bruno' windir = `C:\WINDOWS' POSIXLY_CORRECT = `1' HKEY_CURRENT_USER\Software\Cygnus Solutions HKEY_CURRENT_USER\Software\Cygnus Solutions\Cygwin HKEY_CURRENT_USER\Software\Cygnus Solutions\Cygwin\mounts v2 (default) = `/cygdrive' cygdrive flags = 0x0022 HKEY_CURRENT_USER\Software\Cygnus Solutions\Cygwin\mounts v2\/ (default) = `C:\cygwin' flags = 0x0002 HKEY_CURRENT_USER\Software\Cygnus Solutions\Cygwin\mounts v2\/usr/bin (default) = `C:\cygwin/bin' flags = 0x0002 HKEY_CURRENT_USER\Software\Cygnus Solutions\Cygwin\mounts v2\/usr/lib (default) = `C:\cygwin/lib' flags = 0x0002 HKEY_CURRENT_USER\Software\Cygnus Solutions\Cygwin\Program Options HKEY_LOCAL_MACHINE\SOFTWARE\Cygnus Solutions HKEY_LOCAL_MACHINE\SOFTWARE\Cygnus Solutions\Cygwin HKEY_LOCAL_MACHINE\SOFTWARE\Cygnus Solutions\Cygwin\mounts v2 HKEY_LOCAL_MACHINE\SOFTWARE\Cygnus Solutions\Cygwin\Program Options c: hd NTFS 57231Mb 18% CP CS UN PA FC d: cd N/AN/A y: net NTFS 57857Mb 31% CP CSPAbruno z: net NTFS 57857Mb 31% CP CSPACommun C:\cygwin / userbinmode C:\cygwin/bin /usr/bin userbinmode C:\cygwin/lib /usr/lib userbinmode . /cygdrive userbinmode,cygdrive Not Found: awk Not Found: bash Not Found: cat Not Found: cp Not Found: cpp (good!) Not Found: find Not Found: gcc Not Found: gdb Not Found: grep Not Found: ld Not Found: ls Not Found: make Not Found: mv Not Found: rm Not Found: sed Not Found: sh Not Found: tar Cygwin Package Information No setup information found Use -h to see help about each section
Re: [SOLVED - mysteriously...not quite] Re: xorg installation fails, 99% complete only
Carlo Florendo wrote: Cygwin Rocks! Yes. I tried to reproduce the problem on the WinXP SP2 machine in question and I was unsuccessful in reproducing it... Cygwin installed like a charm =) Now, I don't know why such is the case. Hmmm...To tell the truth, I was actually asking someone to follow instructions from me, (i.e. press this button, select Unix, click ok, etc.). However, when I myself tried do a fresh uninstall and fresh install, all worked well. Fresh install... how fresh is fresh? Maybe there remains some registry key after an uninstall? In any case, I'm now attaching cygcheck.out and setup.log (compressed) so the experts could point out what difference it had with my previous posting of cygcheck.out and setup.log, I could see no significant difference there. My suspicion is that the one whom I asked to install cygwin did something else that she didn't tell me(What I saw when she was installing cygwin was the 99% completed on xorg that never stopped. This, I verified with my eyes. On her following my instructions, I didn't see it with my own eyes. She just said she did.) Gerrit -- =^..^=
Re: Why Xming.exe?
On Thu, 11 Nov 2004, David Fraser wrote: Alexander, I'd love to be kept informed of developments so I think it would be great if this could be done on a mailing list ... I'm sure the majority of cygwin-xfree subscribers wouldn't mind a bit of discussion here until a new home is found ... I'll post them to [EMAIL PROTECTED] See http://sources.redhat.com/win32-x11 for subscription details bye ago -- [EMAIL PROTECTED] http://www.gotti.org ICQ: 126018723
Re: xorg installation fails, 99% complete only
Igor Pechtchanski wrote: On Thu, 11 Nov 2004, Carlo Florendo wrote: [snip] In any case, here's setup.log.bz2. (Compressed since it is greater than 100K) The WinXP SP2 machine is accessible to me off-line. Thus, the delay in the sending of these files. [...] Hmm, the fact that it attempts to reinstall packages, coupled with the fact that it's not renaming postinstall scripts properly, makes me wonder whether your Cygwin directory tree is writable by the user that performs the install. This would also be consistent with the Cannot open c:\cygwin\var\log\setup.log for writing message that others reported. IIRC, Bruno said that he is in the Administrator group, should'nt this be sufficient? Gerrit -- =^..^=
Re: xorg installation fails, 99% complete only
bruno patin wrote: Ok there the status As already stated a loop somewhere that consume memory with time when installing /usr/X11R6/share/doc/Cygwin/xorg-x11-fscl-6.8.1.README the joined files: 1 cygcheck.txt (this mail) This is perhaps a problem: TEMP = `C:\DOCUME~1\bruno\LOCALS~1\Temp' TMP = `C:\DOCUME~1\bruno\LOCALS~1\Temp' TEMP = `c:\DOCUME~1\ericson\LOCALS~1\Temp' TMP = `c:\DOCUME~1\ericson\LOCALS~1\Temp' Plese try setting TEMP and TMP in the Windows global environment to some PATH without spaces in it, eg.: TEMP=c:\TEMP TMP=/tmp Gerrit -- =^..^=
Re: xorg installation fails, 99% complete only
following the work I can say there are two messages and that they are issued in the logFile.cc file of setup and thee messages are numbered : first : LOG : 1 Installing file cygfile second LOG : 2 (see my last mail) hope that helps BPatin
Re: xorg installation fails, 99% complete only
Gerrit P. Haase wrote: bruno patin wrote: Ok there the status As already stated a loop somewhere that consume memory with time when installing /usr/X11R6/share/doc/Cygwin/xorg-x11-fscl-6.8.1.README the joined files: 1 cygcheck.txt (this mail) This is perhaps a problem: TEMP = `C:\DOCUME~1\bruno\LOCALS~1\Temp' TMP = `C:\DOCUME~1\bruno\LOCALS~1\Temp' TEMP = `c:\DOCUME~1\ericson\LOCALS~1\Temp' TMP = `c:\DOCUME~1\ericson\LOCALS~1\Temp' Plese try setting TEMP and TMP in the Windows global environment to some PATH without spaces in it, eg.: TEMP=c:\TEMP TMP=/tmp Gerrit will do as asked but I'm inside the debug session of ollydbg on the problem point. As I do not know really well the tool I did not find at this point how to associate the src files so ... I'm using the assmbly code and beside the setup source code. From this point I will do another run n ten to twenty mns using the new def of tmp and temp. BPatin
Mozilla ready compiled, but not operational
Hi, I have some mozilla binaries, the application starts and the usual window is displayed, but it does nothing, if an URL is entered in the address bar and I hit enter, nothing happens, menus are displayed ok, but clicking on an entry does nothin, tooltips are not working or not visible, there is only a minimal balloon displayed. The GTKMozEmbed test application is working fine and I can browse websites with it. What may be the problem with the mozilla application? Who is interested to help debugging this? Gerrit -- =^..^=
Re: xorg installation fails, 99% complete only
Gerrit P. Haase wrote: bruno patin wrote: Ok there the status As already stated a loop somewhere that consume memory with time when installing /usr/X11R6/share/doc/Cygwin/xorg-x11-fscl-6.8.1.README the joined files: 1 cygcheck.txt (this mail) This is perhaps a problem: TEMP = `C:\DOCUME~1\bruno\LOCALS~1\Temp' TMP = `C:\DOCUME~1\bruno\LOCALS~1\Temp' TEMP = `c:\DOCUME~1\ericson\LOCALS~1\Temp' TMP = `c:\DOCUME~1\ericson\LOCALS~1\Temp' Plese try setting TEMP and TMP in the Windows global environment to some PATH without spaces in it, eg.: TEMP=c:\TEMP TMP=/tmp Gerrit Gerrit, I was thinking of something. I would be surprised that my test with different TMP variables could work because they work for all the packages before. Could it be more something like trying to open more than the permitted number of files ? when trying at this step to open one more handler, setup would fail go into a continuable exception not releasing the memory associated with this corrupted handler and trying again consuming memory ? Is it ridiculous ? Bruno
Re: xorg installation fails, 99% complete only
Gerrit P. Haase wrote: bruno patin wrote: Ok there the status As already stated a loop somewhere that consume memory with time when installing /usr/X11R6/share/doc/Cygwin/xorg-x11-fscl-6.8.1.README the joined files: 1 cygcheck.txt (this mail) This is perhaps a problem: TEMP = `C:\DOCUME~1\bruno\LOCALS~1\Temp' TMP = `C:\DOCUME~1\bruno\LOCALS~1\Temp' TEMP = `c:\DOCUME~1\ericson\LOCALS~1\Temp' TMP = `c:\DOCUME~1\ericson\LOCALS~1\Temp' Plese try setting TEMP and TMP in the Windows global environment to some PATH without spaces in it, eg.: TEMP=c:\TEMP TMP=/tmp Gerrit ok I've done it, it did not work. I'll try something else. I will remove one and only one unitary package in order to see if I block on the same file. If not I will assume my hypothesis based on the number of file opened could be a good one. Bruno
Re: Keyboard unkown
On Thu, 11 Nov 2004, Christian Kratzer wrote: Hello, My keyboard is not recognize. Here are the reference. (--) winConfigKeyboard - Layout: 100C (100c) (EE) Keyboardlayout Swiss French (100C) is unknown. I've added it. the next release will include it. bye ago -- [EMAIL PROTECTED] http://www.gotti.org ICQ: 126018723
Re: Mozilla ready compiled, but not operational
Gerrit P. Haase wrote: Hi, I have some mozilla binaries, the application starts and the usual window is displayed, but it does nothing, if an URL is entered in the address bar and I hit enter, nothing happens, menus are displayed ok, but clicking on an entry does nothin, tooltips are not working or not visible, there is only a minimal balloon displayed. The GTKMozEmbed test application is working fine and I can browse websites with it. What may be the problem with the mozilla application? Who is interested to help debugging this? Gerrit Why not but beware, I do not know the internals of mozilla that well (notat all in fact, only reading what something like linux mag could write). What do you want me to do ? Don't forget I have no operational cygwin working at this step as I'm working on the install problem and each test is a long one. Bruno
Re: xorg installation fails, 99% complete only
Gerrit P. Haase wrote: bruno patin wrote: Ok there the status As already stated a loop somewhere that consume memory with time when installing /usr/X11R6/share/doc/Cygwin/xorg-x11-fscl-6.8.1.README the joined files: 1 cygcheck.txt (this mail) This is perhaps a problem: TEMP = `C:\DOCUME~1\bruno\LOCALS~1\Temp' TMP = `C:\DOCUME~1\bruno\LOCALS~1\Temp' TEMP = `c:\DOCUME~1\ericson\LOCALS~1\Temp' TMP = `c:\DOCUME~1\ericson\LOCALS~1\Temp' Plese try setting TEMP and TMP in the Windows global environment to some PATH without spaces in it, eg.: TEMP=c:\TEMP TMP=/tmp Gerrit I was right, deleting only one file of the list of file to install meant the ability to write one more file on the disk (in my case I'm now blocked on the xorg-x11-fsrv-6.8.1.0.README. I do nt hink that it is a problem specific to xorg. xorg unhappily was in the last packages to be installed. Ami I right ? It would mean that this thread is not well placed in the xfree list. Bruno
Re: xorg installation fails, 99% complete only
Gerrit P. Haase wrote: bruno patin wrote: Ok there the status As already stated a loop somewhere that consume memory with time when installing /usr/X11R6/share/doc/Cygwin/xorg-x11-fscl-6.8.1.README the joined files: 1 cygcheck.txt (this mail) This is perhaps a problem: TEMP = `C:\DOCUME~1\bruno\LOCALS~1\Temp' TMP = `C:\DOCUME~1\bruno\LOCALS~1\Temp' TEMP = `c:\DOCUME~1\ericson\LOCALS~1\Temp' TMP = `c:\DOCUME~1\ericson\LOCALS~1\Temp' Plese try setting TEMP and TMP in the Windows global environment to some PATH without spaces in it, eg.: TEMP=c:\TEMP TMP=/tmp Gerrit something I forgot, the number of handler increases at the same speed as the memory. I suspect that this resource is not released when exiting setup because when doing my experiments my computer is being slower and slower. Bruno
Re: Why Xming.exe?
