Re: Possible bug with mmap on XP?
On Fri, 3 Feb 2006, Christopher Faylor wrote: ... One person even reported that a Cygwin bug caused OS wrapping and he found himself running FreeBSD... Is that the Blue Screen Of Life? Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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/
1.5.19+: symlink bug
Creating links with the same name, but with and without a .exe extension succeeds, but the one with no extension is later ignored. Here's a minimal example: $ ls -l TESTLINK* ls: TESTLINK*: No such file or directory $ ln -s /bin/ls.exe TESTLINK $ ls -l TESTLINK* lrwxrwxrwx 1 igor root 11 Feb 3 09:25 TESTLINK - /bin/ls.exe $ ln -s /bin/echo.exe TESTLINK.exe $ ls -l TESTLINK* lrwxrwxrwx 1 igor root 13 Feb 3 09:25 TESTLINK - /bin/echo.exe lrwxrwxrwx 1 igor root 13 Feb 3 09:25 TESTLINK.exe - /bin/echo.exe $ uname -svr CYGWIN_NT-5.1 1.5.20(0.152/4/2) 2006-01-31 21:09 $ cygcheck -f /bin/ln coreutils-5.3.0-9 $ Let me know if more detail is needed to reproduce this, and I'll attach the cygcheck output, or an strace if needed. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Running bash from cron
On Fri, 3 Feb 2006, Cygnut (sent by Nabble.com) wrote: http://cygwin.com/acronyms/#PCYMTWLL. Thanks. I have had much success running cygwin bash scripts from cron (Kalab's crons.exe), however on new machines I install cygwin on, the bash scripts don't run at all. the scripts are run from crons like 12 12 * * * * c:\cygwin\bin\bash c:\cygwin\tmp\mybash.sh If you're having a problem with a third-party cron daemon, you'd be better off asking their support forum for help. This list only supports official Cygwin packages. If you can demonstrate that there is a problem running bash from a regular cmd.exe window, we'd be interested in hearing it. However, your command above works just fine for me. Is there some enviornment setup I am missing ? This is probably specific to the third-party cron, and thus is off-topic for this list. We just wouldn't know. If I use cygwin's cron would I fare better ? Most likely yes (though you'd have to fall into the habit of using POSIX paths, properly redirecting input and output, etc). Plus, if you use Cygwin cron and have the same problem, we'd be glad to help. :-) Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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/
run-1.1.6-1: argument limit problem
Hi, I use a rather long (many-argument) rxvt command line in one of my shortcuts, and am apparently hitting the arbitrary 20-argument limit that run imposes. Can this limit please be increased? I'll look into submitting a patch that does away with the fixed limit on the number of arguments, but can't promise a timeframe. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Running bash from cron
On Fri, 3 Feb 2006, Cygnut (sent by Nabble.com) wrote: Thanks Igor. I already uninstalled crons and did this cygrunsrv -I cron -p /usr/sbin/cron -a -D cygrunsrv -S cron Why not /usr/bin/cron-config? I put a simple script in there like this 51 * * * * /tmp/doit.sh You forgot to redirect I/O. If this script produces any output (on either stdout or stderr), cron will likely try to mail it to you, and fail, since you probably don't have an smtp daemon running. ran crontab `cat the above file` crontab -l yields above net stop cron net start cron i realize this may not be required, but I am trying everything before writing The two lines above are the same as cygrunsrv -E cron; cygrunsrv -S cron. doit.sh is #!/bin/bash PATH=/usr/bin:/bin echo OK /tmp/OK at 52 minutes no /tmp/OK file Try also adding /bin/date /tmp/OK (just to make sure you know *when* the output was produced). /var/log/cron.log just says cron started Did you also look in the Windows Event Log? How can I troubleshoot this ? I really need it to work or I have to write my program in f*cking vbscript, which I hate. Calm down, take a deep breath. Did you run cron_diagnose.sh? Please read and follow the directions at Problem reports: http://cygwin.com/problems.html especially the part about attaching (as an uncompressed text *attachment*) the output of cygcheck -svr on your system. That will give us more information to work with -- there are just too many things that can produce the results you describe. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: strange cygstart bug with current Cygwin versions
On Fri, 3 Feb 2006, David Picton wrote: I have encountered a strange bug when starting Microsoft Word when it is started by the cygstart command, e.g. cygstart Index.doc, with the current version of the Cygwin dll. The symptoms are as follows: 1. Word appears to start normally, and the file can be edited on screen. 2. However, when Word should be autosaving, a message appears to the effect that the auto-recovery save has been postponed. 3. Attempting to save the file gets no response. The only way to close the window is to exit without saving! To attempt to reproduce the bug, you need to open a Word document using cygstart, change the document, then try to save it. I have seen this bug on two of the three computers which I use - all with the latest release of Cygwin. Reverting to the previous version doesn't help, but the bug goes away if I restore a pre-January version of Cygwin. It also goes away if I invoke the Word binary (winword.exe) directly from the command line or a shell script. Sounds like an instance of http://cygwin.com/ml/cygwin/2005-05/msg00587.html... Does it work if you cygstart cmd and then start word from that cmd shell? If you get the same symptoms, please run set in that cmd window and compare the output with the same in a cmd started via Start-Run... Do you get the same problem if you use run winword.exe filename? Which version of Word are you trying to use? I know that Word uses /tmp for temporary files when started via Cygwin. Check permissions on your /tmp. Otherwise, we'll need more details, as I, for one, can't reproduce your problem. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Running bash from cron
On Fri, 3 Feb 2006, Cygnut (sent by Nabble.com) wrote: Again, http://cygwin.com/acronyms/#PCYMTWLL. Thanks. ok, I spoke too soon, it runs but when I log off it doesn't. I am admin level. There are no logs. What can I check ? Check here: Problem reports: http://cygwin.com/problems.html You might also run cron_diagnose.sh and see what that says. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: cvs is broken/cygwin-bug in mkdir()?
On Sat, 4 Feb 2006, Sven Köhler wrote: i'm trying to use cygwin's cvs in server-mode for another application I get the following error, but on a linux-system, everything is fine! E cannot mkdir /tmp/cvs-serv784/. error No such file or directory As Brian said, this was already reported. The ball is now in the CVS maintainer's court: http://cygwin.com/ml/cygwin/2006-01/msg01383.html. Of course, cvs isn't using the mkdir command i guess, but it's using some mkdir()-call and i guess that cygwin returns the wrong return-code and actually cvs would ignore the file exists error-code, but it gets the no such file or directory error-code. So that results in a broken cvs :-( That's exactly right (see line 1002 of server.c). Is this a bug in cygwin? Yes. Looks like Cygwin is too hasty in assigning the error number: Linux only returns ENOENT if the directory doesn't already exist, but Cygwin will always return it for a trailing dot argument. The same with rmdir, where it would always return EINVAL, even if the directory doesn't exist (in which case Linux returns ENOENT). FWIW, POSIX only specifies an action for rmdir() on a trailing dot (EINVAL). I didn't post earlier because I was trying to test a patch (a variant of CGF's, though his is cleaner). I'm glad this is fixed. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: cvs is broken/cygwin-bug in mkdir()?
On Fri, 3 Feb 2006, Christopher Faylor wrote: On Sat, Feb 04, 2006 at 05:11:08AM +0100, Sven K?hler wrote: As i demonstrated, my Linux returns the error file already exists instead of cygwin's no such file or directory. So the problem is perhaps gone, if cygwin becomes more Linux-like once again ;-) I just checked in a fix to make cygwin return the correct error when trying to create a file with a trailing dot. It will be in the next snapshot. Thanks for doing this. Care to also fix rmdir() to return ENOENT if the directory doesn't exist (see http://cygwin.com/ml/cygwin/2006-02/msg00172.html)? Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Exec and parent environment [attn tcltk maintainer]
On Thu, 2 Feb 2006, Eric Blake wrote: Ugh - top-posting reformatted. http://cygwin.com/acronyms/#TOFU Good point -- I should start including the link to that acronym too... :-) When I type env the environment appears fine. However, when I start a tclsh and type puts [ exec sh -c env ] the environment is almost empty - this is new behaviour, the whole environment got fully passed through until recently. Sounds like YA case of http://sourceware.org/ml/cygwin/2006-01/msg00938.html. tclsh is one of the few programs in /bin that does not link against cygwin1.dll, but uses Windows directly. So it probably needs to be taught how to interact with cygwin's updated environment scheme. Thanks for the feedback. So does this mean that it is the tcl that is bundled with Cygwin that is broken? Actually, on further thought, it might be something that cygwin needs to fix. Do you mount /bin as cygexec? (Hint - attaching the output of cygcheck -svr as a text attachment would have answered this question). Prior to cygwin 1.5.19, it was suggested that mounting /bin as cygexec would speed up operations of cygwin programs, since the bulk of the files in /bin are linked against cygwin1.dll. In 1.5.19, however, cgf improved cygwin to automatically detect wheter an application is linked against cygwin or is a native windows program, and in the process made it so that only native windows programs are handed a native windows environment. I've been thinking along the same lines. Actually, as of Cygwin 1.5.19, any program that links to Cygwin1.dll will behave as if it's on a cygexec mount... I wonder if there needs to be a symmetric option to mount to indicate a non-Cygwin executable... But if cygexec mounting is turned on, then a native windows program in that mount point will be mistakenly treated as a cygwin program, such that cygwin tries to use only the cygwin method for passing the environment to the child program. My guess is that you are calling tclsh from a cygwin shell (bash?), and that cygwin is assuming the child (tclsh) can understand cygwin's environment because it lives in /bin, even though it is a windows native executable and needs the windows environment. Well, tclsh is a *Cygwin* executable that uses native Windows calls (CreateProcess) under the covers (in tcl84.dll). I'm not sure if it would even help to be able to mount it as a non-cygwin executable... CGF? From there, when you do the exec in tclsh, the grandchild sh sees the same environment that tclsh saw. Is there an easy workaround? If my hypothesis was correct, then remount /bin to no longer be cygexec (cygexec mount points no longer make sense, now that cygwin can autodetect programs linked against cygwin). Or wait and see if a future snapshot of cygwin learns to recognize native windows apps that live in a cygexec mountpoint. Since Cygwin automatically detects Cygwin programs, removing the cygexec bit won't help. I guess there isn't an easy workaround at the moment, short of patching tclsh appropriately and building your own. That might not be as hard as it sounds, and I'm sure the maintainer will appreciate a ready and tested patch. Or, if you are ambitious, figure out a way to package two versions of tcltk in cygwin; the native tcltk is a requirement (cgf wants the insight debugger to work no matter what), but a cygwin-aware tcltk would also be nice. The issue has come up before, search the mailing list archives. Until someone comes up with a good solution, tclsh will remain one of the few windows-native programs distributed in /bin. :-) Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Exec and parent environment [attn tcltk maintainer]
On Thu, 2 Feb 2006, Eric Blake wrote: But if cygexec mounting is turned on, then a native windows program in that mount point will be mistakenly treated as a cygwin program, such that cygwin tries to use only the cygwin method for passing the environment to the child program. My guess is that you are calling tclsh from a cygwin shell (bash?), and that cygwin is assuming the child (tclsh) can understand cygwin's environment because it lives in /bin, even though it is a windows native executable and needs the windows environment. Well, tclsh is a *Cygwin* executable that uses native Windows calls (CreateProcess) under the covers (in tcl84.dll). I'm not sure if it would even help to be able to mount it as a non-cygwin executable... CGF? Really? $ cygcheck tclsh Found: C:\cygwin\bin\tclsh.exe C:/cygwin/bin/tclsh.exe $ My understanding of the cygexec magic was that if it was determined that an .exe had a startup dependency on cygwin1.dll, but tclsh has NO dll dependencies at startup (any dlls are pulled in with dynamic loads during execution), so how would cygwin1.dll be able to think that tclsh is a cygwin executable? Because tclsh.exe is a symlink to tclsh84.exe, which has $ ls -l /bin/tclsh.exe lrwxrwxrwx 1 Administrators SYSTEM 11 Jul 30 2004 /bin/tclsh.exe - tclsh84.exe* $ cygcheck tclsh84 Found: C:\cygwin\bin\tclsh84.exe C:/cygwin/bin/tclsh84.exe C:\WINDOWS\System32\KERNEL32.dll C:\WINDOWS\System32\ntdll.dll C:\cygwin\bin\cygwin1.dll C:\WINDOWS\System32\ADVAPI32.DLL C:\WINDOWS\System32\RPCRT4.dll C:\cygwin\bin\tcl84.dll C:\WINDOWS\System32\USER32.dll C:\WINDOWS\System32\GDI32.dll I made the same mistake initially... :-) Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Exec and parent environment [attn tcltk maintainer]
On Thu, 2 Feb 2006, Christopher Faylor wrote: On Thu, Feb 02, 2006 at 05:04:42PM +, Eric Blake wrote: $ cygcheck tclsh Found: C:\cygwin\bin\tclsh.exe C:/cygwin/bin/tclsh.exe $ bash$ ls -l /bin/tclsh* lrwxrwxrwx 1 cgf None11 Mar 15 2005 /bin/tclsh.exe - tclsh84.exe -rwxrwxrwx 1 cgf None 16384 Sep 1 2003 /bin/tclsh84.exe bash$ cygcheck /bin/tclsh84.exe d:/cygwin/bin/tclsh84.exe L:\WINXP\system32\KERNEL32.dll L:\WINXP\system32\ntdll.dll d:/cygwin/bin\cygwin1.dll L:\WINXP\system32\ADVAPI32.DLL L:\WINXP\system32\RPCRT4.dll d:/cygwin/bin\tcl84.dll L:\WINXP\system32\USER32.dll L:\WINXP\system32\GDI32.dll I will look into fixing this. cgf Whatever fix is made, it would be great if cygcheck output the path of symlinks it had to chase to get the executable. Also, outputting Not an executable for scripts and such would help. BTW: $ cygcheck vi Error: could not find vi $ ls -l /usr/bin/vi lrwxrwxrwx 1 igor Users 7 Dec 20 18:21 /usr/bin/vi - vim.exe $ cygcheck awk Found: C:\cygwin\bin\awk.exe C:/cygwin/bin/awk.exe $ ls -l /usr/bin/awk.exe lrwxrwxrwx 1 igor Users 8 Nov 17 18:25 /usr/bin/awk.exe - gawk.exe Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Exec and parent environment [attn tcltk maintainer]
On Thu, 2 Feb 2006, Christopher Faylor wrote: On Thu, Feb 02, 2006 at 12:27:53PM -0500, Igor Peshansky wrote: On Thu, 2 Feb 2006, Christopher Faylor wrote: On Thu, Feb 02, 2006 at 05:04:42PM +, Eric Blake wrote: $ cygcheck tclsh Found: C:\cygwin\bin\tclsh.exe C:/cygwin/bin/tclsh.exe $ bash$ ls -l /bin/tclsh* lrwxrwxrwx 1 cgf None11 Mar 15 2005 /bin/tclsh.exe - tclsh84.exe -rwxrwxrwx 1 cgf None 16384 Sep 1 2003 /bin/tclsh84.exe bash$ cygcheck /bin/tclsh84.exe d:/cygwin/bin/tclsh84.exe L:\WINXP\system32\KERNEL32.dll L:\WINXP\system32\ntdll.dll d:/cygwin/bin\cygwin1.dll L:\WINXP\system32\ADVAPI32.DLL L:\WINXP\system32\RPCRT4.dll d:/cygwin/bin\tcl84.dll L:\WINXP\system32\USER32.dll L:\WINXP\system32\GDI32.dll I will look into fixing this. Whatever fix is made, it would be great if cygcheck output the path of symlinks it had to chase to get the executable. Also, outputting Not an executable for scripts and such would help. I will not look into fixing this. I will look into fixing this. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: mismatched dll
On Thu, 2 Feb 2006, Christopher Faylor wrote: On Thu, Feb 02, 2006 at 09:58:17PM +, Eric Blake wrote: [snip a nice summation of the problems with mismatched dlls along with a cogent suggestion for dealing with the problem] Can I get a gold star here? Oh, you mean for saying that anything in computer science is [possible], if you are willing to write the code and wait long enough? :-) Done. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: g-b-s patch: upstream patch list
On Wed, 1 Feb 2006, Eric Blake wrote: The uncontroversial bits of my patch: 2006-02-01 Eric Blake ebb9atbyu.net * templates/generic-build-script: Add comments throughout. (usage): Document --logs option. (install): manifest.lst is not executable. Applied with (very) minor tweaks. Thanks! Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac
Re: cygwin.bat improvements (was Re: Cygwin on XP disapearing terminal mystery)
On Wed, 1 Feb 2006, Christopher Faylor wrote: On Tue, Jan 31, 2006 at 08:16:15PM -0800, Brian Dessent wrote: Christopher Faylor wrote: Couldn't it just replace the bash line with: if exist bash.exe goto ok echo bash.exe is missing - please install it pause :ok bash --login -i Well that would certainly work for the case of a missing bash. I guess I was thinking that the marker file method would work for a wider range of brokenness where for whatever reason the shell fails to start and/or complete its initialization. Having never experienced this personally I can't really come up with a set of possible scenarios of brokenness, but it seems we get do get a lot of bash window flashes briefly and then disappears reports. Are they all just missing bash.exe? That has always been my impression. Maybe I'm wrong. OTOH, would people complain bitterly if we just put a pause at the end of cygwin.bat. That would be a very minimal change. I would. I don't use cygwin.bat so I can't say how aggravating this would be. Very. This would mean an extra keystroke for every exited bash shell... Yuck. However, most bash invocations should exit with a 0 exit code. So, why not simply do something like the test below? if not errorlevel 1 goto nopause pause :nopause Also, command not found sets error code to 127, so the 1 above can be changed to 127. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: cygwin.bat improvements (was Re: Cygwin on XP disapearing terminal mystery)
On Wed, 1 Feb 2006, Christopher Faylor wrote: On Wed, Feb 01, 2006 at 12:06:08PM -0500, Igor Peshansky wrote: However, most bash invocations should exit with a 0 exit code. So, why not simply do something like the test below? if not errorlevel 1 goto nopause pause :nopause Also, command not found sets error code to 127, so the 1 above can be changed to 127. (actually Dave's right -- it's 9009). Wouldn't that be: if not errorlevel 0 pause Contrary to all common sense, if not errorlevel A means if %ERRORLEVEL% A, not an equality test... So the above will *always* pause. Also, in command.com (Win9x), I believe you can only have a goto after the if errorlevel test... or @if errorlevel 0 goto exit @echo bash was either not executed or did not exit cleanly @echo [more touchy feely stuff here] @pause :exit ? Again, if errorlevel 0 means 0 and above. Yuck! This would be triggered if someone types exit N (where N != 0) in the bash window, though. I suspect that we'll now hear from the cygwin user contingent who always type exit 42 because that was required to work around problems with ZoneAlarm v527. Since exit N from bash will never produce an exit code that's over 256, checking for %ERRORLEVEL%=256 should work well to diagnose bash crashing problems. So, do we go with if not errorlevel 256 goto nopause pause :nopause ? Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: cygwin.bat improvements (was Re: Cygwin on XP disapearing terminal mystery)
On Wed, 1 Feb 2006, Igor Peshansky wrote: On Wed, 1 Feb 2006, Christopher Faylor wrote: Wouldn't that be: if not errorlevel 0 pause Contrary to all common sense, if not errorlevel A means if %ERRORLEVEL% A, not an equality test... So the above will *always* pause. Also, in s/always/never/. D'oh. command.com (Win9x), I believe you can only have a goto after the if errorlevel test... Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: cygwin.bat improvements (was Re: Cygwin on XP disapearing terminal mystery)
On Wed, 1 Feb 2006, Christopher Faylor wrote: On Wed, Feb 01, 2006 at 01:31:08PM -0500, Igor Peshansky wrote: On Wed, 1 Feb 2006, Christopher Faylor wrote: On Wed, Feb 01, 2006 at 12:06:08PM -0500, Igor Peshansky wrote: However, most bash invocations should exit with a 0 exit code. So, why not simply do something like the test below? if not errorlevel 1 goto nopause pause :nopause Also, command not found sets error code to 127, so the 1 above can be changed to 127. (actually Dave's right -- it's 9009). Wouldn't that be: if not errorlevel 0 pause Contrary to all common sense, if not errorlevel A means if %ERRORLEVEL% A, not an equality test... So the above will *always* pause. Also, in command.com (Win9x), I believe you can only have a goto after the if errorlevel test... Ok. I seem to vaguely recall this from MS-DOS days but I don't ever recall that only goto is allowed. I just tried this on Windows 98 and it is possible to put something besides a 'goto' in the if clause. Heh, you're right, and my belief above was wrong. However, in command.com, you definitely cannot put more than one command after the if, so if you want to print the touchy feely stuff, you'll need a goto. Can anyone confirm/deny this on Windows 95? command.com should be the same, FWIW. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Any way to get cpu usage % of a process in Cygwin/windows on the command line?
On Wed, 1 Feb 2006, Karr, David wrote: Is there any way to get the CPU usage percentage of a process in Cygwin or Windows, in a script? The ps command in Cygwin is based on SYSV ps, which doesn't appear to show the CPU usage (whereas BSD ps does). Use procps. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: RFC: [PATCH] Simplify installed.db format (take 3)
On Mon, 23 Jan 2006, Igor Peshansky wrote: : On Sun, 22 Jan 2006, Igor Peshansky wrote: : The attached patch fixes the parsing of installed.db package lines. : Comments welcome. : : 2006-01-23 Igor Peshansky [EMAIL PROTECTED] : : * cygpackage.h (filesize): Remove unused field. : (cygpackage::createInstance): Ditto. : * cygpackage.cc (cygpackage::createInstance): Ditto. : * package_db.cc (packagedb::packagedb): Remove unused variables. : Simplify parsing of installed.db. In fact, let's use the attached patch instead. Ping? Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac
Re: [PATCH] setup: initially expand categories starting with .