On Thu, Nov 11, 2004 at 08:49:48AM +0100, ?yvind Harboe wrote: First of all: I love CygWin! If I didn't, I would not passionatly dive into this discussion. :-) Here are a couple of reasons why I think a discussion of Xming.exe can bring something to CygWin. Basically, I believe that the Xming.exe would never have seen the light of day if if all was well with CygWin. It is a cry for help. Not a rationale for discussing MinGW projects on a Cygwin list. The MinGW project came about due to the Cygwin project. The MinGW project has a separate mailing list. And, they created the mailing list themselves without offering specious arguments about why they should be able to keep discussing things in the Cygwin list. Also, I believe the next useful solution to the set of problems that CygWin is facing is non-local. I think it is in the interest of CygWin to allow discussions and hacks to wander somewhat freely. This is just more of I think it would be nice. Some issues that CygWin could address to make Xming.exe moot: - Side-by-side install. There is a need for e.g. SSH + and an X server in a single package that is easily deployable. This is not a goal of Cygwin any more than it is a goal of Red Hat Linux to be easily deployable with SuSE. This has come up repeatedly in the main list. Maybe you should do some research before you start suggesting things that have already been beaten to death. It is not that hard to install a Cygwin/X/SSH package into an existing cygwin installation. It just requires care. The problem is that no one wants to take the time to do things carefully. Everyone wants to use a simple tar ball or a simple copy to the main disk and have things work. That is not going to work but having an installer which first checked the cygwin version and the cygwin layout and then installed things into that would work. The installer could optionally install the cygwin DLL if required. The problem is that it seems that every single person who whines about needing two versions of cygwin on the same computer lacks either the ability or the concentration to be able to figure out how to do this. Given that no one has stepped forward to provide tools for doing this, I am left to conclude that this basically boils down to I know how to use a baseball bat. I'm trying to trim my hedges. The baseball bat is not working right. Please, Mr. Baseball Bat Manufacturer take fix your bat! However, if it was an issue, then the solution would not be to develop a whole bunch of packages using MinGW. The solution would be to modify *cygwin*. E.g. its more than once that CygWin updates have broken http://sources.redhat.com/ecos tools, without bringing any new features or improvements for those tools. 1) This has absolutely nothing to do with X. 2) If Cygwin is an important platform for eCos, then the solution to this problem is for eCos people to be involved in the cygwin mailing list. - Side-by-side means CygWin next to CygWin, and CygWin next to any Windows program. Cygwin already exists fine next to any Windows program. Cygwin next to Cygwin has nothing to do with a MinGW version of X. - It is currently impossible for the casual programmer to debug CygWin X using GDB. See http://cygwin.com/ml/cygwin-xfree/2004-10/msg00159.html This has nothing to do with discussing a MinGW version of X. If something needs to be fixed in cygwin or in gdb, you fix cygwin or gdb. You don't immediately jump to a new project. - CygWin is a pain to install. It takes long to download(many files more than size with a broad connection), takes long to install and many things can and do go wrong. So, should we allow discussion of Xceed here, too? I assume it's easy to install. In any event - CygWin messes with Windows environment variables, breaking other programs. Cygwin does not mess with windows environment variables. - CygWin is broken by other programs using the Windows environment variables. Examples? - CygWin is always in motion. There is no CygWin distribution version like the Linux distributions operate with. There is no way to install a specific old boring CygWin to run some tools that never change. Wow, your logic really escapes me. You haven't provided one fact-based rationale for why it would be advantageous to discuss a non-cygwin program in a cygwin mailing list. All of your arguments are the grasping at straws variety. Your arguments boil down to: 1) Xming would not have come about if not for Cygwin. 1) You don't really understand how cygwin works. (cygwin messes with Windows environment variables) 2) Cygwin is hard to install. 3) You can't put as many versions of Cygwin on the computer as you want to (without understanding what you are doing). 4) When there is a Cygwin update it can break programs. 5) You think it is a good idea. Ranting about Cygwin problems is not a rationale for allowing discussion of non-cygwin solutions. If Cygwin has problems you FIX THE
Re: xorg installation fails, 99% complete only
Christopher Faylor wrote: On Thu, Nov 11, 2004 at 01:15:16PM +0100, bruno patin wrote: Last experiment, I use the ollydbg tool in order to begin to see what is the problem inside setup. It traps also the debug messae of setup and one ofit consists of saying (as it is given by this tool) : failed too open cygfile:///usr/X11R6/share/doc/Cygwin/xorg-x11-fscl-6.8.1.0.README for writing Does the path c:/cygwin/usr/X11R6/share/doc/Cygwin exist at this point? cgf yes, I followed it rather closely and the creation of the preceding n.README files. Wht do you think of my follwing mails ? do you think it could be the problem ? Bruno
Re: Why Xming.exe?
On Thu, Nov 11, 2004 at 10:30:48AM -0500, Christopher Faylor wrote: In any event Sorry. This message obviously slipped out before I was done with it. I blame my spastic index finger. It likes to press the y key. In any event, what I was going to say above was that In any event it is certainly possible to craft a Cygwin/X installation which only installs a minimal number of packages. No one has stepped forward to do this. Many of the observations made here are fixable just by having someone step forward to do the work in Cygwin. If no one is willing to do that, that's fine. You can do related work elsewhere but touting MinGW programs as a solution to Cygwin problems doesn't work. cgf
Re: Why Xming.exe?
Christopher Faylor wrote: On Thu, Nov 11, 2004 at 10:30:48AM -0500, Christopher Faylor wrote: In any event Sorry. This message obviously slipped out before I was done with it. I blame my spastic index finger. It likes to press the y key. In any event, what I was going to say above was that In any event it is certainly possible to craft a Cygwin/X installation which only installs a minimal number of packages. No one has stepped forward to do this. Many of the observations made here are fixable just by having someone step forward to do the work in Cygwin. If no one is willing to do that, that's fine. You can do related work elsewhere but touting MinGW programs as a solution to Cygwin problems doesn't work. OK thats fine, just a minor point: there is one Cygwin problem that this solves, which is that people keep on requesting this kind of thing on the cygwin-xfree mailing list :-) Also its constructed from the same source code as cygwin/X so its arguably the same thing. But anyway we'll happy carry on discussion on a different list David
Re: Why Xming.exe?
On Thu, Nov 11, 2004 at 06:04:45PM +0200, David Fraser wrote: Christopher Faylor wrote: On Thu, Nov 11, 2004 at 10:30:48AM -0500, Christopher Faylor wrote: In any event Sorry. This message obviously slipped out before I was done with it. I blame my spastic index finger. It likes to press the y key. In any event, what I was going to say above was that In any event it is certainly possible to craft a Cygwin/X installation which only installs a minimal number of packages. No one has stepped forward to do this. Many of the observations made here are fixable just by having someone step forward to do the work in Cygwin. If no one is willing to do that, that's fine. You can do related work elsewhere but touting MinGW programs as a solution to Cygwin problems doesn't work. OK thats fine, just a minor point: there is one Cygwin problem that this solves, which is that people keep on requesting this kind of thing on the cygwin-xfree mailing list :-) Yes, and people keep requesting that Cygwin shouldn't be GPLed because it is inconvenient for them. People ask for su to work correctly. People occasionally want to discuss Xceed here, too. This isn't a ill-informed minority gets to decide mailing list. Also its constructed from the same source code as cygwin/X so its arguably the same thing. Is anyone here *at all* familiar with MinGW? Apparently few of you are or you wouldn't be making arguments like this. Much of MinGW is based on the same source code as what Cygwin uses. Using this logic, since Cygwin/X is based on the same source code that runs on many different platforms, apparently we should just shut this mailing list down and move everyone over to the main Xorg mailing lists. But anyway we'll happy carry on discussion on a different list But not before trying to get off a few more shots, eh? cgf
Re: Why Xming.exe?