On Sun, 22 Jan 2006, Igor Peshansky wrote: On Thu, 17 Nov 2005, Igor Pechtchanski wrote: (see http://cygwin.com/ml/cygwin-apps/2005-11/msg00141.html) 2005-11-16 Igor Peshansky [EMAIL PROTECTED] * PickView.cc (PickView::setViewMode): Auto-expand category if its name starts with a dot. Ping. YA ping. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac
Re: [PATCH] setup: Handle the package validation exception
On Mon, 23 Jan 2006, Igor Peshansky wrote: In the scenarios you've given above, if network install is selected and the locally cached copy is corrupt, it will be re-downloaded. In case of local install, the corrupt copy will result in the same behavior as not having the package on your machine. Frankly, I fail to see the problem... Perhaps I'm missing something. Ping? Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac
Re: [PATCH] setup: fix abnormal exit test for postinstall scripts
On Mon, 23 Jan 2006, Igor Peshansky wrote: Just noticed that the postinstall mechanism in setup inverts the test for abnormal exit. The attached one-liner patch fixes that. ChangeLog below. Dave's advice in http://cygwin.com/ml/cygwin/2006-01/msg01475.html, made me realize that setup will always rename the script to .done, even if it didn't complete properly (i.e., exited with a non-zero exit code). Now that we detect the code correctly, let's also use that to not rename failed scripts for easier detection (patch attached). ChangeLog below. Note that this will put a bit of extra burden on package maintainers to ensure that their scripts always return a zero exit code. Most scripts already do that, though. Igor == 2006-01-31 Igor Peshansky [EMAIL PROTECTED] * script.cc (Script::run): Fix inverted test for abnormal exit. Do not rename to .done unless completed successfully. -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de BergeracIndex: script.cc === RCS file: /cvs/cygwin-apps/setup/script.cc,v retrieving revision 2.21 diff -u -p -r2.21 script.cc --- script.cc 14 Oct 2005 04:00:50 - 2.21 +++ script.cc 31 Jan 2006 14:39:40 - @@ -243,14 +243,15 @@ Script::run() const if (to_log !file_out.isEmpty ()) log(LOG_BABBLE) file_out endLog; - if (!retval) + if (retval) log(LOG_PLAIN) abnormal exit: exit code= retval endLog;; /* if file exists then delete it otherwise just ignore no file error */ io_stream::remove (String (cygfile://) + scriptName + .done); - io_stream::move (String (cygfile://) + scriptName, - String (cygfile://) + scriptName + .done); + if (!retval) +io_stream::move (String (cygfile://) + scriptName, + String (cygfile://) + scriptName + .done); return retval; }
RE: 1.5.19-4 exec family of functions find wrong file to execute
On Tue, 31 Jan 2006, Dave Korn wrote: On 30 January 2006 19:23, Martin wrote: Any response? TIA --- Martin [EMAIL PROTECTED] wrote: I am attempting to invoke a command with execvp/execlp. If a file appears in my PATH before the executable desired and has the same name as the executable, the first occurrence of the file name is used as the executable to invoke. Even though the first file is NOT marked as executable. Is this normal behavior for execlp/execvp? Shouldn't the execution permission be set in order to execute it? Depends on your mountpoints. If the file is found under a mountpoint with the cygexec flag set, it will be assumed to be executable regardless of the x bit in the file's perms. Also, on FAT filesystems, there are no permission bits[*]... Your file may be tripping the executable file detection heuristics, though. Igor [*] Cygwin can simulate them for you on pure FAT with ntea, but with FAT32 you're SOL. -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: 1.5.19-4 exec family of functions find wrong file to execute
Ugh, top-posting... Reformatted. On Tue, 31 Jan 2006, Martin wrote: --- Igor Peshansky [EMAIL PROTECTED] wrote: http://cygwin.com/acronyms/#PCYMTNQREAIYR. Thanks. On Tue, 31 Jan 2006, Dave Korn wrote: On 30 January 2006 19:23, Martin wrote: Any response? TIA --- Martin [EMAIL PROTECTED] wrote: I am attempting to invoke a command with execvp/execlp. If a file appears in my PATH before the executable desired and has the same name as the executable, the first occurrence of the file name is used as the executable to invoke. Even though the first file is NOT marked as executable. Is this normal behavior for execlp/execvp? Shouldn't the execution permission be set in order to execute it? Depends on your mountpoints. If the file is found under a mountpoint with the cygexec flag set, it will be assumed to be executable regardless of the x bit in the file's perms. Also, on FAT filesystems, there are no permission bits[*]... Your file may be tripping the executable file detection heuristics, though. Igor [*] Cygwin can simulate them for you on pure FAT with ntea, but with FAT32 you're SOL. In response to your comments: I ran the testcase from ~/testcases directory. I believe my cygwin.out file indicates /home is mounted from D:\home (HOME=/home/test). Also, D: is NTFS. Also, there is nothing special about mount point (no -x flag was used). This is all shown in cygwin.out. Then you'll just have to run this under strace, and see why it selects that file to execute. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Installation Errors
Ugh, top-posting... Reformatted. On Tue, 31 Jan 2006, jgriffin wrote: On Mon, 30 Jan 2006 21:19:01 -0700, Eric Blake wrote According to jgriffin on 1/30/2006 11:29 AM: I'm trying to install Cygwin on a stand alone computer. Here is what I'm doing: Most of this looked okay. However, you probably installed without wiping out the previous installation, or with a second cygwin1.dll somewhere in the path. Attaching the output of cygcheck -svr, as a text attachment, would help. Errors Most of the errors are coming from the post-install scripts. I get an error dialog stating that cygpopt-0.dll could not be found. Somehow, you missed a dependency when building your cd. Without all the required packages, you can't expect the installation to work properly. Try again (also, you can use cygcheck -p cygpopt-0.dll to find out what package you missed). Eric, I appreciate your response. When I run setup.exe, I'm downloading to a specfic folder. I have installed Cygwin numerous times this way and never had any problems. The one thing I'm doing different is burning the packages (along with the setup.exe app) to a CD. So, in my mind, I need to ensure that the cygpopt-0.dll is being transfered 1) from the mirror site to the harddrive and then 2) from the hard drive to the CD - because I know its not being installed. One thing to keep in mind (and which people trying to make Cygwin CDs using setup.exe keep forgetting) is that setup.exe is not a mirroring tool. It will not download packages that are already installed on the system, even in download mode. So, if you have the package that contains cygpopt-0.dll (see below) installed on your system, but the tarball for it is somehow missing from your package cache, it will not be re-downloaded. If you want to make setup download stuff, move your Cygwin directory out of the way before downloading (you can move it back afterwards without any loss of information -- just delete the newly-created Cygwin root folder). What package contains the dll? As Eric said, cygcheck -p should work for a sufficiently recent version of Cygwin (1.5.19+). Otherwise, there's always the Cygwin package search page at http://cygwin.com/packages/. I did run cygcheck -svr and compared it to a working cygwin install - nothing jumped out ( I can't post the file for other reasons). Unless you store plaintext passwords in the environment, nothing in the cygcheck output should be a security hole. However, we should probably have an FAQ page on what can be sanitized in the cygcheck output and how to sanitize it without losing information... Any volunteers? :-) The -p option is not recogized by cygcheck. It is for Cygwin 1.5.19 or newer. HTH, Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: g-b-s patch: upstream patch list
On Mon, 30 Jan 2006, Eric Blake wrote:
I finally made my upstream patching generic, following Igor's suggestion
of providing foo-ver-rel.patch.tar.bz2 if there is more than one patch
file being applied to foo-ver.tar.*. I used this patch on
readline-5.1-2, and have been testing it on my (soon-to-be-released)
bash-3.1-2. The idea is that you create a single file,
CYGWIN-PATCHES/upstream_patches.lst, which is a listing of pathnames
relative to ${srcdir} containing upstream patches. Then g-b-s can use
the contents of that file to manipulate a .patches directory to
create/unpack the patch tarball.
Heh, this is cool, thanks for doing this. I was working on the same kind
of functionality for pdksh over the weekend, and also have a whole bunch
of changes to the g-b-s, with essentially a very similar approach. Some
of your stuff looks cleaner, while some of mine may be more general --
I'll need to review and compare these in more detail. Some initial
comments on your patch below.
2006-01-30 Eric Blake ebb9atbyu.net
* templates/generic-build-script: Add ability to apply upstream
patches, listed in CYGWIN-PATCHES/upstream_patches.lst.
(mkdirs): Use new .patches directory.
(fixup): New function.
(prep): Unpack patch tarball.
(mkpatch): Apply upstream patches before doing diff.
(spkg): Make patch tarball if needed.
(checksig): Look at patch tarball sig.
(install): manifest.lst is not executable.
First off, please, please don't bundle unrelated changes into the same
patch. I like the idea of the leading comments, and of fixing up the mode
on manifest.lst, but could that be a separate patch, with a separate
ChangeLog? That makes it much easier to review the changes.
FWIW, in the comments, I'd replace g-b-s by the script everywhere,
since they will also appear in the actual build scripts. Also, I think
it's worth mentioning that depend() only lists the *static* DLL
dependencies.
Now, for the upstream patches functionality, I think it would be easier if
I actually posted my changes and we could compare on-list. I'll clean
them up and post them in the next couple of days.
Igor
--
http://cs.nyu.edu/~pechtcha/
|\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED]
ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!)
|,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski
'---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow!
Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte.
But no -- you are no fool; you call yourself a fool, there's proof enough in
that! -- Rostand, Cyrano de Bergerac
Re: g-b-s patch: upstream patch list
On Mon, 30 Jan 2006, Yitzchak Scott-Thoennes wrote: On Mon, Jan 30, 2006 at 10:45:10AM -0500, Igor Peshansky wrote: 2006-01-30 Eric Blake ebb9atbyu.net * templates/generic-build-script: Add ability to apply upstream patches, listed in CYGWIN-PATCHES/upstream_patches.lst. Now, for the upstream patches functionality, I think it would be easier if I actually posted my changes and we could compare on-list. I'll clean them up and post them in the next couple of days. FWIW, when I looked at this, it looked easiest just to apply the patches in unpack(). Maybe, but I'd rather keep unpack() as it is -- i.e., a helper function to replace an unconditional tar xf. That way, if a package has non-tar.gz source archives, the maintainers don't have to rack their brains to figure out how to edit unpack(). Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac
Re: [maybe-ITP] gamin
On Mon, 30 Jan 2006, Yaakov S (Cygwin Ports) wrote: Lapo Luchini wrote: I needed this patch in order to have it working on FAT32 disks, but both on FAT and NTFS the test results seem quite consistent to me: all is good on NTFS and all-but-test-9 on FAT. I'd rather not outright #ifndef __CYGWIN__ these if there's a better solution. I don't know if there's a way to determine if the drive is FAT or NTFS, I'm sure there is -- Corinna posted a test program to cygwin@ earlier this month that prints various attributes of the drives. It was meant for shares, but I'm sure it also applies to local drives. Besides, cygcheck prints out what kind of drive it is, so you can get some code from there too. but AFAICS there is a way to determine if Windows is NT or not (and presume that NT is running on NTFS and not FAT, which should be the norm), for example: Umm, you don't want to assume that any disk is NTFS on NT... See below. http://cygnome2.sourceforge.net/install/release/nautilus/nautilus-2.4.2-cygwin.patch Although this raises an interesting question: Linux can also use FAT disks, so what happens then?? Exactly, and so can NT. Better to find out what disk we're working with. Test 9 checks if a directory mtime gets modified when you write a file in it and I guess FAT just doesn't work that way. Possibly, but I only have NTFS drives at my disposal. Floppies are usually FAT. How about using one? :-) HTH, Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac
Re: [maybe-ITP] gamin
On Mon, 30 Jan 2006, Christopher Faylor wrote: On Mon, Jan 30, 2006 at 08:29:05PM -0500, Igor Peshansky wrote: On Mon, 30 Jan 2006, Yaakov S (Cygwin Ports) wrote: Possibly, but I only have NTFS drives at my disposal. Floppies are usually FAT. How about using one? :-) FAT but not FAT32. There can be differences. Good point. Can you format a floppy as FAT32? According to http://support.microsoft.com/?kbid=253774, no: Floppy disks are formatted as FAT12, even if you have formatted your hard disk to FAT32. The utility, Format.com is prohibited from making a FAT32 file system on a floppy disk. It should be possible to create a small partition for testing, but it's probably much easier to do in VMware... Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac
Re: No Terminal Windows appear
On Mon, 30 Jan 2006, Juanjo Carmona-Serradilla wrote: After I install cygwin and type starx in bash, I do not get the terminal windows. See log below. Thanks in advance. I am running Windows XP professional. Juanjo Juanjo [EMAIL PROTECTED] ~ $ startx [snip] startx will only open a terminal if one is present in ~/.Xsession. Why not use startxwin.sh? Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://x.cygwin.com/docs/ FAQ: http://x.cygwin.com/docs/faq/
Re: 1.5.19: pdksh 5.2.14-3 tab-complete and shell metacharacters
On Thu, 26 Jan 2006, Svend Sorensen wrote: When I use pdksh to expand the name of a file with spaces or other shell metachars in it, the filename is expanded without escaping the metacharacters. Thanks for the report. I've been able to reproduce this. There's an upstream patch that fixes the issue, and I'm currently working on integrating that and other patches into the build to produce a new release (but no ETA yet). The pdksh version strings are identical. This is unfortunate. What is the patchlevel shown for the pdksh package on NetBSD? FWIW, I will include the full list of patches in the README for the new release. Igor Peshansky, volunteer PDKSH maintainer for Cygwin -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Cygwin Bash window disappear?!
On Mon, 30 Jan 2006, Dave Korn wrote: On 28 January 2006 04:19, Daniel mark wrote: Hello Eric: // Cygwin Configuration Diagnostics Path: C:\Program Files\texmf\miktex\bin Path = 'C:\Program Files\texmf\miktex\bin;C:\WINDOWS\system32;C:\WINDOWS;C:\WINDOWS\System32\Wbem; First thing to try is to get rid of all those clashing win32-versions of *nix apps from your %PATH% and re-run setup. Huh? Since when does MikTeX conflict with Cygwin? I wonder if cygcheck should add c:\cygwin\bin to the PATH if it isn't there (e.g., when run from CMD and /bin isn't in the system PATH), just so we know if the Not Found: sh message means that the bash postinstall didn't complete, or simply that /bin isn't in the PATH. We still haven't gotten the exact error message that the OP gets when running c:\cygwin\Cygwin.bat from a CMD window. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Cygwin Bash window disappear?!
On Mon, 30 Jan 2006, Dave Korn wrote: On 30 January 2006 15:54, Igor Peshansky wrote: On Mon, 30 Jan 2006, Dave Korn wrote: On 28 January 2006 04:19, Daniel mark wrote: Hello Eric: // Cygwin Configuration Diagnostics Path: C:\Program Files\texmf\miktex\bin Path = 'C:\Program Files\texmf\miktex\bin;C:\WINDOWS\system32;C:\WINDOWS;C:\WINDOWS\Sys tem32\Wbem; First thing to try is to get rid of all those clashing win32-versions of *nix apps from your %PATH% and re-run setup. Huh? Since when does MikTeX conflict with Cygwin? I could see the included makeinfo interfering with /etc/postinstall/update-info-dir.sh. And people are always complaining about seeming-hangs in setup at the post-texmf.sh script. update-info-dir.sh doesn't use makeinfo. Besides, postinstall scripts should not rely on the path containing /bin -- they should use explicit paths (or else set the PATH themselves). Plus, I believe setup prepends /bin to the PATH before running the scripts. post-texmf.sh regenerates all of the binary LaTeX format files -- it's *supposed* to take a long time, especially with slower disks. Perhaps it should be renamed to post-texmf-THIS-MAY-TAKE-A-WHILE.sh (so people at least get a visual clue in setup)? :-) We still haven't gotten the exact error message that the OP gets when running c:\cygwin\Cygwin.bat from a CMD window. Igor Yep, we don't have enough information to diagnose the problem yet, that suggestion above was only a suggestion for something to try. Well, let's hope this doesn't take the OP off on a tangent and make him forget to actually post some useful output. :-) Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: 1.5.19: pdksh 5.2.14-3 tab-complete and shell metacharacters
Ugh, top-posting... Reformatted. On Mon, 30 Jan 2006, Svend Sorensen wrote: On 1/30/06, Igor Peshansky [EMAIL PROTECTED] wrote: http://cygwin.com/acronyms/#PCYMTNQREAIYR. Thanks. On Thu, 26 Jan 2006, Svend Sorensen wrote: When I use pdksh to expand the name of a file with spaces or other shell metachars in it, the filename is expanded without escaping the metacharacters. Thanks for the report. I've been able to reproduce this. There's an upstream patch that fixes the issue, and I'm currently working on integrating that and other patches into the build to produce a new release (but no ETA yet). The pdksh version strings are identical. This is unfortunate. What is the patchlevel shown for the pdksh package on NetBSD? FWIW, I will include the full list of patches in the README for the new release. The NetBSD system I tested has pdksh-5.2.14nb1 installed. There are two patches in pkgsrc, one which removes a Makefile check for pdksh in the /etc/shells file, and one which removes a declaration of errorno. Other than that, it uses the vanilla pdksh-5.2.14.tar.gz sources. Hmm, this is extremely surprising, as 5.2.14 doesn't contain any code to quote shell metacharacters on tab completion. Are you sure those are vanilla sources? I know OpenBSD uses its own CVS repository for PDKSH -- does NetBSD do this too? Can you compare the tarball with ftp://ftp.cs.mun.ca/pub/pdksh/pdksh-5.2.14.tar.gz? I have attached the patches from pkgsrc. These patches don't seem to do anything interesting. The next release of PDKSH on Cygwin will be built with a metacharacter quoting patch (from Debian). I'd still like to investigate this to see why it works for you on NetBSD without that patch... Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: 1.5.19: pdksh 5.2.14-3 tab-complete and shell metacharacters
On Mon, 30 Jan 2006, Svend Sorensen wrote: On 1/30/06, Igor Peshansky [EMAIL PROTECTED] wrote: Again, http://cygwin.com/acronyms/#PCYMTNQREAIYR. Thanks. Ugh, top-posting... Reformatted. On Mon, 30 Jan 2006, Svend Sorensen wrote: On 1/30/06, Igor Peshansky [EMAIL PROTECTED] wrote: http://cygwin.com/acronyms/#PCYMTNQREAIYR. Thanks. On Thu, 26 Jan 2006, Svend Sorensen wrote: When I use pdksh to expand the name of a file with spaces or other shell metachars in it, the filename is expanded without escaping the metacharacters. Thanks for the report. I've been able to reproduce this. There's an upstream patch that fixes the issue, and I'm currently working on integrating that and other patches into the build to produce a new release (but no ETA yet). The pdksh version strings are identical. This is unfortunate. What is the patchlevel shown for the pdksh package on NetBSD? FWIW, I will include the full list of patches in the README for the new release. The NetBSD system I tested has pdksh-5.2.14nb1 installed. There are two patches in pkgsrc, one which removes a Makefile check for pdksh in the /etc/shells file, and one which removes a declaration of errorno. Other than that, it uses the vanilla pdksh-5.2.14.tar.gz sources. Hmm, this is extremely surprising, as 5.2.14 doesn't contain any code to quote shell metacharacters on tab completion. Are you sure those are vanilla sources? I know OpenBSD uses its own CVS repository for PDKSH -- does NetBSD do this too? Can you compare the tarball with ftp://ftp.cs.mun.ca/pub/pdksh/pdksh-5.2.14.tar.gz? It turns out I was running NetBSD's base ksh, which must be a patched pdksh. Ah, I thought so. I tested the version of pdksh in pkgsrc, and it has no metacharacter expansion. The pkgsrc version of pdksh uses that source tarball you mentioned. Right. If you could get a pointer at the NetBSD CVS repository for pdksh, or at least a list of NetBSD-specific patches on top of the base sources, that would be great. I have attached the patches from pkgsrc. These patches don't seem to do anything interesting. The next release of PDKSH on Cygwin will be built with a metacharacter quoting patch (from Debian). I'd still like to investigate this to see why it works for you on NetBSD without that patch... NetBSD must patch their stock pdksh, which would make this a feature request, not a bug report. Nevertheless, I've been meaning to add upstream patches for a while... They address some very nasty bugs in the vanilla sources. Should NetBSD modify their base KSH_VERSION reporting to show it is patched? I'd say yes, if only to avoid the confusion you had at the start of this thread (with the identical version numbers). Thanks for your help, Igor. Don't thank me yet -- you'll probably be one of the people testing the new release. :-) Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: where is SCP
On Mon, 30 Jan 2006, Scott Purcell wrote: Hello, I need to scp some files to a remote machine on our network, and I cannot find the scp module? When I loaded cygwin, I loaded most default stuff, but even today, I am trying to search for it, and cannot find scp? Exactly where were you trying to search for it? The answer to most queries of where is program foo is the Cygwin package search page at http://cygwin.com/packages/ (also mentioned in the FAQ entry http://cygwin.com/faq/faq.setup.html#faq.setup.what-packages). You can use Perl regular expressions to get better matches for short names, e.g., \bscp\b for a whole-word scp search. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: MSYS path behavior in Cygwin
On Mon, 30 Jan 2006, Brian Dessent wrote:
tom wrote:
I guess the two will be mutually exclusive, unless you Cygwin devs see
some reason to merge some of Earnie's work. I'm have to believe
there's some reason why it hasn't been done though. I'm sure it's
non-trivial to say the least.
Thanks for the help anyways! Cygwin and msys make life in Win32land
bearable.
To do that would go against the grain of what Cygwin is trying to
accomplish: provide a full posix environment. So to have
argument-translation code sort of defeats that purpose since all Cygwin
programs are supposed to recognise posix paths.
Now in reality, everyone has to run a non-Cygwin program from time to
time, so of course there will be times where you run up against this.
But the question becomes, when should the library translate paths, and
when should it leave them alone? Because you can't just unconditionally
do it. The way MSYS handles this is by assuming that everything in the
MSYS bin directory is a MSYS binary that can handle posix paths, and
that everything else is a win32/mingw app that needs win32 paths. Now
that's a rather stark and arbitrary distinction. It works for MSYS
since it's intended to be a rather minimal system, just enough to build
packages using auto{conf,make},libtool. But I think for Cygwin this
would be way too restrictive.
Unfortunately, sometimes you can't do this even if you know that a Win32
program is being invoked. The problem is that an arbitrary string with a
/ in it may or may not be a path to a file -- and the only way to know
is to understand the semantics of the arguments. Since you don't always
know those semantics, doing this in general is impossible.
The workaround that I think most people use is a wrapper script that
essentially just runs cygpath -w on each argument and then execs $1.
So you can type wrapper win32program /posix/path/to/file and it ends
up running win32program c:/win32/path/to/file. If you do it right,
you can make this quite generic, so that you just prepend wrapper (or
whatever you want to call it.)
Below is one that I use in my .bashrc that calls perl. It's probably
got bugs, and it's not perfect -- for example if you pass something like
-I/usr/local/bin it will not know how to translate that.
[snip]
For more comprehensive behavior, you need to have program-specific wrapper
scripts. For example, see my Java wrappers in the cygwin-apps CVS:
http://cygwin.com/cgi-bin/cvsweb.cgi/wrappers/java/?cvsroot=cygwin-apps.
Notice that even those scripts don't do a complete job, since arguments
given to the Java programs themselves are left as-is (by design, because
those scripts can't just go about mangling arbitrary strings).
Igor
P.S. Is dodos short for extinct birds? :-)
--
http://cs.nyu.edu/~pechtcha/
|\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED]
ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!)
|,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski
'---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow!
Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte.