Christopher Faylor wrote: On Thu, Nov 11, 2004 at 06:04:45PM +0200, David Fraser wrote: Christopher Faylor wrote: On Thu, Nov 11, 2004 at 10:30:48AM -0500, Christopher Faylor wrote: In any event Sorry. This message obviously slipped out before I was done with it. I blame my spastic index finger. It likes to press the y key. In any event, what I was going to say above was that In any event it is certainly possible to craft a Cygwin/X installation which only installs a minimal number of packages. No one has stepped forward to do this. Many of the observations made here are fixable just by having someone step forward to do the work in Cygwin. If no one is willing to do that, that's fine. You can do related work elsewhere but touting MinGW programs as a solution to Cygwin problems doesn't work. OK thats fine, just a minor point: there is one Cygwin problem that this solves, which is that people keep on requesting this kind of thing on the cygwin-xfree mailing list :-) Yes, and people keep requesting that Cygwin shouldn't be GPLed because it is inconvenient for them. People ask for su to work correctly. People occasionally want to discuss Xceed here, too. This isn't a ill-informed minority gets to decide mailing list. Fine, but at least we can have arguments about it :-) Also its constructed from the same source code as cygwin/X so its arguably the same thing. Is anyone here *at all* familiar with MinGW? Apparently few of you are or you wouldn't be making arguments like this. Much of MinGW is based on the same source code as what Cygwin uses. I use MinGW the whole time, yes, I know what it is. Using this logic, since Cygwin/X is based on the same source code that runs on many different platforms, apparently we should just shut this mailing list down and move everyone over to the main Xorg mailing lists. OK fine :-) But anyway we'll happy carry on discussion on a different list But not before trying to get off a few more shots, eh? Yes, you see I've managed to generate an extra 3 useless emails! And without your reply I wouldn't have managed to generate this one too ... You could actually compile Cygwin/X using cygwin GCC with the -mnocygwin option and that apparently would be off topic too. So I won't :-) Don't mean to hassle you so lets leave it at that David
Re: Can't get startwin.bat to work
Have you tried to open up a cygwin shell, and then starting up an xterm, to see if it might just be there, awaiting connections? If you open in multiwindow mode (native mode by default), then I wouldn't expect anything to happen until your X Server actually has a client window to display. open up a cygwin bash shell, make sure your DISPLAY is set, or give the -display option to some xclient (xterm, xclock, etc), and see if it pops up on your desktop. On Wed, 10 Nov 2004 04:49:47 -0800 (PST), Mike McCollister wrote: I have been using cygwin for years and the X client under cygwin for about a year. In order to get X to work, I have to make sure that I am running bash and then run startxwin.sh to get it to work. Well, I've been trying to get it to work with just startxwin.bat but I am not successful. It seems to have a problem with the run commands. Here are the run commands: run XWin -clipboard -silent-dup-error run xterm -sl 1000 -sb -rightbar -ms red -fg yellow -bg black -e /usr/bin/bash -l I do have cygwin installed in c:\progra~1\cygwin and I have modified startxwin.bat so that SET CYGWIN_ROOT=c:\progra~1\cygwin The last time I updated anything in cygwin as last week (first week of November). Does anyone know how to get the startxwin.bat to work? Just a little more background. When I mean that it does not work, I mean that the X shows up as an icon next to the clock but nothing else happens. If I right click on the X, nothing happens. I have to use the Task Manger to kill XWin.exe. Thanks, Mike __ Do you Yahoo!? Check out the new Yahoo! Front Page. www.yahoo.com
Re: xorg installation fails, 99% complete only
On Wed, Nov 10, 2004 at 09:22:16PM +0100, bruno patin wrote: Good evening again, install in two phases worked well. I can deinstall again and verify if there is the problem again ? could it be useful ? joined : an analysis of my system the cygcheck result hope that helps I'm noticing a bunch of XFree86 modules in this cygcheck, too. Where are they coming from? I wouldn't have expected many XFree86 modules in a clean install. cgf
Re: xorg installation fails, 99% complete only
Christopher Faylor wrote: On Wed, Nov 10, 2004 at 09:22:16PM +0100, bruno patin wrote: Good evening again, install in two phases worked well. I can deinstall again and verify if there is the problem again ? could it be useful ? joined : an analysis of my system the cygcheck result hope that helps I'm noticing a bunch of XFree86 modules in this cygcheck, too. Where are they coming from? I wouldn't have expected many XFree86 modules in a clean install. cgf I don't know why but it is probably of no importance for the problem at hand. Could you say me if my mails joined are a bad analysis of this problem ? BPatin ---BeginMessage--- Gerrit P. Haase wrote: bruno patin wrote: Ok there the status As already stated a loop somewhere that consume memory with time when installing /usr/X11R6/share/doc/Cygwin/xorg-x11-fscl-6.8.1.README the joined files: 1 cygcheck.txt (this mail) This is perhaps a problem: TEMP = `C:\DOCUME~1\bruno\LOCALS~1\Temp' TMP = `C:\DOCUME~1\bruno\LOCALS~1\Temp' TEMP = `c:\DOCUME~1\ericson\LOCALS~1\Temp' TMP = `c:\DOCUME~1\ericson\LOCALS~1\Temp' Plese try setting TEMP and TMP in the Windows global environment to some PATH without spaces in it, eg.: TEMP=c:\TEMP TMP=/tmp Gerrit Gerrit, I was thinking of something. I would be surprised that my test with different TMP variables could work because they work for all the packages before. Could it be more something like trying to open more than the permitted number of files ? when trying at this step to open one more handler, setup would fail go into a continuable exception not releasing the memory associated with this corrupted handler and trying again consuming memory ? Is it ridiculous ? Bruno ---End Message--- ---BeginMessage--- Gerrit P. Haase wrote: bruno patin wrote: Ok there the status As already stated a loop somewhere that consume memory with time when installing /usr/X11R6/share/doc/Cygwin/xorg-x11-fscl-6.8.1.README the joined files: 1 cygcheck.txt (this mail) This is perhaps a problem: TEMP = `C:\DOCUME~1\bruno\LOCALS~1\Temp' TMP = `C:\DOCUME~1\bruno\LOCALS~1\Temp' TEMP = `c:\DOCUME~1\ericson\LOCALS~1\Temp' TMP = `c:\DOCUME~1\ericson\LOCALS~1\Temp' Plese try setting TEMP and TMP in the Windows global environment to some PATH without spaces in it, eg.: TEMP=c:\TEMP TMP=/tmp Gerrit ok I've done it, it did not work. I'll try something else. I will remove one and only one unitary package in order to see if I block on the same file. If not I will assume my hypothesis based on the number of file opened could be a good one. Bruno ---End Message--- ---BeginMessage--- Gerrit P. Haase wrote: bruno patin wrote: Ok there the status As already stated a loop somewhere that consume memory with time when installing /usr/X11R6/share/doc/Cygwin/xorg-x11-fscl-6.8.1.README the joined files: 1 cygcheck.txt (this mail) This is perhaps a problem: TEMP = `C:\DOCUME~1\bruno\LOCALS~1\Temp' TMP = `C:\DOCUME~1\bruno\LOCALS~1\Temp' TEMP = `c:\DOCUME~1\ericson\LOCALS~1\Temp' TMP = `c:\DOCUME~1\ericson\LOCALS~1\Temp' Plese try setting TEMP and TMP in the Windows global environment to some PATH without spaces in it, eg.: TEMP=c:\TEMP TMP=/tmp Gerrit I was right, deleting only one file of the list of file to install meant the ability to write one more file on the disk (in my case I'm now blocked on the xorg-x11-fsrv-6.8.1.0.README. I do nt hink that it is a problem specific to xorg. xorg unhappily was in the last packages to be installed. Ami I right ? It would mean that this thread is not well placed in the xfree list. Bruno ---End Message---
Re: xorg installation fails, 99% complete only
bruno patin wrote: something I forgot, the number of handler increases at the same speed as the memory. I suspect that this resource is not released when exiting setup because when doing my experiments my computer is being slower and slower. What is the actual limit for NT? I never was aware that there is a limit at all. Gerrit -- =^..^=
Re: xorg installation fails, 99% complete only
Gerrit P. Haase wrote: bruno patin wrote: something I forgot, the number of handler increases at the same speed as the memory. I suspect that this resource is not released when exiting setup because when doing my experiments my computer is being slower and slower. What is the actual limit for NT? I never was aware that there is a limit at all. Gerrit In fact I'm a unix guy and I don't really know of what windows is doing. My understanding is a 'black box' one seeing what happens with the few win tools I have and the setup code. I can perhaps search a little but I think the cygwin coder are far more better than me. That's the reason why I'd like an advice fromm christopher Faylor that knows certainly what could happen on nt. Bruno
Re: xorg installation fails, 99% complete only
On Thu, Nov 11, 2004 at 06:42:44PM +0100, bruno patin wrote: Gerrit P. Haase wrote: bruno patin wrote: something I forgot, the number of handler increases at the same speed as the memory. I suspect that this resource is not released when exiting setup because when doing my experiments my computer is being slower and slower. What is the actual limit for NT? I never was aware that there is a limit at all. Gerrit In fact I'm a unix guy and I don't really know of what windows is doing. My understanding is a 'black box' one seeing what happens with the few win tools I have and the setup code. I can perhaps search a little but I think the cygwin coder are far more better than me. That's the reason why I'd like an advice fromm christopher Faylor that knows certainly what could happen on nt. I'm thinking that there is something odd about setup's handing of files in directories which have not been explicitly created in the tar archive. That's why I was asking if c:/cygwin/usr/X11R6/share/doc/Cygwin existed. It didn't seem like the tar archive was creating it specifically. I don't know why this would affect some people but not others, though. When I install cygwin from scratch in a base install + all X packages selected, I don't see this. So maybe this is a red herring. Anyway, to the end of confirming my theory, I have (I HOPE) regenerated all of the X11 files in the distribution and made sure that directory creation precedes filename creation in every case. I haven't bumped the version numbers, though, since this is purely a cosmetic change. This change will take a while to hit mirrors since many files were affected but if you check that a mirror contains files with today's date, then it is worth trying a full install to see if your problem mysteriously goes away. So, Bruno, could you periodically check: ftp://mirrors.sunsite.dk/pub/cygwin/release/X11/xorg-x11-fscl/ wait until the file xorg-x11-fscl-6.8.1.0-2.tar.bz2 shows today's date, and then try your test case again to see if it hangs? I think that setup.exe should do the right thing and download the changed files but, if it doesn't, you'll have to delete them from your disk. As far as handles are concerned, I don't think that there is any upper limit on the number that can be allocated. The only limitation is 32 bits and the amount of available memory. It seems like people are hitting this limit if they allow setup to run long enough. cgf
Re: xorg installation fails, 99% complete only
So, Bruno, could you periodically check: ftp://mirrors.sunsite.dk/pub/cygwin/release/X11/xorg-x11-fscl/ wait until the file xorg-x11-fscl-6.8.1.0-2.tar.bz2 shows today's date, and then try your test case again to see if it hangs? I think that setup.exe should do the right thing and download the changed files but, if it doesn't, you'll have to delete them from your disk. As far as handles are concerned, I don't think that there is any upper limit on the number that can be allocated. The only limitation is 32 bits and the amount of available memory. It seems like people are hitting this limit if they allow setup to run long enough. cgf I'll do it but remember that I realize a test where nnn-fscl-nnn was not anymore concerned only by removing one file of the setup llist. Bruno
Re: xorg installation fails, 99% complete only
On Thu, Nov 11, 2004 at 07:27:32PM +0100, bruno patin wrote: So, Bruno, could you periodically check: ftp://mirrors.sunsite.dk/pub/cygwin/release/X11/xorg-x11-fscl/ wait until the file xorg-x11-fscl-6.8.1.0-2.tar.bz2 shows today's date, and then try your test case again to see if it hangs? I think that setup.exe should do the right thing and download the changed files but, if it doesn't, you'll have to delete them from your disk. As far as handles are concerned, I don't think that there is any upper limit on the number that can be allocated. The only limitation is 32 bits and the amount of available memory. It seems like people are hitting this limit if they allow setup to run long enough. I'll do it but remember that I realize a test where nnn-fscl-nnn was not anymore concerned only by removing one file of the setup llist. On Thu, Nov 11, 2004 at 01:20:57PM -0500, Christopher Faylor wrote: I have (I HOPE) regenerated all of the X11 files in the distribution ^^ I was merely providing a reference point. cgf
Re: xorg installation fails, 99% complete only