But no -- you are no fool; you call yourself a fool, there's proof enough in
that! -- Rostand, Cyrano de Bergerac
--
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: Heads-up: G-B-S changes (bash, logging)
On Sat, 28 Jan 2006, Yitzchak Scott-Thoennes wrote: On Sat, Jan 28, 2006 at 03:43:43PM -0500, Igor Peshansky wrote: I've just committed a change that turns off logging functionality in the generic-build-script by default. Um, why? Because it's still in the testing phase, and because not everyone wants logs in their packages (I don't, for example). Since I also added both an in-script and command-line control mechanism, it's easy enough to turn on. The decision to turn the logs off by default was made after I tried to use the g-b-s to build a package recently. If enough people complain, I'll add a symmetric mechanism for turning logging off and use that in my packages instead. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac
Re: Cygwin Bash window disappear?!
Ugh, top-posting... Reformatted. On Sun, 29 Jan 2006, Hammond, Robin-David%KB3IEN wrote: On Fri, 27 Jan 2006, Daniel mark wrote: Date: Fri, 27 Jan 2006 21:04:29 + (UTC) From: Daniel mark [EMAIL PROTECTED] To: [EMAIL PROTECTED] Daniel, http://cygwin.com/acronyms/#PCYMTNQREAIYR. Thanks. Subject: Cygwin Bash window disappear?! Hello all: I met a very strange problem as follows: After I finish installing the latest version of cygwin on my computer, I try to run C:\cygwin\cygwin.bat. The only thing I got is a black window splash for a second and then immediately disappear. I have fully installed all cygwin package and test all package without problem by using the package test tool provided by cygwin. I have tried to install it for 3 times!! still the same problem My machine is using AMD chip with XP professional I too have this problem, using win XP patch level 2. Try invoking sh, see if you can get a stable shell then. ls wont work (if your in the same boad i am) but `echo *` will have about the same effect sh in Cygwin is the same as bash, so that will have (almost) exactly the same behavior. You have a problem with your installation, whether you think setup completed successfully or not. Please read and follow the directions at Problem reports: http://cygwin.com/problems.html Please pay particular attention to the part that asks you to attach (as an uncompressed text *attachment*) the output of cygcheck -svr on your machine. Also, as others have said, open a CMD shell, run C:\cygwin\Cygwin.bat from there, and post any error messages you receive. HTH, Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: bizarre g++ behavior after reinstalling cygwin
On Sun, 29 Jan 2006, Vadim Oganesyan wrote: I just reinstalled cygwin on my laptop (running winXP) and g++ is acting most peculiar. The code that has been working fine for thepast 2~3 months no longer works (the programs are still working on other machines). the seem to compile OK, but don't generate output files, or generate nonsense, or get trapped in loops indefinitely. We'll need to see actual code snippets that misbehave before we can help. However, judging by your example below, I can guess that most of these are against the general C++ coding practices. GCC 3.4 is much stricter in that regard than previous versions, and does not necessarily try to make sense of such code, which is probably what's happening in your case. Did you try compiling with -Wall? Digging in I found that none of the i=i++ calls inside for-loops work. I have never heard of anything of sorts, still changes ++ to +1, this fixed the loops, but apparently there are still problems... i=i++ is either a no-op or nonsense, depending on the interpretation. You should either use i=i+1 or i++ (those should be equivalent in modern compilers). please help, have i lost my mind? That's for you to say, but I suspect your code would benefit from some review and rewrite. Google for C++ code style guide. HTH, Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_ [EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-'old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: bizarre g++ behavior after reinstalling cygwin
On Sun, 29 Jan 2006, Brett Serkez wrote: snip i=i++ is either a no-op or nonsense, depending on the interpretation. You should either use i=i+1 or i++ (those should be equivalent in modern compilers). The strick interpretation of i=i++ starts by evaluating the right side i++ to compute a r-value (right value). Since the ++ is after the i, it is a post-operation, vs. if it were before the i, in which case it would be a pre-operation. So first the value of i would be saved, then i incremented, then the saved value would be assigned to the l-value (left value) which in this case is i, setting i back to its original value. Yes, that was the nonsence interpretation. :-) You could use i=++i which would increment i, then save the r- value and assign to the l-value, but then again, i++ would be much simplier, no? Right. But there may be other style mishaps in Vadim's code that gcc 3.4 doesn't like. We just don't know. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: [ITP] bsdiff-4.3
On Sat, 28 Jan 2006, Brian Dessent wrote: Lapo Luchini wrote: http://cyberx.lapo.it/cygwin/bsdiff/ http://cyberx.lapo.it/cygwin/bsdiff/bsdiff-4.3-1-src.tar.bz2 http://cyberx.lapo.it/cygwin/bsdiff/bsdiff-4.3-1.tar.bz2 bsdiff-4.3.README contains VER and REL from the template. It is my understanding that these should be substituted for the actual values in the README. Your understanding is correct. Also, the binary package contains the empty directory /usr/share/doc/bsdiff-4.3/. It seems pointless to create this if there is no documentation to go there, and the user might think something was missing. This is actually g-b-s creating the directory unconditionally. I'm testing a rework of the logging code now, and will try to insert a check for documentation after that. Not a showstopper, IMO. Otherwise everything looks fine. Building from the source package and a quick test of the binary both seemed ok. This looks like a useful app, so I guess +1 from me. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac
Re: g-b-s patch: detached signature if package
On Sat, 28 Jan 2006, Igor Peshansky wrote: On Sat, 28 Jan 2006, Lapo Luchini wrote: If signatures are wanted, it seems sensible to me to also create ones for source and binary package, not only inside the source package itself: more useful for the security-savvy end-user, while the latter are more useful for the advanced user that want to re-compile the package itself. First off, you're missing a ChangeLog. I have a comment on the patch itself -- I know that sigfile() checks that $SIG eq 1 internally, but it would still be nice to wrap the calls to it in another check. This is so I can later add a flag for signing without having to set environment variables. Also, I'm in the process of testing a new logging setup in the g-b-s. I won't be able to apply your patch until that code stabilizes and is checked in. Igor Oh, and thanks for the patch, of course! Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac
Re: g-b-s patch: detached signature if package
On Sat, 28 Jan 2006, Lapo Luchini wrote: If signatures are wanted, it seems sensible to me to also create ones for source and binary package, not only inside the source package itself: more useful for the security-savvy end-user, while the latter are more useful for the advanced user that want to re-compile the package itself. First off, you're missing a ChangeLog. I have a comment on the patch itself -- I know that sigfile() checks that $SIG eq 1 internally, but it would still be nice to wrap the calls to it in another check. This is so I can later add a flag for signing without having to set environment variables. Also, I'm in the process of testing a new logging setup in the g-b-s. I won't be able to apply your patch until that code stabilizes and is checked in. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac
Re: [ITP] bsdiff-4.3
On Sat, 28 Jan 2006, Lapo Luchini wrote:
Igor Peshansky wrote:
bsdiff-4.3.README contains VER and REL from the template. It is
my understanding that these should be substituted for the actual
values in the README.
Your understanding is correct.
OK then, here you go... version 4.3-3:
http://cyberx.lapo.it/cygwin/bsdiff/
and relevant patch to sed inline those evil tags...
diff -b -u -r1.45 generic-build-script
--- templates/generic-build-script 22 Jan 2006 04:35:42 - 1.45
+++ templates/generic-build-script 28 Jan 2006 15:43:02 -
@@ -295,14 +295,15 @@
fi \
/usr/bin/install -m 755 ${srcdir}/CYGWIN-PATCHES/preremove.sh \
${instdir}${sysconfdir}/preremove/${PKG}.sh ; \
- fi
+ fi \
if [ -f ${srcdir}/CYGWIN-PATCHES/manifest.lst ] ; then
if [ ! -d ${instdir}${sysconfdir}/preremove ]; then
mkdir -p ${instdir}${sysconfdir}/preremove ;
fi
/usr/bin/install -m 755 ${srcdir}/CYGWIN-PATCHES/manifest.lst \
${instdir}${sysconfdir}/preremove/${PKG}-manifest.lst ;
- fi )
+ fi \
+ sed -i -e s/VER/$VER/;s/REL/$REL/
${instdir}${prefix}/share/doc/Cygwin/${SHORTPKG}.README )
^^^
The patch got wrapped, FWICS.
}
strip() {
(cd ${instdir} \
FWIW, there was a patch from Alan Miles two years ago to do a more general
version of this, but he didn't follow through on it. Here are the links
to the relevant messages:
http://cygwin.com/ml/cygwin/2003-11/msg00912.html
http://cygwin.com/ml/cygwin-apps/2003-12/msg00292.html
http://cygwin.com/ml/cygwin-apps/2004-01/msg00042.html
http://cygwin.com/ml/cygwin-apps/2004-02/msg00090.html
http://cygwin.com/ml/cygwin-apps/2004-02/msg00091.html
http://cygwin.com/ml/cygwin-apps/2004-02/msg00297.html
http://cygwin.com/ml/cygwin-apps/2004-02/msg00298.html
(the last message was never followed up on).
As part of that discussion, a bug in sed -i was discovered: if /tmp is
not writable, sed would go into an infinite loop. AFAICS, a workaround
would be to always invoke sed -i with TMP (or TMPDIR, need to check
that) set to a known-writable directory (e.g., .). Feel like changing
the patch while you're at it?
Igor
--
http://cs.nyu.edu/~pechtcha/
|\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED]
ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!)
|,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski
'---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow!
Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte.
But no -- you are no fool; you call yourself a fool, there's proof enough in
that! -- Rostand, Cyrano de Bergerac
Heads-up: G-B-S changes (bash, logging)
To everyone who uses G-B-S: I've just committed a change that turns off logging functionality in the generic-build-script by default. If you still want logs, please invoke your scripts with the with_logs (or --logs) flag, or insert a call to enablelogging() before the option processing loop. I've also fixed the wrong exit code when logging bug, so logging should be safe to use even for potentially failing builds. Note that some bash-isms were needed to make this work correctly, so the g-b-s is now a bash script (but since Cygwin's sh is bash, this shouldn't be a problem). Please test. Comments and suggestions welcome. Also, if you have g-b-s patches pending, please resubmit against CVS HEAD. Thanks, Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac
Re: Problems with cygwin cvs over ssh.