On Thu, 11 Nov 2004, Gerrit P. Haase wrote: Igor Pechtchanski wrote: On Thu, 11 Nov 2004, Carlo Florendo wrote: [snip] In any case, here's setup.log.bz2. (Compressed since it is greater than 100K) The WinXP SP2 machine is accessible to me off-line. Thus, the delay in the sending of these files. [...] Hmm, the fact that it attempts to reinstall packages, coupled with the fact that it's not renaming postinstall scripts properly, makes me wonder whether your Cygwin directory tree is writable by the user that performs the install. This would also be consistent with the Cannot open c:\cygwin\var\log\setup.log for writing message that others reported. IIRC, Bruno said that he is in the Administrator group, should'nt this be sufficient? Nope. It's possible to create files that aren't readable/writeable by the Administrators group. Being in Administrators allows you to take ownership of the files or change the mode on them, but AFAIK that's not done in the course of a normal write. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_[EMAIL PROTECTED] |,4- ) )-,_. ,\ ( `'-' Igor Pechtchanski, Ph.D. '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! The Sun will pass between the Earth and the Moon tonight for a total Lunar eclipse... -- WCBS Radio Newsbrief, Oct 27 2004, 12:01 pm EDT
Re: xorg installation fails, 99% complete only
Christopher, I'm not understanding really well what to do (you know my proficiency in english is not so perfect as to understand really everything). I understand that you think that to open for writing a name where the directory do no exist could provoke problems. It would explain the non blocking exception (am I right ?) that occurs and loops indefinitely (is it what you think ?). What of the fact that the error occurs only after being able to write a given number of file in this directory (around ten files where created) ? what of the fact that when you do not install only one file before (like cron) you are able to write one more file in this directory ? I'm not sufficiently proficient in the cygwin coding to answer these questions. So my work is to begin a new complete install with the new packages. Do I have to reload all of them ? only xorg related packages ? What I intend to do is begin again the first test described under with the reloaded x11 files. Is it what you want ? --- To summarize what I did when testing and the problems I saw (I do that not only for repeating but also to have all the datas on only one mail) 1/ complete test I have a local version of all the cygwin distro (got at mirrors.sunsite.dk) I begin by deleting my cygwin root on the disk but not the entries of the registry I launch setup selecting everything Arriving at the file: /usr/X11R6/share/doc/Cygwin/xorg-x11-fscl-6.8.1.0.README setup enters a loop delivering the log message one after the other indefintely (view through ollydbg) LOG : 1 Installing file cygfile:///usr/X11R6/share/doc/Cygwin/xorg-x11-fscl-6.8.1.0.README for writing LOG : 2 failed too open cygfile:///usr/X11R6/share/doc/Cygwin/xorg-x11-fscl-6.8.1.0.README for writing . The second message is issued through an exception You can see two things increasing with the same speed, the memory consumed and the handlers. when cancelled, the message issued by setup is cannot open log file c:\cygwin/var/log/setup.log for writing another box opens but disappear too fast for me to give you the message With this test, no setup.log is created at all This test has been conducted with these different changes: I changed the rights of the /var directory with chmod a+w with no effect (because of the absence of setup) I changed the TEMP and TMP variables values (erasing blanks of the names) 2/ successive setup I conducted a first setup without X11 and a second with only X11 .It worked fine with setup.log created 3/ complete test with one file (cron) not installed The behaviour is exactly the same as in first test with only the file on which it stopped as a difference. That is t loops on the next file to write: xorg-x11-fsrv-6.8.1.0.README --- Well, perhaps I forgot something but no one is perfect :-) Bruno
Re: xorg installation fails, 99% complete only
On Thu, 11 Nov 2004, bruno patin wrote: Christopher, I'm not understanding really well what to do (you know my proficiency in english is not so perfect as to understand really everything). I understand that you think that to open for writing a name where the directory do no exist could provoke problems. It would explain the non blocking exception (am I right ?) that occurs and loops indefinitely (is it what you think ?). What of the fact that the error occurs only after being able to write a given number of file in this directory (around ten files where created) ? what of the fact that when you do not install only one file before (like cron) you are able to write one more file in this directory ? I'm not sufficiently proficient in the cygwin coding to answer these questions. So my work is to begin a new complete install with the new packages. Do I have to reload all of them ? only xorg related packages ? What I intend to do is begin again the first test described under with the reloaded x11 files. Is it what you want ? --- To summarize what I did when testing and the problems I saw (I do that not only for repeating but also to have all the datas on only one mail) 1/ complete test I have a local version of all the cygwin distro (got at mirrors.sunsite.dk) I begin by deleting my cygwin root on the disk but not the entries of the registry I launch setup selecting everything Arriving at the file: /usr/X11R6/share/doc/Cygwin/xorg-x11-fscl-6.8.1.0.README setup enters a loop delivering the log message one after the other indefintely (view through ollydbg) LOG : 1 Installing file cygfile:///usr/X11R6/share/doc/Cygwin/xorg-x11-fscl-6.8.1.0.README for writing LOG : 2 failed too open cygfile:///usr/X11R6/share/doc/Cygwin/xorg-x11-fscl-6.8.1.0.README for writing . The second message is issued through an exception You can see two things increasing with the same speed, the memory consumed and the handlers. when cancelled, the message issued by setup is cannot open log file c:\cygwin/var/log/setup.log for writing another box opens but disappear too fast for me to give you the message With this test, no setup.log is created at all This test has been conducted with these different changes: I changed the rights of the /var directory with chmod a+w with no effect (because of the absence of setup) I changed the TEMP and TMP variables values (erasing blanks of the names) 2/ successive setup I conducted a first setup without X11 and a second with only X11 .It worked fine with setup.log created 3/ complete test with one file (cron) not installed The behaviour is exactly the same as in first test with only the file on which it stopped as a difference. That is t loops on the next file to write: xorg-x11-fsrv-6.8.1.0.README --- Well, perhaps I forgot something but no one is perfect :-) One question that should have been asked long before: is your disk full? Do you perhaps have quotas enabled? Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_[EMAIL PROTECTED] |,4- ) )-,_. ,\ ( `'-' Igor Pechtchanski, Ph.D. '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! The Sun will pass between the Earth and the Moon tonight for a total Lunar eclipse... -- WCBS Radio Newsbrief, Oct 27 2004, 12:01 pm EDT
Re: xorg installation fails, 99% complete only
On Thu, 11 Nov 2004, Christopher Faylor wrote: On Wed, Nov 10, 2004 at 09:22:16PM +0100, bruno patin wrote: Good evening again, install in two phases worked well. I can deinstall again and verify if there is the problem again ? could it be useful ? joined : an analysis of my system the cygcheck result hope that helps I'm noticing a bunch of XFree86 modules in this cygcheck, too. Where are they coming from? I wouldn't have expected many XFree86 modules in a clean install. Installing everything should also grab all of the empty XFree86 upgrade helpers. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_[EMAIL PROTECTED] |,4- ) )-,_. ,\ ( `'-' Igor Pechtchanski, Ph.D. '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! The Sun will pass between the Earth and the Moon tonight for a total Lunar eclipse... -- WCBS Radio Newsbrief, Oct 27 2004, 12:01 pm EDT
Re: xorg installation fails, 99% complete only
One question that should have been asked long before: is your disk full? Do you perhaps have quotas enabled? Igor For this one, it is easy, my box is a completely new one. I have plenty of room but for any quota mechanism (is there one on xp ?) Bruno
Re: xorg installation fails, 99% complete only
Installing everything should also grab all of the empty XFree86 upgrade helpers. Igor I'll have to read closely the doc on setup as I do not understand what you mean by empty XFreee86 upgrade. Bruno
Re: xorg installation fails, 99% complete only
On Thu, Nov 11, 2004 at 03:39:23PM -0500, Igor Pechtchanski wrote: On Thu, 11 Nov 2004, Christopher Faylor wrote: On Wed, Nov 10, 2004 at 09:22:16PM +0100, bruno patin wrote: Good evening again, install in two phases worked well. I can deinstall again and verify if there is the problem again ? could it be useful ? joined : an analysis of my system the cygcheck result hope that helps I'm noticing a bunch of XFree86 modules in this cygcheck, too. Where are they coming from? I wouldn't have expected many XFree86 modules in a clean install. Installing everything should also grab all of the empty XFree86 upgrade helpers. Ok. I guess I've lost track of who's doing what. I didn't realize that this was an absolutely everything install. cgf
Re: xorg installation fails, 99% complete only
One question that should have been asked long before: is your disk full? Do you perhaps have quotas enabled? Igor Igor, Ok, I verified, there are quotas on xp but on MY box they are deactivated. Bruno
RE: xorg installation fails, 99% complete only
hope that helps I'm noticing a bunch of XFree86 modules in this cygcheck, too. Where are they coming from? I wouldn't have expected many XFree86 modules in a clean install. Installing everything should also grab all of the empty XFree86 upgrade elpers. Igor, it was grabbing the wrong packages from Xfree. It was grabbing full packages, xfree-bin comes to mind, as that was the only one it picked up on. Bobby BTW it was straight out off the net, full, all users, unix. I'll try again later. This time without the xorg.
RE: xorg installation fails, 99% complete only
ftp://mirrors.sunsite.dk/pub/cygwin/release/X11/xorg-x11-fscl/ WIW, http://mirrors.rcn.com/pub/sourceware/cygwin/release/X11/xorg-x11-fscl/ hows that this file is now updated on the mirrors.rcn.net mirror. Bobby, do you want to try that one? cgf Trying it now. Let me see here.
Re: Problem with setup
Christopher Faylor wrote: On Thu, Nov 11, 2004 at 05:26:34PM -0600, Bobby McNulty wrote: Guys, theres a major problem with setup. According to the error I just got, setup is not closing any of its files after writing them. I finally got past xorg, and got up to update_info. I'll rerun setup to get a cygcheck. So, does this mean that the changes I made to the xorg distribution worked, then? If so, why are you reporting this here rather than in the cygwin-xfree mailing list where this has been discussed at great length? cgf -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/ Yes, it worked. Now I'm getting a report that setup is not closing the files. It can't get past the update_info now. Something has happened to setup. Either that, or Cygwin is bigger now than when the current version came out. I lost the URL to the test version of setup. I know they have been working on. I see it in the CVS logs every day.
Re: xorg installation fails, 99% complete only
On Thu, Nov 11, 2004 at 09:32:51PM +0100, bruno patin wrote: So my work is to begin a new complete install with the new packages. Do I have to reload all of them ? only xorg related packages ? You have to do whatever it is that you do to cause the problem. The mirror has been updated (I created the tar files in a way that made them unreadable by setup earlier but that's fixed now) so you should be able to install as normal. cgf
Re: xorg installation fails, 99% complete only
Christopher Faylor wrote: On Thu, Nov 11, 2004 at 09:32:51PM +0100, bruno patin wrote: So my work is to begin a new complete install with the new packages. Do I have to reload all of them ? only xorg related packages ? You have to do whatever it is that you do to cause the problem. The mirror has been updated (I created the tar files in a way that made them unreadable by setup earlier but that's fixed now) so you should be able to install as normal. cgf OK, I reload all and do a new install (it will take a little time as I do not have a really fast connection). I'll let you know when it is finished. Cheers Bruno
[Patch] Fixing the PROCESS_DUP_HANDLE security hole.
Now that 1.5.12 is out, here is a patch to fix the
PROCESS_DUP_HANDLE security hole. It uses a new approach
to reparenting: the parent duplicates the exec'ed process
handle when signaled by the child.
It also handles correctly the case of a quick re-exec
(2 simultaneous reparenting), which is a weak point of
the current version.
Pierre
P.S.: I have no news about the recent patch to /bin/kill -f
2004-11-12 Pierre Humblet [EMAIL PROTECTED]
* pinfo.h (_pinfo::isreparenting): New element.
(_pinfo::ppid_sendsig): Ditto.
(_pinfo::exit): Suppress second argument.
* child_info.h: Update CURR_CHILD_INFO_MAGIC.
(child_info::pppid_sendsig): New element.
* sigproc.h: Add __SIGREPARENT.
(enum procstuff): Add PROC_REPARENT.
* pinfo.cc (_pinfo::exit): Suppress second argument.
If required, send reparenting signal and wait.
* spawn.cc (spawn_guts): Implement new reparenting strategy.
* sigproc.cc (proc_subproc): Reduce access to vchild-pid_handle
and vchild-ppid_handle. Set ppid_sendsig by duplication.
Add PROC_REPARENT case and simplify PROC_CHILDTERMINATED case.
(sig_send): Use ppid_sendsig to signal parent.
(init_child_info): Set pppid_sendsig.
(wait_sig): Add __SIGREPARENT case.
* dcrto.cc (dll_crt0_0): Close pppid_sendsig.
Index: pinfo.h
===
RCS file: /cvs/src/src/winsup/cygwin/pinfo.h,v
retrieving revision 1.64
diff -u -p -r1.64 pinfo.h
--- pinfo.h 12 Sep 2004 03:55:42 - 1.64
+++ pinfo.h 29 Oct 2004 18:56:55 -
@@ -36,12 +36,16 @@ public:
constants below. */
DWORD process_state;
- /* If hProcess is set, it's because it came from a
- CreateProcess call. This means it's process relative
- to the thing which created the process. That's ok because
- we only use this handle from the parent. */
+ /* hProcess comes from a CreateProcess or DuplicateHandle call.
+ This means it's process relative to the thing which created
+ the process. That's ok because we only use this handle in
+ the parent, duplicating it if was temporarily set by the
+ child during reparenting. */
HANDLE hProcess;
+ /* A reparenting operation is in progress */
+ volatile bool isreparenting;
+
#define PINFO_REDIR_SIZE ((char *) myself.procinfo-hProcess - (char *)
myself.procinfo)
/* Handle associated with initial Windows pid which started it all. */
@@ -88,7 +92,7 @@ public:
HANDLE tothem;
HANDLE fromthem;
- void exit (UINT n, bool norecord = 0) __attribute__ ((noreturn, regparm(2)));
+ void exit (UINT n) __attribute__ ((noreturn, regparm(1)));
inline void set_has_pgid_children ()
{
@@ -118,6 +122,7 @@ public:
/* signals */
HANDLE sendsig;
+ HANDLE ppid_sendsig;
private:
sigset_t sig_mask;
CRITICAL_SECTION lock;
Index: sigproc.h
===
RCS file: /cvs/src/src/winsup/cygwin/sigproc.h,v
retrieving revision 1.70
diff -u -p -r1.70 sigproc.h
--- sigproc.h 12 Sep 2004 03:47:57 - 1.70
+++ sigproc.h 29 Oct 2004 18:56:55 -
@@ -26,7 +26,8 @@ enum
__SIGDELETE = -(NSIG + 5),
__SIGFLUSHFAST= -(NSIG + 6),
__SIGHOLD= -(NSIG + 7),
- __SIGNOHOLD = -(NSIG + 8)
+ __SIGNOHOLD = -(NSIG + 8),
+ __SIGREPARENT= -(NSIG + 9)
};
#endif
@@ -38,7 +39,8 @@ enum procstuff
PROC_CHILDTERMINATED = 2,// a child died
PROC_CLEARWAIT = 3,// clear all waits - signal arrived
PROC_WAIT= 4,// setup for wait() for subproc
- PROC_NOTHING = 5 // nothing, really
+ PROC_REPARENT= 5,// reparenting signal arrived
+ PROC_NOTHING = 6 // nothing, really
};
struct sigpacket
Index: child_info.h
===
RCS file: /cvs/src/src/winsup/cygwin/child_info.h,v
retrieving revision 1.45
diff -u -p -r1.45 child_info.h
--- child_info.h12 Sep 2004 18:10:15 - 1.45
+++ child_info.h29 Oct 2004 18:56:55 -
@@ -29,7 +29,7 @@ enum
#define EXEC_MAGIC_SIZE sizeof(child_info)
-#define CURR_CHILD_INFO_MAGIC 0x19c16fb6U
+#define CURR_CHILD_INFO_MAGIC 0xb2f6689eU
/* NOTE: Do not make gratuitous changes to the names or organization of the
below class. The layout is checksummed to determine compatibility between
@@ -46,6 +46,7 @@ public:
HANDLE user_h;
HANDLE parent;
HANDLE pppid_handle;
+ HANDLE pppid_sendsig;
init_cygheap *cygheap;
void *cygheap_max;
DWORD cygheap_reserve_sz;
Index: pinfo.cc
===
RCS file: /cvs/src/src/winsup/cygwin/pinfo.cc,v
retrieving revision 1.121
diff -u -p -r1.121 pinfo.cc
--- pinfo.cc5 Oct 2004 02:10:15 - 1.121
+++ pinfo.cc29 Oct 2004 18:56:56 -
@@ -102,22 +102,38
Re: [Patch] Fixing the PROCESS_DUP_HANDLE security hole.
On Thu, Nov 11, 2004 at 10:48:57PM -0500, Pierre A. Humblet wrote: P.S.: I have no news about the recent patch to /bin/kill -f That is because I was sure that I'd used 'kill -f' to kill windows pids in the past and wanted to check your patch. I haven't been near a WinMe system in a while, though. My vmware version isn't working currently. cgf
Re: [Patch] Fixing the PROCESS_DUP_HANDLE security hole.
On Thu, Nov 11, 2004 at 10:48:57PM -0500, Pierre A. Humblet wrote: Now that 1.5.12 is out, here is a patch to fix the PROCESS_DUP_HANDLE security hole. It uses a new approach to reparenting: the parent duplicates the exec'ed process handle when signaled by the child. Can you refresh my memory (a URL is fine) on the PROCESS_DUP_HANDLE security hole? I'm not 100% certain but I think if you cast back into the dim recesses of cygwin's past, you might find that this is the way things used to be done, to some degree. cgf
Re: [Patch] Fixing the PROCESS_DUP_HANDLE security hole.
At 11:24 PM 11/11/2004 -0500, Christopher Faylor wrote: On Thu, Nov 11, 2004 at 10:48:57PM -0500, Pierre A. Humblet wrote: P.S.: I have no news about the recent patch to /bin/kill -f That is because I was sure that I'd used 'kill -f' to kill windows pids in the past and wanted to check your patch. I haven't been near a WinMe system in a while, though. My vmware version isn't working currently. Funny, I had the same feeling. But this is what happens now: ~: ps PIDPPIDPGID WINPID TTY UIDSTIME COMMAND 606855 1 606855 4294360441 con 740 23:06:35 /c/PROGRAM FILES/CYGWIN/BIN/RXVT 537691 606855 537691 42945045690 740 23:06:36 /c/PROGRAM FILES/CYGWIN/BIN/BASH 460171 537691 460171 42942146850 740 23:24:07 /c/PROGRAM FILES/CYGWIN/BIN/PS ~: /bin/kill -f 4294504569 couldn't open pid 2147483647 2147483647 = 0x7FFF, due to strtol saturating. I just researched the ChangeLog and found a possible cause: 2003-09-20 Christopher Faylor [EMAIL PROTECTED] * kill.cc (main): Allow negative pids (indicates process groups). Another complication is due to #define CW_NEXTPID 0x8000 /* or with pid to get next one */ Pierre
Re: [Patch] Fixing the PROCESS_DUP_HANDLE security hole.
On Thu, Nov 11, 2004 at 11:36:32PM -0500, Pierre A. Humblet wrote: At 11:24 PM 11/11/2004 -0500, Christopher Faylor wrote: On Thu, Nov 11, 2004 at 10:48:57PM -0500, Pierre A. Humblet wrote: P.S.: I have no news about the recent patch to /bin/kill -f That is because I was sure that I'd used 'kill -f' to kill windows pids in the past and wanted to check your patch. I haven't been near a WinMe system in a while, though. My vmware version isn't working currently. Funny, I had the same feeling. But this is what happens now: ~: ps PIDPPIDPGID WINPID TTY UIDSTIME COMMAND 606855 1 606855 4294360441 con 740 23:06:35 /c/PROGRAM FILES/CYGWIN/BIN/RXVT 537691 606855 537691 42945045690 740 23:06:36 /c/PROGRAM FILES/CYGWIN/BIN/BASH 460171 537691 460171 42942146850 740 23:24:07 /c/PROGRAM FILES/CYGWIN/BIN/PS ~: /bin/kill -f 4294504569 couldn't open pid 2147483647 2147483647 = 0x7FFF, due to strtol saturating. That's right. I have seen that from time to time. I just researched the ChangeLog and found a possible cause: 2003-09-20 Christopher Faylor [EMAIL PROTECTED] * kill.cc (main): Allow negative pids (indicates process groups). If that is the cause then bypassing that code when -f is specified should work. But, nevertheless, go ahead and check in your patch. Thanks. cgf
Re: [Patch] Fixing the PROCESS_DUP_HANDLE security hole.
At 11:33 PM 11/11/2004 -0500, Christopher Faylor wrote: On Thu, Nov 11, 2004 at 10:48:57PM -0500, Pierre A. Humblet wrote: Now that 1.5.12 is out, here is a patch to fix the PROCESS_DUP_HANDLE security hole. It uses a new approach to reparenting: the parent duplicates the exec'ed process handle when signaled by the child. Can you refresh my memory (a URL is fine) on the PROCESS_DUP_HANDLE security hole? It starts with http://cygwin.com/ml/cygwin-developers/2003-09/msg00078.html Eventually things were broken down in several patches. The part about the tty gave rise to your archetype and the abandon of vfork. Very long story. I'm not 100% certain but I think if you cast back into the dim recesses of cygwin's past, you might find that this is the way things used to be done, to some degree. The patch relies heavily on your implementation of signals using a pipe, which allows to carry extra info. Pierre
Re: vpath search failures in Clearcase when files do actually exist
Pierre A. Humblet schrieb: On Wed, Nov 10, 2004 at 01:09:47PM -0500, Christopher Faylor wrote: On Wed, Nov 10, 2004 at 12:47:26PM -0500, Pierre A. Humblet wrote: On Tue, Nov 09, 2004 at 07:51:23PM -0500, Christopher Faylor wrote: On Tue, Nov 09, 2004 at 07:21:00PM -0500, Jeremy Broughton wrote: Cyg ver:1.3.10 (although I have reproduced the same problem with 1.5.7) OS : Windows 2000 File system: MVFS (running gnumake inside a Rational Clearcase dynamic view) cygcheck: (See attached file: cygcheck.out) It appears as though there is a problem using the /cygdrive/m type paths inside MVFS when there are extremely large dependency trees. Unless you can point to problems with a standard, supported W2K filesystem, please don't expect anyone to try to fix this. Cygwin uses standard Win32 APIs for detemining if something is a directory. If those APIs are not working for Clearcase then it is unlikely that anyone will be motivated to fix the problem on the Cygwin side. The OP was suggesting both a problem determining if something is a directory and a problem related to pathnames. Running strace my shed some light on the root cause. Would it have made you feel better if I had said for dealing with paths and determining if something is a directory? The related to pathnames was refering to the last paragraph in the initial post, which points to Cygwin (although I can't imagine how): As shown above, using vpath . ../.. /cygdrive/m/top fails, even though it should work. Using vpath . /cygdrive/m/top fails with the same errors as above. However, vpath . M:/top succeeds in the dependency check, even though M:/top should apparently be written as /cygdrive/m/top to comply to cygwin standards. If the OP is willing to do this then I hope he does it on a standard filesystem like NTFS, as I originally suggested. If there is a problem there then it is obviously not a ClearCase problem. Right. But then if it is a ClearCase problem, it can't be identified... My request would be to trace under the latest Cygwin. BTW: Can someone provide testcode to identify if the current drive as MVFS please. At least I think that a clearcase view is mounted as driveletter. Does it show up as GetLogicalDrives()? (should so) In detail, what does GetVolumeInformation().lpFileSystemNameBuffer return for such a MVFS drive? For the proposed mounttable (getmntent) changes. http://cygwin.com/ml/cygwin/2004-10/msg01449.html user|system = mnt_opts GetVolumeInformation().lpFileSystemNameBuffer as mnt_type. -- Reini Urban http://xarch.tu-graz.ac.at/home/rurban/ -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
Re: What to do when setup fails?
Luke Kendall schrieb: On 11 Nov, luke wrote: By the time it had crashed with the above panel, Task Manager was showing setup of using only 10MB of memory. Yet as soon as I clicked on OK for the runtime error, the pages were very quickly handed back to the system and it returned to about 100MB of memory in use (no page faults). So it sounds like setup.exe was the culprit. This is an old version of setup. I notice that my post-install stuff detected the problem of a failed install (symptom: missing /etc/profile file). The sanest thing to do will be to reboot, and try installing a much more recent Cygwin. If that fails, I'll scrub the installed files and reg keys, and try again. Hmm, using a mirror that's a few days old shows similar behaviour. This time though it's saying it's installing zip-2.3-6 - all the other details seem the same. I watched in task manager as setup's memory use climbed from 50MB to 127MB over a period of a few minutes. When next I looked nothing had visibly changed, except task manager reckoned setup was now using only 12MB (again climbing steadily upwards at about 5MB per minute), even as the Performance task manager view showed system memory use had climbed to 880MB (and still climbing). Looks like it crawls your local-package-dir tree (aka Select download directory) to look for all available setup.ini's and tar.bz2 packages. You should really check your /var/log/setup.log.full what's going on and not only taskman, to see that's something going on. You really should provide a meaningful local-package-dir otherwise it will crawl your entire root, which could last 30 mins if it's cygfile:/// But maybe there's an undetected recursive loop possibility somewhere. Have to investigate. (Note: the fromcwd() step) Do you have junctions, directory hardlinks under your local-package-dir? Sounds like the first attempted install, with the insufficient permissions and the Mount: command completed successfully might have left things in a somehow strange state. I'll force a system disc check and reboot and try again. Then scrub Cygwin and try a fresh install if that's no better. -- Reini Urban http://xarch.tu-graz.ac.at/home/rurban/ -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
Re: library `/usr/lib/libncurses.la' was moved.