On Sat, 28 Jan 2006, Corinna Vinschen wrote: On Jan 27 17:01, Igor Peshansky wrote: Ugh, top-posting... Reformatted. On Fri, 27 Jan 2006, Frank-Michael Moser wrote: Igor Peshansky wrote: The problem is with the CVS server running on Cygwin. CVS in client mode works just fine. I have encountered the same problem, have investigated it a bit and came out with two interesting facts: A) Without changing anything else than replacing cygwin1.dll, using the snaphot cygwin1-20050928.dll works fine while using cygwin1-20050929.dll produces the problem. B) Also without changing anything else than replacing cygwin1.dll, using the snaphot cygwin1-20050928.dll mkdir /tmp/foo/. runs fine while with cygwin1-20050929.dll you see: $ mkdir /tmp/foo/. mkdir: cannot create directory `/tmp/foo/.': No such file or directory Right. I missed the . in the original message. The change that prompted this behavior seems to be http://cygwin.com/ml/cygwin-cvs/2005-q3/msg00224.html. I'm assuming the motivation for this patch was to duplicate Linux's behavior (which doesn't allow trailing . in a path passed to mkdir). Indeed. Eric mentioned that the coreutils testsuite tests thsi behaviour explicitely. Since there's not much impact speedwise, we just added appropriate checks to be POSIX compatible here. So I guess it's now up to the cvs maintainer to try tracking this down (or reporting this as a bug upstream). FWIW, I've looked at the sources of 1.11.17 and didn't see where that . was added. Did anyone try this with a test cvs-1.11.21-1? Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: curses.h (Attn: bash and setup maintainers)
On Tue, 24 Jan 2006, Eric Blake wrote: According to Igor Peshansky on 1/23/2006 9:34 AM: Most of the messages are diagnostics to track script progress. The only exceptions are the lines below: ./00bash.sh.done: line 12: ./01bash.bat: No such file or directory cp: cannot stat `./01bash.bat.t': No such file or directory rm: cannot remove `./01bash.bat.t': No such file or directory The first set is from 00bash.sh (it should probably have a test for the existence the .bat file). Also, in-place edits should work just fine (using 'sed -i -e /^echo/d; s,REM BASHPATH,$bashpath, 01bash.bat'). Duly noted; bash-3.1-2 will improve on this situation One quick note: I chanced upon my comment from 2 years ago regarding a bug in sed that makes it go into an infinite loop with -i if /tmp is not writable. Don't know if it still applies, but this is something to watch out for. A workaround is to invoke sed with TMP (or TMPDIR) set to a known-writable directory (e.g., /etc/postinstall or .). Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: multimonitor on laptops if monitor added after X starts (canresize down, but not up)
Hey, Eric, great to hear you're using Cygwin... :-) On Fri, 27 Jan 2006, Freudenthal, Eric wrote: I am so impressed with your fullscreen, multiplemonitors, and clipboard support. Bravo! Just one nagging nit: -multiplemonitors works great on my xp laptop (dell/intel chipset) if the additional display is added prior to Xwin starting. I can even switch to single monitor and go back, with everything working. Looks like the multiplemonitors code relies on GetSystemMetrics(), which will only return information about the attached displays. Attaching the extra display afterwards probably generates a Windows message that the code doesn't handle (it should probably redo the check when handling display resolution changes, etc). IOW, it's a design bug... However, if I start Xwin prior to adding an additional monitor, Xwin won't handle the increased number of rows/cols properly. Perhaps a fix would just require that I specify my max rows/cols at Xwin's startup (but how?). For example, XWin :0 -screen 0 2048 768 But this will likely screw up window positioning and offset computation if you decide not to attach the other monitor. HTH, Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://x.cygwin.com/docs/ FAQ: http://x.cygwin.com/docs/faq/
Re: Calling shell script from DOS
On Thu, 26 Jan 2006, Pinaki Mukherjee wrote: I have installed Cygwin. I want to invoke a shell script from the DOS/Windows command prompt (instead of opening Cygwin Window first and calling it from there). How can I do this? In addition to what others said, use bash -c instead of bash to honor the shebang (#!) line. This would then work for tcsh, ksh, perl, python, [you name it] scripts, as well as symlinks. In fact, whenever you need to invoke a Cygwin program from the DOS prompt, and you don't know if it's a script, a symlink, or an actual .exe, you can't go wrong with bash -c progname. Adding --login (-l) is optional, but may be useful for scripts that make assumptions about your environment. HTH, Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Problems with cygwin cvs over ssh.
Ugh, top-posting... Reformatted. On Fri, 27 Jan 2006, Frank-Michael Moser wrote: Igor Peshansky wrote: The problem is with the CVS server running on Cygwin. CVS in client mode works just fine. I have encountered the same problem, have investigated it a bit and came out with two interesting facts: A) Without changing anything else than replacing cygwin1.dll, using the snaphot cygwin1-20050928.dll works fine while using cygwin1-20050929.dll produces the problem. B) Also without changing anything else than replacing cygwin1.dll, using the snaphot cygwin1-20050928.dll mkdir /tmp/foo/. runs fine while with cygwin1-20050929.dll you see: $ mkdir /tmp/foo/. mkdir: cannot create directory `/tmp/foo/.': No such file or directory Right. I missed the . in the original message. The change that prompted this behavior seems to be http://cygwin.com/ml/cygwin-cvs/2005-q3/msg00224.html. I'm assuming the motivation for this patch was to duplicate Linux's behavior (which doesn't allow trailing . in a path passed to mkdir). There must be some Cygwin-specific code in cvs that adds that trailing dot. I'll take a look at the cvs sources later. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: How to compile the cygwin dll's using a functional Cygwin system?
On Thu, 26 Jan 2006, Keith Christian wrote: I'm looking for the instructions (I've searched with Google etc.) about how to compile the Cygwin DLL's from source using non-Microsoft compilers, e.g. those included with a functional Cygwin system. Did you try the FAQ? http://cygwin.com/faq/faq.programming.html#faq.programming.building-cygwin Is this possible, or do the DLL's have to be built using either free or low cost Microsoft compilers? You got it backwards. I don't think you *can* build Cygwin using Microsoft's compilers, while it certainly builds using Cygwin tools. It would be interesting to find out where you got the idea that Cygwin needs a Microsoft compiler to build... Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: submission rules page proposal number 3
On Wed, 25 Jan 2006, Corinna Vinschen wrote: On Jan 25 08:26, Lapo Luchini wrote: Corinna Vinschen wrote: FYI, I was going to upload the new setup.html file, but I can't access your server right now. Yes, my home ADSL seem to have misbehaved for some hours yesterday (11:00-17:00 circa). It should now work (well, if you receive this message, it certainly does ^_^). Yep, I've uploaded the new setup.html page to sourceware.org now. I think reworking this page deserves a gold star. Igor? Is that a question? If so, I certainly agree... :-) Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac
Re: RFC: [PATCH] Simplify installed.db format (take 3)
On Wed, 25 Jan 2006, Christopher Faylor wrote: : On Sun, 22 Jan 2006, Igor Peshansky wrote: : The attached patch fixes the parsing of installed.db package lines. : Comments welcome. Will this require a change to cygcheck also? No. Cygcheck already looks only at the first two space-delimited fields in each line of installed.db -- I'm proposing to change the third. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac
Re: [UPLOAD] Base-files 3.7-1
On Wed, 25 Jan 2006, John Morrison wrote:
Change Log
--
3.7-1
[snip]
* Follow links in /etc/profile.d - Max Bowsher
One small problem with this change: FWICS, the files will be found all
right, but the symlinks will later be ignored by the if test. For this
to work, 'if [ -f ${f} ]; then' needs to be changed to 'if [ -f ${f}
-o -l ${f} ]; then'. Actually, the whole if test seems redundant to
me anyway, now that you're doing find -type f...
Igor
--
http://cs.nyu.edu/~pechtcha/
|\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED]
ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!)
|,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski
'---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow!
Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte.
But no -- you are no fool; you call yourself a fool, there's proof enough in
that! -- Rostand, Cyrano de Bergerac
Re: [Patch] regtool: Add load/unload commands and --binary option
On Wed, 25 Jan 2006, Brian Dessent wrote: Christian Franke wrote: At least when regtool is used interactively, it is IMO not very useful to have modem-line-noise-like output for REG_BINARY, but human readable output for the other value types. But this is the current behavior of regtool get Instead of an explicit -b flag, perhaps it should just call isatty() and if being run interactively, output human readable hex dump, otherwise output raw binary. What if you want to redirect the hex dump to a file? IMO, isatty() checks are only useful if the output won't change qualitatively on redirection (e.g., for coloring). Otherwise, it's always better to use an explicit flag. That said, attempting to regtool get a binary value is no different from attempting to cat a binary file -- and we're not changing cat to output hex dumps. What would be useful instead is a way to query the type of a value via regtool. We could also add a -h (for --human-readable) flag to always output the value in human-readable form (see below). FWIW, 'A=`regtool get KEY1\\VAL`; regtool set KEY2\\VAL $A' will already break for REG_MULTI_SZ type values. It would be good to make regtool's behavior consistent: regtool set -f FILE reads the value from FILE, which would work for all value types. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac
Re: Shell (bash, (pd)ksh, zsh, /not/ ash) + exec + here-doc +redirect == trouble!
On Wed, 25 Jan 2006, Bas van Gompel wrote: Hi, Try the following script: === begin testexec.sh === #!/bin/ksh exec 50 /bin/ksh EOSH echo First exec: Done. exec 05 echo Second exec: Done. exit 0 EOSH end testexec.sh You might want to quote 'EOSH', but that's not the cause of your problem. Here's a funny thing: #!/bin/bash exec 50 /bin/bash 'EOSH' echo First exec: Done. cat exec 05 echo Second exec: Done. exit 0 EOSH works for me. Don't ask me why, though. Perhaps Eric will chime in. Has anybody got a clue? Is this cygwin-specific? That's easy to check. It isn't. I get the same behavior with bash-3.0 on Linux. Are all these shells borrowing code from each other? Most definitely not. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Errors compiling cdrtools under cygwin 1.5.19
On Wed, 25 Jan 2006, pobox wrote: I like this slogan: Cygwin: nice, but not reliable. We should adopt this, and refer 90% of the mailing list complaints to it. :-) Igor I am not really sure how to understand the joke. Only that Cygwin cannot be expected to be fully tested and reliable without input from the users. People using Cygwin for critical tasks should either not upgrade, or test the snapshots, so that they are not caught by surprise at release time. cygwin worked flawlessly for me for very long time. May be I have been lucky, but this made me develop an attitude of blind trust in cygwin. Blind trust is always a bad idea. I'm not aware of any software product (free or not) that will not change its behavior from time to time. Then came this 'getline' problem and what is confusing is not that things broke, but the reaction of the cygwin people. What reaction? We have not done anything differently than Linux? It's true. FWIW, some applications have Cygwin-specific patches in their codebase that are supposed to work around previous deficiencies in Cygwin. As Cygwin evolves to be more like Linux, those programs should remove those kludges. Some don't. I'm not saying this is what happened in this particular case, but that's one possible way in which Cygwin changes can break another application. What I saw is that it is possible that cygwin breaks from one day to the next in critical area, and the reaction will be that everything that breaks is at its own fault. An area that is critical for you may not be critical for others. If it *is* that critical, why did you not test the snapshots? There were multiple calls for this. Linux has an equivalent getline() call. Apache (presumably, I haven't checked) builds on Linux. Thus, either Cygwin's getline declaration is somehow different from Linux's (in which case it should be fixed in Cygwin), or Apache is doing something different in Cygwin than in Linux (in which case it should be fixed in Apache). But someone has got to find out where the fix should go, and the resources of Cygwin developers are stretched pretty thin as it is. That someone will have to be the person for whom building Apache is critical (e.g., you), or a person that volunteers his time (in which case, you'd have to wait for them to find enough time to work on this). Since you mentioned using VMWare, why not install Linux in it and try building Apache? Then compare the logs to see what's different in the builds. The problem is that if Apache have not fixed their 'getline' since 2002, it is very likely that they will not fix it soon, if at all. Certainly not if nobody advocates fixing this. Make some noise -- you never know, it might have just fallen through the cracks. cygwin has quite a marginal place in comparison to all other operating systems they should address, an cygwin should be - due to this - a bit more careful when challenging the word. Like any other software, Cygwin has bugs. These bugs ought to be fixed, but first they need to be tracked down. Apache doesn't build is not the bug. The bug would be Apache doesn't build because some code that it expects to be this way is that way. Reducing it down to a simple test case (a small program that uses getline() and builds on Linux but not on Cygwin) would help. It's also possible, as I said above, that the bug *is* in Apache, but did not manifest until now. At least for me with cygwin is over. We will gladly refund the full purchase price of this software. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Problems with cygwin cvs over ssh.