Gerrit P. Haase schrieb: Getting this warning: libtool: link: warning: library `/usr/lib/libncurses.la' was moved. $ cat /usr/lib/libncurses.la [...] # Directory that this library needs to be installed in: libdir='/usr/src/ncurses/ncurses-5.4/.inst/usr/lib' We should really add some kind of .la test to our gbs. Or maybe into setup.exe, for those people who already have their gbs customized or not using it? Or provide a gbs-library? It would be easier to upgrade then. I had a lot such broken .la's, which forced me to post my fixups last month of a couple of such .la's. libcurl also. -- Reini Urban http://xarch.tu-graz.ac.at/home/rurban/ -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
Re: rlogind vs. smb
On Wed, Nov 10, 2004 at 11:02:40AM -0500, Igor Pechtchanski wrote: 1. does anybody know why rlogind doesn't like SMB shares? http://cygwin.com/cygwin-ug-net/ntsec.html#NTSEC-SWITCH, or see /usr/share/doc/Cygwin/inetutils-1.3.2.README I think this is my problem... It sure sounds like it Arguably it's not a good idea to create a user that has the create process level token privilige, as which the inetd service will run, so that it can setuid /AND/ access shares... I think i'll just keep local copies of .rhosts as part of the installation procedure... *sigh* Should be in /var/log/inetd.log or the Windows Event Log (Application). /var/log/inetd.log contains spawn info for the rlogind process, but not much more... As for the event log, it contains somewhat sparse info. Maybe this is a configuration problem? Should all syslogish output really be directed to the event log? (not that rlogin is very helpful in it's logging... =P) If by promiscuous you mean persistent, go to the Service Control Manager and set Startup Type for the inetd service to Automatic. promiscuous mode, in pam's rlogin auth lib, means that the '+' wild cards are OK. By default, on most distributions, unless the word 'promiscuous' is in one of it's options, these pluses are disregarded. I haven't reached the stage where i've figured out whether this is really required or not... ;-) -- Yuval Kogman, Sysadmin HyperRoll Israel, Ltd. -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
RE: stack overflow in cygwin_nt-5.0
-Original Message- From: cygwin-owner On Behalf Of Marcel Telka Sent: 10 November 2004 20:42 To: cygwin Subject: Re: stack overflow in cygwin_nt-5.0 Don't declare huge data structures on the stack. I have to use large matrices (even larger) in my application. So, how to do it? malloc() ? Or make them static, instead of local variables. Or use the compiler flag -Wl,--stack=[big number goes here] cheers, DaveK -- Can't think of a witty .sigline today -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
if construct doesn't work in makefile
Hi, In my make file I am trying to check whether a directory exists or not then set a path differently if doesn't exist. I am using cygwin installed on windows 2000. The if construct I have is as follows. if test [-dc:/tools]; then DRIVE = c:/tools; else DRIVE = c:/altTools; fi pls help me out. Regards, sitaram -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
howto register process
Hi, I have an application that is created to work on Unix but it is supposed to run on Windows as well. The problem is that we have a daemon that spawns a few children and later it needs to kill them. The parent process is supposed to communicate with its children by signals and to eventually terminate them. I would like to use kill(pid, signal) to kill the processes because they rely on the signals mechanizm. The problem is that kill returns not such pid. Is there a way to register a process in the cygwin process table so that later be able to kill it with kill()? TIA. Regards, -- Kamen TOMOV UNIX Software Engineer -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
RE: if construct doesn't work in makefile
[EMAIL PROTECTED] wrote: Hi, In my make file I am trying to check whether a directory exists or not then set a path differently if doesn't exist. I am using cygwin installed on windows 2000. The if construct I have is as follows. if test [-dc:/tools]; then DRIVE = c:/tools; else DRIVE = c:/altTools; fi pls help me out. Regards, sitaram Does the following work better? if test [ -d /cygdrive/C/tools ]; then DRIVE=c:/tools; else DRIVE= By the way: this seems not to be a cygwin problem and it would be very helpful to provide at least the error message. See http://cygwin.com/problems.html matthias -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
Re: library `/usr/lib/libncurses.la' was moved.
Reini Urban wrote: Gerrit P. Haase schrieb: Getting this warning: libtool: link: warning: library `/usr/lib/libncurses.la' was moved. $ cat /usr/lib/libncurses.la [...] # Directory that this library needs to be installed in: libdir='/usr/src/ncurses/ncurses-5.4/.inst/usr/lib' We should really add some kind of .la test to our gbs. Or maybe into setup.exe, for those people who already have their gbs customized or not using it? Naah, it is a libtool bug, for sure. Sanity check is good, but it shouldn't be neccessary. Or provide a gbs-library? It would be easier to upgrade then. I don't get this point. Would you explain, please? I had a lot such broken .la's, which forced me to post my fixups last month of a couple of such .la's. libcurl also. Wrong libtool version...;) Gerrit -- =^..^= -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
Re: if construct doesn't work in makefile
Prasad, Kanuparthi wrote: Hi, In my make file I am trying to check whether a directory exists or not then set a path differently if doesn't exist. I am using cygwin installed on windows 2000. The if construct I have is as follows. if test [-dc:/tools]; then DRIVE = c:/tools; else DRIVE = c:/altTools; fi pls help me out. Try if test [-dc:/tools] ; then DRIVE=c:/tools; else DRIVE=c:/altTools; fi instead, and/or use cygwin syntax to access drives. Gerrit -- =^..^= -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
Re: looking for an arm9 cross-compiler
At 05:20 AM 11/10/2004, you wrote: Hi, I used to work with Linux and I'm trying Cygwin. Does anyone know if there's an existing cross compiler for arm9 package? I could not find it in the package list. Otherwise, what is the solution? Do I have to compile an arm-gcc with the cygwin library? I'm sure there's already an existing arm compiler I could use without having to die trying to make my own. You may be right but as you've found, there isn't one that's officially available via 'setup.exe'. There has been occasional inquiries like this in the past, with much the same response as I'm giving you now (either that or no response ;-) ). Building a cross compiler in general shouldn't be that painful. You can find some general pointers on how to do that in the email archives. Or there's always www.objsw.com/CrossGCC. -- Larry Hall http://www.rfk.com RFK Partners, Inc. (508) 893-9779 - RFK Office 838 Washington Street (508) 893-9889 - FAX Holliston, MA 01746 -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
RE: if construct doesn't work in makefile
-Original Message-
From: cygwin-owner On Behalf Of Gerrit P. Haase
Sent: 11 November 2004 11:32
Prasad, Kanuparthi wrote:
Hi,
In my make file I am trying to check whether a directory
exists or not
then set a path differently if doesn't exist.
I am using cygwin installed on windows 2000.
The if construct I have is as follows.
if test [-dc:/tools]; then DRIVE = c:/tools; else DRIVE =
c:/altTools; fi
pls help me out.
Try
if test [-dc:/tools] ; then DRIVE=c:/tools; else DRIVE=c:/altTools; fi
instead, and/or use cygwin syntax to access drives.
There's nothing wrong with the command syntax in general. The problem is that
in bash, you can't put a space between the VARIABLE and the = sign, or bash
thinks it's an executable to search the $PATH for instead of a variable to
assign to.
[EMAIL PROTECTED] /tmp/bgcc unset FRED
[EMAIL PROTECTED] /tmp/bgcc if test [-dc:/baar]; then FRED=yes; else FRED=no;
fi
[EMAIL PROTECTED] /tmp/bgcc echo ${FRED}
yes
[EMAIL PROTECTED] /tmp/bgcc unset FRED
[EMAIL PROTECTED] /tmp/bgcc if test [-dc:/baar]; then FRED = yes; else FRED =
no; fi
bash: FRED: command not found
[EMAIL PROTECTED] /tmp/bgcc if test [-dc:/baar]; then FRED =yes; else FRED
=no; fi
bash: FRED: command not found
[EMAIL PROTECTED] /tmp/bgcc if test [-dc:/baar]; then FRED=yes; else FRED =no;
fi
[EMAIL PROTECTED] /tmp/bgcc
cheers,
DaveK
--
Can't think of a witty .sigline today
--
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Problem reports: http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ: http://cygwin.com/faq/
Re: [ANNOUNCEMENT] Updated: docbook-xml42-4.2-3
Marcel Telka schrieb: I've updated the docbook-xml42 package to version 4.2-3. docbook-xml42 package contains Docbook XML DTD version 4.2 as published by OASIS. Changes since 4.2-2: - Added Public Identifier association with an URI into /etc/xml/catalog. - Moved DTD files to /usr/share/xml/docbook/4.2 directory. Hi Marcel, thanks for your good work. Since I expected problems with my /etc/xml/catalog file, I kept an eye on it. And indeed it got corrupted by the update. Here is what happened: - the new publicId was inserted ok - the location of the DTD was updated ok - existing entries were not altered BUT: - all group definitions were lost - all identation was lost I use group definitions to set a common xml:base for all group elements. As a result, the paths of every former group element got corrupted. Can you please fix the catalog updating script? Cheers, Patrick -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
RE: if construct doesn't work in makefile
-Original Message- From: cygwin-owner On Behalf Of Morche Matthias Sent: 11 November 2004 12:22 cygwin-ownerOOOPScygwinBADOIN!com wrote: Matthias, http://cygwin.com/acronyms#PCYMTNQREAIYR, thanks! In my make file I am trying to check whether a directory exists or not then set a path differently if doesn't exist. I am using cygwin installed on windows 2000. The if construct I have is as follows. if test [-dc:/tools]; then DRIVE = c:/tools; else DRIVE = c:/altTools; fi pls help me out. Regards, sitaram Does the following work better? if test [ -d /cygdrive/C/tools ]; then DRIVE=c:/tools; else DRIVE= I thought it would be that as well, but in fact test seems to be happy with no spaces around the square brackets. I think it may be only if you want to use the implicit form of test that the brackets need to be separated with spaces from the test inside them, so that bash spots them as a separate token. [EMAIL PROTECTED] /tmp/bgcc if [-dc:/baar]; then FRED=yes; else FRED=no; fi bash: [-dc:/baar]: No such file or directory [EMAIL PROTECTED] /tmp/bgcc if [ -dc:/baar ]; then FRED=yes; else FRED=no; fi [EMAIL PROTECTED] /tmp/bgcc if test [-dc:/baar]; then FRED=yes; else FRED=no; fi [EMAIL PROTECTED] /tmp/bgcc Yep, that looks fairly conclusive to me. cheers, DaveK -- Can't think of a witty .sigline today -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
Re: library `/usr/lib/libncurses.la' was moved.
Gerrit P. Haase schrieb: Reini Urban wrote: Gerrit P. Haase schrieb: Getting this warning: libtool: link: warning: library `/usr/lib/libncurses.la' was moved. $ cat /usr/lib/libncurses.la [...] # Directory that this library needs to be installed in: libdir='/usr/src/ncurses/ncurses-5.4/.inst/usr/lib' We should really add some kind of .la test to our gbs. Or maybe into setup.exe, for those people who already have their gbs customized or not using it? Naah, it is a libtool bug, for sure. Sanity check is good, but it shouldn't be neccessary. Or provide a gbs-library? It would be easier to upgrade then. I don't get this point. Would you explain, please? Modularization of the gbs: Provide a central place for some common gbs functions, like /usr/share/cygwin/gbs-functions.sh, which would be sourced by each build script. Any updates on this library will help all build scripts. Currently you have to fix/update every single instance. I'm thinking of functions which are very likely not touched. The init step which parses the PKG/VER/REL, the name defaults, install_docs, mkdirs, prep, autoreconf, reconf, mkpatch, build, certain install steps, strip, list, depend, requires, pkg, spkg, finish, sigfile, checksig, ... But it's good to have some sane defaults which will get inherited to all new and old build scripts your are working with. I had a lot such broken .la's, which forced me to post my fixups last month of a couple of such .la's. libcurl also. Wrong libtool version...;) But not on my side. Certain wrong devel packages install strange la's into /lib. I hope mine are correct. -- Reini Urban http://xarch.tu-graz.ac.at/home/rurban/ -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
RE: howto register process
-Original Message- From: cygwin-owner On Behalf Of Kamen TOMOV Sent: 11 November 2004 11:01 To: cygwin Subject: howto register process Hi, I have an application that is created to work on Unix but it is supposed to run on Windows as well. The problem is that we have a daemon that spawns a few children and later it needs to kill them. The parent process is supposed to communicate with its children by signals and to eventually terminate them. I would like to use kill(pid, signal) to kill the processes because they rely on the signals mechanizm. The problem is that kill returns not such pid. Is there a way to register a process in the cygwin process table so that later be able to kill it with kill()? TIA. You can't kill a windows process with cygwin kill. Presumably your software is using the windows native functions to spawn child processes? That's not necessary: cygwin gives you unix syscalls on windows. If you used the standard posix functions (fork/exec/etc) to start the children, kill would work fine on them, and you'd have a lot more common code and a lot less differences to keep track of between your unix and 'doze versions of your code. cheers, DaveK -- Can't think of a witty .sigline today -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
RE: looking for an arm9 cross-compiler
-Original Message- From: cygwin-owner On Behalf Of Larry Hall Sent: 11 November 2004 12:44 Hi, I used to work with Linux and I'm trying Cygwin. Does anyone know if there's an existing cross compiler for arm9 package? I could not find it in the package list. Otherwise, what is the solution? Do I have to compile an arm-gcc with the cygwin library? I'm sure there's already an existing arm compiler I could use without having to die trying to make my own. You may be right but as you've found, there isn't one that's officially available via 'setup.exe'. There has been occasional inquiries like this in the past, with much the same response as I'm giving you now (either that or no response ;-) ). Building a cross compiler in general shouldn't be that painful. You can find some general pointers on how to do that in the email archives. Or there's always www.objsw.com/CrossGCC. That site is years obsolete and out of date. More recent is Bill Gatliff's crossgcc site at http://crossgcc.billgatliff.com which has a script that builds arm-elf-gcc, and even more up-to-date would be Dan Kegel's crosstool, another build script, available at http://www.kegel.com/crosstool/ cheers, DaveK -- Can't think of a witty .sigline today -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
Re: Creating superuser (root) in Cygwin
Ok, ok. Those who know more than I do. Today. I'm trying to compile and install kalarm, which I have downloaded and decompressed. Now the instructions go as follows: cd kalarm-1.1.6 ./configure make [ Log in as root ] - !!! HOW !!! cd /path/kalarm-1.1.6 make install My question is. How does a Cygwin user (without a root account) get round this kidda problem? Please resist the urge to flame. (Some of us are just new, that's all.) Thanks, - Olumide - -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
RE: if construct doesn't work in makefile
-Original Message-
From: cygwin-owner On Behalf Of Prasad, Kanuparthi
Sent: 11 November 2004 11:16
In my make file I am trying to check whether a directory
exists or not
then set a path differently if doesn't exist.
I am using cygwin installed on windows 2000.
The if construct I have is as follows.
if test [-dc:/tools]; then DRIVE = c:/tools; else DRIVE =
c:/altTools; fi
Oh, and by the way, another way you can do this, using only makefile syntax so
it doesn't involve the shell at all, is to use the $(wildcard ...) function with
a full filename to see if it exists, and then use that in a $(if ...) test.
Like so:
DRIVE=$(if $(wildcard c:/tools), c:/tools, c:/altTools)
This technique has the advantage of working whether you're using make in --win32
or --unix mode:
[EMAIL PROTECTED] /tmp/bgcc ls -la C:/tools
ls: C:/tools: No such file or directory
[EMAIL PROTECTED] /tmp/bgcc cat makefoo
DRIVE=$(if $(wildcard c:/tools), c:/tools, c:/altTools)
all:
echo drive is ${DRIVE}
[EMAIL PROTECTED] /tmp/bgcc make -f makefoo
echo drive is c:/altTools
drive is c:/altTools
[EMAIL PROTECTED] /tmp/bgcc make --win32 -f makefoo
echo drive is c:/altTools
drive is c:/altTools
[EMAIL PROTECTED] /tmp/bgcc mkdir C:/tools
[EMAIL PROTECTED] /tmp/bgcc make -f makefoo
echo drive is c:/tools
drive is c:/tools
[EMAIL PROTECTED] /tmp/bgcc make --win32 -f makefoo
echo drive is c:/tools
drive is c:/tools
[EMAIL PROTECTED] /tmp/bgcc
since it depends only on the cygwin dll's path-handling abilities, and not on
the nature of the shell that is being used to execute the commands.
cheers,
DaveK
--
Can't think of a witty .sigline today
--
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Problem reports: http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ: http://cygwin.com/faq/
RE: Creating superuser (root) in Cygwin
-Original Message- From: cygwin-owner On Behalf Of Olumide Sent: 11 November 2004 13:23 Ok, ok. Those who know more than I do. Today. I'm trying to compile and install kalarm, which I have downloaded and decompressed. Now the instructions go as follows: cd kalarm-1.1.6 ./configure make [ Log in as root ] - !!! HOW !!! cd /path/kalarm-1.1.6 make install My question is. How does a Cygwin user (without a root account) get round this kidda problem? Just make sure the username that you've logged into windows with has the required admin privs; FAPP you probably can just ignore the requirement. Please resist the urge to flame. Aww.. :) http://cygwin.com/acronyms#WJM cheers, DaveK -- Can't think of a witty .sigline today -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
RE: if construct doesn't work in makefile
Hi, I tried, but it doesn't work. I tried using cygwin syntax to access drives. but invain. Also I tried by typing this script at the prompt. but it doesn't work the way it is expected. Should I need to execute it in a separate shell, if so, how is it done. Thanks Regards, sitaram -Original Message- From: Gerrit P. Haase [mailto:[EMAIL PROTECTED] Sent: 11 November 2004 11:32 To: Prasad, Kanuparthi Cc: '[EMAIL PROTECTED]' Subject: Re: if construct doesn't work in makefile Prasad, Kanuparthi wrote: Hi, In my make file I am trying to check whether a directory exists or not then set a path differently if doesn't exist. I am using cygwin installed on windows 2000. The if construct I have is as follows. if test [-dc:/tools]; then DRIVE = c:/tools; else DRIVE = c:/altTools; fi pls help me out. Try if test [-dc:/tools] ; then DRIVE=c:/tools; else DRIVE=c:/altTools; fi instead, and/or use cygwin syntax to access drives. Gerrit -- =^..^= -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
stderr and clearmake
Hello, I'm using clearmake, and the Metaware C++ compiler (don't ask...I
don't like metaware either) and whenever I execute a clearmake from
within cygwin, I don't have any stderr output. This was first discovered
while I was in an Xterm, and didn't affect the main window. However,
recently, for some reason, it's started happening in the cygwin window
(by cygwin window, I mean bash.exe) as well.
I've written a program to print out a message on stdout and then on
stderr. The output is as follows:
[EMAIL PROTECTED] ~
$ ./stderr-test.exe
This is stderr!
This is stdout!
The code is:
#include stdio.h
int main(){
fprintf(stdout, This is stdout!\n);
fprintf(stderr, This is stderr!\n);
return 0;
}
As you can see, for some reason stderr and stdout are being manipulated
in some odd fashion, which may or may not be related to the problem I'm
having. I've already tried manipulating the cygwin variable, without
success.
I can attach information about the cygwin version I'm running.
running: Windows 2000 Professional Ver 5.0 Build 2195 Service Pack 4
Cygwin version: 1.5.11-1
--
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Problem reports: http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ: http://cygwin.com/faq/
Re: if construct doesn't work in makefile
Prasad, Kanuparthi wrote: Hi, I tried, but it doesn't work. I tried using cygwin syntax to access drives. but invain. Also I tried by typing this script at the prompt. but it doesn't work the way it is expected. Should I need to execute it in a separate shell, if so, how is it done. A snippet from the Makefile you're using would help, please see also the other postings with more tipps - http://cygwin.com/ml/cygwin/2004-11/ Gerrit -- =^..^= -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
RE: if construct doesn't work in makefile
Gerrit P. Haase wrote on Thursday, November 11, 2004 3:13 PM: [snip] A snippet from the Makefile you're using would help, please see also the other postings with more tipps - http://cygwin.com/ml/cygwin/2004-11/ ... or older: http://www.cygwin.com/ml/cygwin/2000-06/msg01318.html - Jörg -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
[ANNOUNCEMENT] New Package: mathomatic-11.5-1
I've made a new package available on sourceware: mathomatic See http://www.mathomatic.com Mathomatic is a highly portable, general purpose symbolic math program that can solve, simplify, combine, differentiate, integrate, and compare algebraic equations. It can do standard, complex number, and polynomial arithmetic. It is designed to be as general as possible, with few options. It is a console mode application that compiles and installs easily. It also prints the results in a cool, easily readable and colored console layout. License: LPGL 2.1 Sample: ; iterative root approximation equation: x_n = x_o*y/(x_o^n))-1)/n)+1) simp repl x_n x_o with x x This is just the mathomatic console client from the mathomatic package. Not distributed are the samples for prime number generation: twin, sumsq, primes and around. If you want these also, get the src packages and compile them by yourself. maxima (on clisp) is also suitable and maybe better, but harder to install. And there exist dozens of other symbolic algebra packages. -- Reini Urban Please report any bugs to the Cygwin mailing list, following the procedure in http://cygwin.com/problems.html. To update your installation, click on the Install Cygwin now link on the http://cygwin.com/ web page. This downloads setup.exe to your system. Then, run setup and answer all of the questions. *** CYGWIN-ANNOUNCE UNSUBSCRIBE INFO *** If you want to unsubscribe from the cygwin-announce mailing list, look at the List-Unsubscribe: tag in the email header of this message. Send email to the address specified there. It will be in the format: [EMAIL PROTECTED] If you need more information on unsubscribing, start reading here: http://sources.redhat.com/lists.html#unsubscribe-simple Please read *all* of the information on unsubscribing that is available starting at the above URL. -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
Re: stderr and clearmake
On Thu, Nov 11, 2004 at 09:08:04AM -0500, Aaron Conole wrote:
Hello, I'm using clearmake, and the Metaware C++ compiler (don't ask...I
don't like metaware either) and whenever I execute a clearmake from
within cygwin, I don't have any stderr output. This was first discovered
while I was in an Xterm, and didn't affect the main window. However,
recently, for some reason, it's started happening in the cygwin window
(by cygwin window, I mean bash.exe) as well.
I've written a program to print out a message on stdout and then on
stderr. The output is as follows:
[EMAIL PROTECTED] ~
$ ./stderr-test.exe
This is stderr!
This is stdout!
The code is:
#include stdio.h
int main(){
fprintf(stdout, This is stdout!\n);
fprintf(stderr, This is stderr!\n);
return 0;
}
As you can see, for some reason stderr and stdout are being manipulated
in some odd fashion, which may or may not be related to the problem I'm
having. I've already tried manipulating the cygwin variable, without
success.
I can attach information about the cygwin version I'm running.