On Wed, 25 Jan 2006, James Courtier-Dutton wrote: CVS_RSH=/bin/ssh CVS -d [EMAIL PROTECTED]:/test-repository co test1 ...password cannot mkdir /tmp/cvs-server1396/. No such file or directory. I am running OpenSSH on a cygwin windows server. The CVS repository is on the windows server. From CYGWIN: OpenSSH_4.2p1 CVS 1.11.17 So, the problem is the ssh and cvs combined. If I ssh, then run cvs it checks out fine. The problem is getting the information back through the ssh pipe. Are you able to mkdir /tmp/cvs-server1396 from an ssh session? What are the permissions on /tmp? Also, please read and follow the Cygwin problem reporting guidelines at http://cygwin.com/problems.html, particularly the part that tells you to attach (as an uncompressed text *attachment*) the output of cygcheck -svr on your system. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Losing key bindings in Bash 3.00.16(14)-release
On Wed, 25 Jan 2006, Keith Christian wrote: Hi Folks, Versions: Windows XP SP2, RXVT 2.7.10, Bash 3.00.16(14)-release, Cygwin DLL 1.5.18. Since Bash 3.x became the default version installed with Cygwin, on my system, Bash inexplicably loses the command line edit bindings. This usually happens when a window has been open for a day or more and I've done alot of work in it. The problem shows itself when I press Ctrl-A or Ctrl-E or other command line editing keys, and instead of going to the start or the end of the line, I see ^A or ^E in the command line at the cursor position. Are you getting the control characters for just one command, or for more than one? This sounds weird enough that you'd probably have to debug this yourself. Also, does this happen in the regular bash window, rather than rxvt? Does this happen in an xterm? I haven't discovered how to restore editing bindings to the BASH command line other than closing the RXVT window and re-opening it. Try set -o emacs. HTH, Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Shell (bash, (pd)ksh, zsh, /not/ ash) + exec + here-doc + redirect == trouble!
On Wed, 25 Jan 2006, Eric Blake wrote: You might want to quote 'EOSH', but that's not the cause of your problem. Here's a funny thing: #!/bin/bash exec 50 /bin/bash 'EOSH' echo First exec: Done. cat exec 05 echo Second exec: Done. exit 0 EOSH works for me. Don't ask me why, though. Perhaps Eric will chime in. Define works for me. $ ./exectest.sh First exec: Done. exec 05 echo Second exec: Done. exit 0 D'oh! I was thrown off by the fact that I used bash -xv in my test script. Sorry for the noise. Here, the cat uses fd 0 (which is still set to the pipe from the here-doc), and consumes it by echoing the remainder of the here-doc to stdout, so that when /bin/bash goes to read the next line, fd 0 is at EOF, and you never execute the exec 05 line in the shell. Thus, the process never tries to duplicate fd 5 back to fd 0, and your example no longer hangs waiting for input from the terminal. But it doesn't do what was originally intended (that is, allow /bin/bash to execute the remainder of the here-doc). Right. Your redirection order explanation is correct. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Cygwin Setup: Fatal Error: Uncaught Exception
On Wed, 25 Jan 2006, Dave Korn wrote: Igor wrote: I've looked at this a bit. Here's the weird part: the error says Uncaught Exception, but all the throws of that exception appear to be properly wrapped in try/catch blocks. So a simple change exception into an mbox kind of solution won't work here. More debugging is needed. That's because the io_stream code throws an invalid_argument, which is derived from std::exception, but our code only catches Exception, which is our custom exception type. Nope. It's the same old Exception, and it is indeed uncaught -- I just missed one use. See the rest of the thread, plus the patch on cygwin-apps. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Problems with cygwin cvs over ssh.
On Wed, 25 Jan 2006, James Courtier-Dutton wrote: On 25/01/06, Karl-Heinz Wirth [EMAIL PROTECTED] wrote: http://cygwin.com/acronyms/#PCYMTNQREAIYR. Thanks. Hi On Wed, 25 Jan 2006, James Courtier-Dutton wrote: CVS_RSH=/bin/ssh CVS -d [EMAIL PROTECTED]:/test-repository co test1 ...password cannot mkdir /tmp/cvs-server1396/. No such file or directory. I am having the same problem now, after upgrading to cygwin 1.5.19-4 from 1.5.18-1. I am working with the Eclipse IDE on Windows which uses cygwins cvs and ssh in the background So, the problem is the ssh and cvs combined. If I ssh, then run cvs it checks out fine. The problem is getting the information back through the ssh pipe. The problem is with cvs server. I was able to reproduce this without ssh at all by using the following as input to cvs server: Root /var/cvs/BLAH Valid-responses ok error Valid-requests Checked-in New-entry Checksum Copy-file Updated Created Update-existing Merged Patched Rcs-diff Mode Mod-time Removed Remove-entry Set-static-directory Clear-static-directory Set-sticky Clear-sticky Template Notified Module-expansion Wrapper-rcsOption M Mbinary E F MT valid-requests UseUnchanged Global_option -q Global_option -r Argument . Directory . /var/cvs/BLAH expand-modules Argument -N Argument -P Argument -- Argument . Directory . /var/cvs/BLAH co (this was captured from a cvs co over ssh). The funny thing is that the directory in the message *is* created before the commands are even read, and yet cvs still complains that it can't create it because of ENOENT. cvs init over ssh completed fine. I will investigate further when I have the time (though hopefully the above steps for reproducing this will help someone else debug this). cvs with ssh works fine in my installation, if i reinstall cygwin 1.5.18-1. So, going back to cygwin 1.5.18-1 is my current workaround. For those that might be interested, my current work around is to uninstall cygwin cvs and use CVSNT instead on the server. So, it looks very much like a cygwin cvs problem at the moment. Although, on another project where we were using ssh to run scripts remotely and return the output, the script was running fine, but the output never reached back. This points me towards a problem with openssh on cygwin. It might be a problem with pipes under cygwin, as I assume it is the unix feature of pipes that redirects the output from commands back through the ssh session. E.g. This works: local $ ssh [EMAIL PROTECTED] host $ run-script- This displays output correctly on local. host $ exit This fails: local $ ssh [EMAIL PROTECTED] 'run-script'- No output displayed. To be more exact, we were nesting the ssh commands, i.e. run-script contained a ssh [EMAIL PROTECTED] 'run-script2' line. It was the output from this nested ssh that failed to be returned. When I get a moment, I will try to post some example scripts to demonstrate the problem. This looks unrelated. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Problems with cygwin cvs over ssh.
On Wed, 25 Jan 2006, James Courtier-Dutton wrote: On 25/01/06, Igor Peshansky xxx wrote: On Wed, 25 Jan 2006, James Courtier-Dutton wrote: On 25/01/06, Karl-Heinz Wirth [EMAIL PROTECTED] wrote: http://cygwin.com/acronyms/#PCYMTNQREAIYR. Thanks. Does anyone know how to configure gmail to do this? I'm not aware of a way to configure any webmail service for this. However, if enough people lobby for this feature, it's likely to be added. Note: Special manual deletion of yours. Thanks, appreciated. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Cygwin in any directory by menuchoice
On Wed, 25 Jan 2006, Fredrik Lundgren wrote: Hello list, In win xp you can get the chose to start cmd.exe in any directory of your choice with rightclick and a menuchoice with ### Windows registry Editor Version 5.00 [HKEY_CLASSES_ROOT\Folder\shell\Cmd Here] @=Command Prompt Here [HKEY_CLASSES_ROOT\Folder\shell\Cmd Here\command] @=cmd.exe /k pushd %L ### Can something similar be achieved with Cygwin+ That's what the chere package is for: sdesc: Cygwin Prompt Here context menus ldesc: Shell script to manage shell here context menus for Cygwin HTH, Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Problems with cygwin cvs over ssh.
On Wed, 25 Jan 2006, René Berber wrote: James Courtier-Dutton wrote: CVS_RSH=/bin/ssh CVS -d [EMAIL PROTECTED]:/test-repository co test1 --^^^ Is this correct? or it was really _cvs_. Both should work unless you use check_case:strict or a shell that does case-sensitive hashing. ...password cannot mkdir /tmp/cvs-server1396/. No such file or directory. I am running OpenSSH on a cygwin windows server. The CVS repository is on the windows server. [snip] From the manual: If the access method is omitted, then if the repository starts with '/', then :local: is assumed. Try: cvs -d :ext:[EMAIL PROTECTED]:/test-repository co test1 FWICS, [EMAIL PROTECTED] doesn't start with a '/'. No, that's no cockpit error -- it's a real bug of some sort. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Problems with cygwin cvs over ssh.
On Wed, 25 Jan 2006, René Berber wrote: Igor Peshansky wrote: [snip] From the manual: If the access method is omitted, then if the repository starts with '/', then :local: is assumed. Try: cvs -d :ext:[EMAIL PROTECTED]:/test-repository co test1 FWICS, [EMAIL PROTECTED] doesn't start with a '/'. fred is not in the repository part of the command. To the OP: Just try it using :ext: and report back, perhaps I can learn something. FWIW, I could reproduce the original problem, either with or without :ext:. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Problems with cygwin cvs over ssh.
On Wed, 25 Jan 2006, René Berber wrote: Igor Peshansky wrote: [snip] FWIW, I could reproduce the original problem, either with or without :ext:. The combination cvs/ssh has no problem : $ echo $CVS_RSH ssh $ echo $CVSROOT :ext:[EMAIL PROTECTED]:/export/home0/cvsrep $ cvs co junit-test cvs server: Updating junit-test U junit-test/.classpath U junit-test/.project cvs server: Updating junit-test/lib ... Remote CVS server is Solaris, local ssh and Cygwin dll versions are latest released. Then you have misunderstood the original problem. The problem is with the CVS server running on Cygwin. CVS in client mode works just fine. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Updated: coreutils-5.93-2 [Attn base-files maintainer]
On Tue, 24 Jan 2006, Eric Blake wrote: [snip] An unrelated comment: Due to reinstating su.exe, coreutils now depends on crypt. To date, no package in Base has had a dependency on crypt, so the new coreutils does increase the size of the minimal cygwin install. Speak up now if anyone has heartburn with this. Gasp! How will I live with those extra 48k on my disk? We might want to move crypt into Base, though... Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac
Re: HELP: cygwin setup.exe woes...
On Tue, 24 Jan 2006, KevinGPO wrote: [snip] Now I want to download a complete cygwin (Interpreters, Development, etc.) however it won't do that. All I get is 10MB downloaded files. Somehow it has remembered my previous downloads... How can I get around this? Cygwin never used to do this? or did it?? Heh, if you think these are woes, you should see the complaints about the Uncaught Exception... :-) Setup is primarily an installer, not a download tool. As such, it won't even bother downloading packages that are already installed on the system. It knows what's installed by looking at /etc/setup/installed.db in your Cygwin tree. So, the simplest way to get it to temporarily forget what it has previously downloaded is to rename installed.db while downloading. Setup should re-create it, but if you only download (and not install) things, the new copy will be empty and can be replaced by the old. I wouldn't recommend actually installing anything with no installed.db, as it wouldn't be trivial to bring your system to a consistent state afterwards. One other thing that seemed weird is you first downloading and then installing packages. If all you want is to install things, setup can do it in one shot (still retaining the cached download directory). HTH, Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: HELP: cygwin setup.exe woes...