I've noticed that when I connect my TV to the cable box I can't get
channel 7 although when I connect it directly to the cable I have no
problem. What is wrong with my TV? I've actually tried several TVs and
all of them seem to have this weird problem.
cgf
--
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Problem reports: http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ: http://cygwin.com/faq/
RE: stderr and clearmake
-Original Message-
From: cygwin-owner On Behalf Of Aaron Conole
Sent: 11 November 2004 14:08
I've written a program to print out a message on stdout and then on
stderr. The output is as follows:
[EMAIL PROTECTED] ~
$ ./stderr-test.exe
This is stderr!
This is stdout!
The code is:
#include stdio.h
int main(){
fprintf(stdout, This is stdout!\n);
fprintf(stderr, This is stderr!\n);
return 0;
}
As you can see, for some reason stderr and stdout are being
manipulated in some odd fashion,
No they aren't. That's exactly the output you'd expect to see from that
program. What on earth are you expecting to happen?
cheers,
DaveK
--
Can't think of a witty .sigline today
--
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Problem reports: http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ: http://cygwin.com/faq/
Re: vpath search failures in Clearcase when files do actually exist
On Thu, Nov 11, 2004 at 09:33:21AM +0100, Reini Urban wrote: BTW: Can someone provide testcode to identify if the current drive as MVFS please. At least I think that a clearcase view is mounted as driveletter. Does it show up as GetLogicalDrives()? (should so) In detail, what does GetVolumeInformation().lpFileSystemNameBuffer return for such a MVFS drive? Here is the output of GetVolumeInformation: y:\ Ret 1 Volume: CCase serial: 36984713 component: 255 flags: 3 name: MVFS HTH Pierre -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
[ANNOUNCEMENT] Updated: cygwin-1.5.11-1
I've made a new version of the Cygwin DLL and associated utilities available for download. As usual, a list of what has changed is below. To update your installation, click on the Install Cygwin now link on the http://cygwin.com/ web page. This downloads setup.exe to your system. Then, run setup and answer all of the questions. If you have questions or comments, please send them to the Cygwin mailing list at: [EMAIL PROTECTED] . *** CYGWIN-ANNOUNCE UNSUBSCRIBE INFO *** If you want to unsubscribe from the cygwin-announce mailing list, look at the List-Unsubscribe: tag in the email header of this message. Send email to the address specified there. It will be in the format: [EMAIL PROTECTED] If you need more information on unsubscribing, start reading here: http://sources.redhat.com/lists.html#unsubscribe-simple Please read *all* of the information on unsubscribing that is available starting at this URL. Christopher Faylor TimeSys, Inc. Changes since 1.5.11-1: - Fix problem where stdout was closed when a thread exits. (Christopher Faylor) - Fix pipe problems with Windows 95. (Bas van Gompel) - Fix race condition in fork which occasionally caused bash to think a forked process had died prematurely. (Christopher Faylor) - Properly deal with CTRL-C while a cygwin process is in startup code. (Pierre Humblet) - Handle RTLD_DEFAULT in dlsym. (Sam Steingold) - Don't try to use Windows delete on close for files on remote shares. (Corinna Vinschen) - Properly handle trailing dots in windows file names. (Pierre Humblet) - Make the default /cygdrive binary mode. (Christopher Faylor) - Add siginterrupt declaration to signal.h. (Christopher Faylor) - Use proper wait value for tty reads. (Mark Paulus) - Properly reset SIGCHLD blocking after a spawn when there was no previous signal blocking. (Christopher Faylor) - Work hard to ensure that SYSTEMROOT environment variable is always passed to new processes even when it has ostensibly been deleted. This should fix some strange network problems. (Corinna Vinschen) - Fix network slowdowns experienced by some firewall users. (Christopher Faylor) - Fix return value for unbuffered fread. (Jason Tishler) - Properly load registry hives and insure access to system mounts. (Pierre Humblet) - Cleanup cygcheck and getfacl output. (Bas van Gompel) - Add cygcheck warning about multiple cygwin DLLs. (Bas van Gompel) -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
[ANNOUNCEMENT] Re: Updated: cygwin-1.5.*12*-1
Sorry. That previous announcement was obviously for cygwin-1.5.12-1 *not* the already-released cygwin-1.5.11-1. cgf -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
RE: looking for an arm9 cross-compiler
I've had some luck with the ECOS version. You can find it at http://sources.redhat.com/ecos/ -Jerry -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
Re: Squid Patch for Cygwin 1.5
I'd like to take the compile option. However, don't know which file and line to modify. Can you elaborate more ? Ngai Mo -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
RE: stderr and clearmake
-Original Message-
From: Aaron Conole
Sent: 11 November 2004 15:17
To: Dave Korn
Subject: Re: stderr and clearmake
:) Aaron, let's keep this on the mailing list, I charge extortionate rates for
private consultancy, and anyway, 1) there are a lot more people able to help and
offer advice on the list than just me and they might think of or know things I
don't, 2) if I make any mistakes or offer bad advice someone will pull me on it
fairly sharpish, 3) and of course by keeping the conversation on the list it
goes into the web archives, where it is searchable and may help out anyone who
runs up against the same problem in the future. Thanks. Now on with the code:
Dave Korn wrote:
-Original Message-
From: cygwin-owner On Behalf Of Aaron Conole
Sent: 11 November 2004 14:08
I've written a program to print out a message on stdout and then on
stderr. The output is as follows:
[EMAIL PROTECTED] ~
$ ./stderr-test.exe
This is stderr!
This is stdout!
The code is:
#include stdio.h
int main(){
fprintf(stdout, This is stdout!\n);
fprintf(stderr, This is stderr!\n);
return 0;
}
As you can see, for some reason stderr and stdout are being
manipulated in some odd fashion,
No they aren't. That's exactly the output you'd expect to
see from that
program. What on earth are you expecting to happen?
I'm expecting stderr after stdout. I'm seeing stderr before
stdout. At
first I assumed that it had something to do with the way buffers were
being flushed, but under a cmd window (djgpp compiled
executable) I get
stdout before stderr, as I expect.
I don't think you have any right to expect that. IIRC, there are _zero_
guarantees provided by the C language standard and the C stdio library
specification about the order in which stdio buffers will be flushed. It's
entirely at the whim of the runtime library, and as you can see, when a program
reaches the termination routines under cygwin, it chooses to flush the stderr
buffer before the stdout buffer.
Use fflush explicitly if you need to manage the order:
int main(){
fprintf(stdout, This is stdout!\n);
fflush (stdout); /* force that line to be emitted */
fprintf(stderr, This is stderr!\n);
fflush (stderr); /* now force the next line to be emitted */
return 0;
}
cheers,
DaveK
--
Can't think of a witty .sigline today
--
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Problem reports: http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ: http://cygwin.com/faq/
/cygdrive command line completion?
Is this a FAQ? Can't find it. How do I get /cygdrive to auto-complete on the command line? -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
RE: /cygdrive command line completion?
-Original Message- From: cygwin-owner On Behalf Of Raye Raskin Sent: 11 November 2004 17:08 Is this a FAQ? Can't find it. How do I get /cygdrive to auto-complete on the command line? mkdir /cygdrive WFM, anyway. cheers, DaveK -- Can't think of a witty .sigline today -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
Re: stderr and clearmake
On Thu, Nov 11, 2004 at 05:02:40PM -, Dave Korn wrote: 2) if I make any mistakes or offer bad advice someone will pull me on it fairly sharpish, I don't think you have any right to expect that. IIRC, there are _zero_ guarantees provided by the C language standard and the C stdio library specification about the order in which stdio buffers will be flushed. It's entirely at the whim of the runtime library, and as you can see, when a program reaches the termination routines under cygwin, it chooses to flush the stderr buffer before the stdout buffer. stderr is not normally buffered. stdout is normally unbuffered or line buffered if is attached to a tty. So, stderr should have been flushed before hitting exit. So should stdout in the normal case. On Cygwin, I get stdout first, stderr second with and without setting CYGWIN=tty. If I pipe or redirect the output somewhere then I get the opposite order, which is what I would expect. Which, means, of course, that my cable box analogy still holds. cgf -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
Re: /cygdrive command line completion?
On Thu, Nov 11, 2004 at 05:10:50PM -, Dave Korn wrote: -Original Message- From: cygwin-owner On Behalf Of Raye Raskin Sent: 11 November 2004 17:08 Is this a FAQ? Can't find it. How do I get /cygdrive to auto-complete on the command line? mkdir /cygdrive WFM, anyway. If you type ls /cygdrive/tab it should just work without creating the directory. The directory is only necessary if you want to do: ls /cygtab and have the drive part added. cgf -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
RE: /cygdrive command line completion?
-Original Message- From: cygwin-owner On Behalf Of Christopher Faylor Sent: 11 November 2004 17:18 On Thu, Nov 11, 2004 at 05:10:50PM -, Dave Korn wrote: -Original Message- From: cygwin-owner On Behalf Of Raye Raskin Sent: 11 November 2004 17:08 Is this a FAQ? Can't find it. How do I get /cygdrive to auto-complete on the command line? mkdir /cygdrive WFM, anyway. If you type ls /cygdrive/tab it should just work without creating the directory. Ah, didn't guess that. I had tried ls /cygdrivetab and it didn't work, so The directory is only necessary if you want to do: ls /cygtab and have the drive part added. Unfortunately, the presence of cygwin.bat and cygwin.ico in the root tend to make a mess of that, and even making them both non-executable doesn't help. cheers, DaveK -- Can't think of a witty .sigline today -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
Re: /cygdrive command line completion?
Is this a FAQ? Can't find it. How do I get /cygdrive to auto-complete on the command line? mkdir /cygdrive WFM, anyway. WFM, too. Thanks. If you type ls /cygdrive/tab it should just work without creating the directory. Ha! The whole point was to avoid having to type /cygdrive ! This won't work for me anyway since I have 'c' and 'd' drives. Ah, didn't guess that. I had tried ls /cygdrivetab and it didn't work, so The directory is only necessary if you want to do: ls /cygtab and have the drive part added. Unfortunately, the presence of cygwin.bat and cygwin.ico in the root tend to make a mess of that, and even making them both non-executable doesn't help. Rename cygwin.bat to Cygwin.bat and cygwin.ico to Cygwin.ico. WFM, anyway. ;-) Raye. cheers, DaveK -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
RE: if construct doesn't work in makefile
if test [-dc:/tools]; then ... ... in fact test seems to be happy with no spaces around the square brackets. I think it may be only if you want to use the implicit form of test that the brackets need to be separated with spaces from the test inside them, so that bash spots them as a separate token. No, you *always* need to separate test tokens with spaces. The constructs ... test expr and [ expr ] ... are equivalent. Use either the test command or square brackets, but not both. If you do attempt to use square brackets in the expression, then they have no special significance. With a single argument, test just determines if the argument string is non-null. So the command ... test [-dc:/tools] ... always returns a true exit status, simply because the single argument string [-dc:/tools] is not empty. You can prove this by doing ... bash$ test [-d/does-not-exist] echo yes yes ... or even ... bash$ test [1=2] echo yes yes -- Bob -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
Re: vpath search failures in Clearcase when files do actually exist
Pierre A. Humblet schrieb: On Thu, Nov 11, 2004 at 09:33:21AM +0100, Reini Urban wrote: BTW: Can someone provide testcode to identify if the current drive as MVFS please. At least I think that a clearcase view is mounted as driveletter. Does it show up as GetLogicalDrives()? (should so) In detail, what does GetVolumeInformation().lpFileSystemNameBuffer return for such a MVFS drive? Here is the output of GetVolumeInformation: y:\ Ret 1 Volume: CCase serial: 36984713 component: 255 flags: 3 name: MVFS HTH Yes, it helped. I thought it should be MVFS, but now I'm sure. -- Reini Urban http://xarch.tu-graz.ac.at/home/rurban/ -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