On Tue, 24 Jan 2006, Igor Peshansky wrote: On Tue, 24 Jan 2006, KevinGPO wrote: [snip] Now I want to download a complete cygwin (Interpreters, Development, etc.) however it won't do that. All I get is 10MB downloaded files. Somehow it has remembered my previous downloads... How can I get around this? Cygwin never used to do this? or did it?? Heh, if you think these are woes, you should see the complaints about the Uncaught Exception... :-) Setup is primarily an installer, not a download tool. As such, it won't even bother downloading packages that are already installed on the system. It knows what's installed by looking at /etc/setup/installed.db in your Cygwin tree. So, the simplest way to get it to temporarily forget what it has previously downloaded is to rename installed.db while downloading. ^ s/while/before/ Setup should re-create it, but if you only download (and not install) things, the new copy will be empty and can be replaced by the old. I wouldn't recommend actually installing anything with no installed.db, as it wouldn't be trivial to bring your system to a consistent state afterwards. One other thing that seemed weird is you first downloading and then installing packages. If all you want is to install things, setup can do it in one shot (still retaining the cached download directory). Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: curses.h (Attn: bash and setup maintainers)
On Tue, 24 Jan 2006, Eric Blake wrote: According to Igor Peshansky on 1/23/2006 9:34 AM: Most of the messages are diagnostics to track script progress. The only exceptions are the lines below: ./00bash.sh.done: line 12: ./01bash.bat: No such file or directory cp: cannot stat `./01bash.bat.t': No such file or directory rm: cannot remove `./01bash.bat.t': No such file or directory The first set is from 00bash.sh (it should probably have a test for the existence the .bat file). Also, in-place edits should work just fine (using 'sed -i -e /^echo/d; s,REM BASHPATH,$bashpath, 01bash.bat'). Duly noted; bash-3.1-2 will improve on this situation (whenever I get time to get that working; I'm still struggling with gracefully incorporating upstream patches without wiping them out by rerunning g-b-s prep). Does http://cygwin.com/ml/cygwin-apps/2006-01/msg00064.html make sense? Actually, now that setup.exe has been updated to use /bin/bash and not /bin/sh, the technical reason for me needing two postinstall scripts in the first place has been overcome, so I may just simplify back to a single script that does it all, rather than staging it through a .bat. You can't, unfortunately. Some people may have older versions of setup.exe on their systems, and blatantly disregard setup's warning that a newer version is available. Those systems will then fail to install sh properly. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: curses.h
On Tue, 24 Jan 2006, Bob Rossi wrote: In any case, looks like all the postinstall scripts ran for you, so you should be good to go. Hi Igor, So do you think that I broke CGDB somehow? When I compile and run it on Cygwin, it's display in the terminal is not correct. However, if I run a pre-compiled older version, it's display is fine. I'm going to build an older version today, and see if it still works. There are multiple possibilities. Many Cygwin packages have Cygwin-specific patches to compensate for either upstream non-portability or Cygwin's idiosyncracies. It's also possible that you just need to re-run configure for the newer version, as the old run may not have picked up the right libraries due to your installation mishap. Did a clean build fail for you too? Do you get the same problems when setting TERM to something widely used, e.g., ansi or xterm? Do you get the same problem in rxvt? Alternatively, something may have indeed changed in either CGDB or Cygwin that caused a bug to manifest. I think at this point we veered off the original topic of this thread... Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: ssh starting problems.
On Tue, 24 Jan 2006, JC Oosthuizen wrote: Did you use /usr/bin/ssh-host-config to set up sshd on the Win 2003 server? Currently I am running sshd on two such servers and set them up using the script. The script should detect that you are using Win 2003 and will ask if you want it to create a sshd_server user account and assign the privileges it needs under Local Security Policy to run properly. The sshd service should then be run under this account. ssh was setup using ssh-host-config and worked correctly. I then upgraded cygwin by removing it and reinstalling it. In this process the sshd_server account was not deleted and was not recreated either. This caused the problems that occured. I just removed the sshd_server user and then ran ssh-host-config again and all worked as it should. Ah, this is a clue. By removing Cygwin, you probably also removed /etc/passwd. AIUI, without sshd_server in /etc/passwd, user context switching won't work. Re-running ssh-host-config added sshd_server to /etc/passwd again. I'm not too well-versed in the details of ntsed -- Corinna should be able to confirm or deny the above. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Errors compiling cdrtools under cygwin 1.5.19
On Tue, 24 Jan 2006, pobox wrote: This falls into the same category as my previous discussion about _POSIX_SOURCE. If a program builds without problem on linux, the goal is for it to build without problem on cygwin. It seems like the unconditional addition of getline to the headers moves us a step back from that goal. Is this also valid for Apache 1.3.34? I think it compiles without problem on Linux, but does not compile on cygwin since getline() was implemented last month. This is a simple name clash. Linux also defines getline (also with an incompatible prototype). htpasswd.c defines its own getline(), and it should have the same problem on any system with getline(). You have to check why it compiles (in particular, check the pre-processed form of htpasswd.c using gcc -E). I do not want to heat the discussion, but getline() in cygwin played very hard against me. I used cygwin happily for very long time to compile apache/php/postgresql and enjoy symlinks, and now I am cut-off from one day to the next. The apache folks do not seem to care. The bug I submitted is still without reply - http://issues.apache.org/bugzilla/show_bug.cgi?id=38364 That's because the Apache community has apparently been ignoring this issue since 2002: http://issues.apache.org/bugzilla/show_bug.cgi?id=9894 (which, BTW, contains a patch that you can apply to fix this issue). Apache 2.x/php 5.x do not want to play on cygwin so far. I seem to recall either Brian or Max reporting that they were able to compile PHP 5.x with the Cygwin Apache2 package. Search the cygwin-apps archives. So I am three days in the dark and testing like hell vmware and minigw to save my skin. Seems this getline() breaks quite a lot and I am not quite sure this is _very_ positive for cygwin. People just get left alone in the dark (no everybody can debug and patch) and the pride of cygwin is somehow self focused. I would expect such dramatic moves to be done with more care. Otherwise I could call cygwin nice, but not reliable. I like this slogan: Cygwin: nice, but not reliable. We should adopt this, and refer 90% of the mailing list complaints to it. :-) Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Cygwin Setup: Fatal Error: Uncaught Exception
On Tue, 24 Jan 2006, Yitzchak Scott-Thoennes wrote: On Mon, Jan 23, 2006 at 02:56:39PM -0500, Igor Peshansky wrote: Moving to cygwin-apps, as this is likely to get technical. On Mon, 23 Jan 2006, Brian Dessent wrote: Igor Peshansky wrote: I've looked at this a bit. Here's the weird part: the error says Uncaught Exception, but all the throws of that exception appear to be properly wrapped in try/catch blocks. So a simple change exception into an mbox kind of solution won't work here. More debugging is needed. The reason for the box is that the md5 checking was changed to run in a different thread than originally designed (now in the main thread instead of the download thread IIRC) and that thread does not have any particular catch handler for that exception, only the TOPLEVEL_CATCH which punts with the generic error. Do you mean packagemeta::ScanDownloadedFiles() calling packageversion::scan(), which calls check_for_cached()? Then yes, this isn't properly wrapped in a try/catch. I'd like to verify that this is the culprit (when someone sends me the corrupt tarball), but I think I see a proper fix for this. Will submit a patch shortly. Just to reemphasize, these are *not* corrupt tarballs. They are tarballs exactly as downloaded, extracted, and installed. It's just that later the versions on the cygwin mirror became different while keeping the same version/filename. I verified in a couple of the cases that the newer version actually had executables rebuilt, with slightly different file sizes and timestamps. I don't think I have any of them around any more, but if you were to pick a current tarball in your local package directory and un-bzip2 it and re-bzip2 it with a different compression level, you should see the problem. What Brian said. I've since managed to reproduce the problem with a zero-sized tarball (but you're right, anything with a mismatched size or md5 hash would have worked -- or, rather, not worked) and posted a patch. I would appreciate some comments on the discussion we had with Brian (on cygwin-apps). Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Errors compiling cdrtools under cygwin 1.5.19
On Tue, 24 Jan 2006, Christopher Faylor wrote: On Tue, Jan 24, 2006 at 10:28:50PM -0500, Christopher Faylor wrote: On Tue, Jan 24, 2006 at 06:46:31PM -0800, Yitzchak Scott-Thoennes wrote: The only care that really could be taken to prevent things like this is more users testing pre-release versions. Development snapshots of cygwin with getline() have been available for a long time now. Note that this isn't just addressed to you; if package maintainers heeded the release coming soon, please test a snapshot messages cgf sends out by testing that their packages build and run with the snapshot, there'd be less scope for problems. Thank you, for making this point Yitzchak! Can I get two gold stars for over here? Done. translation: for over here == for Yitzchak Rats, now I had to remove a.k.a. quot;over herequot; from after Yitzchak's name before checking this in... Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: RFC: [PATCH] Simplify installed.db format (take 3)
On Mon, 23 Jan 2006, Buzz wrote:
Op Sun, 22 Jan 2006 23:28:04 -0500 (EST) schreef Igor Peshansky
in Pine.GSO.4.63.0601222324410.18707ataccess1.cims.nyu.edu:
: On Sun, 22 Jan 2006, Igor Peshansky wrote:
: The attached patch fixes the parsing of installed.db package lines.
: Comments welcome.
+ sscanf (line, %s %s, pkgname, inst);
Won't this parse the `` 0'' into ``inst''? Did you mean
+ sscanf (line, %s %s , pkgname, inst);
perhaps?
Umm, yes, good point. In fact, let's use the attached patch instead.
Igor
--
http://cs.nyu.edu/~pechtcha/
|\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED]
ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!)
|,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski
'---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow!
Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte.
But no -- you are no fool; you call yourself a fool, there's proof enough in
that! -- Rostand, Cyrano de BergeracIndex: setup/cygpackage.cc
===
RCS file: /cvs/cygwin-apps/setup/cygpackage.cc,v
retrieving revision 2.15
diff -u -p -r2.15 cygpackage.cc
--- setup/cygpackage.cc 14 May 2005 15:57:29 - 2.15
+++ setup/cygpackage.cc 23 Jan 2006 15:06:52 -
@@ -66,7 +66,7 @@ cygpackage::createInstance (const String
packageversion
cygpackage::createInstance (String const pkgname, String const filename,
- size_t const fs, String const version,
+ String const version,
package_status_t const newstatus,
package_type_t const newtype)
{
@@ -75,7 +75,6 @@ cygpackage::createInstance (String const
temp-fn = filename;
temp-status = newstatus;
temp-type = newtype;
- temp-filesize = fs;
temp-setCanonicalVersion (version);
return packageversion(temp);
}
Index: setup/cygpackage.h
===
RCS file: /cvs/cygwin-apps/setup/cygpackage.h,v
retrieving revision 2.15
diff -u -p -r2.15 cygpackage.h
--- setup/cygpackage.h 14 May 2005 15:57:29 - 2.15
+++ setup/cygpackage.h 23 Jan 2006 15:06:52 -
@@ -72,7 +72,7 @@ public:
const package_type_t type);
static packageversion createInstance (String const , String const ,
- size_t const, String const ,
+ String const ,
package_status_t const,
package_type_t const);
@@ -92,7 +92,6 @@ private:
package_type_t type;
io_stream *listdata, *listfile;
- size_t filesize;
};
#endif /* SETUP_CYGPACKAGE_H */
Index: setup/package_db.cc
===
RCS file: /cvs/cygwin-apps/setup/package_db.cc,v
retrieving revision 2.30
diff -u -p -r2.30 package_db.cc
--- setup/package_db.cc 14 Oct 2005 04:10:26 - 2.30
+++ setup/package_db.cc 23 Jan 2006 15:06:53 -
@@ -57,8 +57,8 @@ packagedb::packagedb ()
if (!db)
return;
/* flush_local_db_package_data */
- char line[1000], pkgname[1000], inst[1000], src[1000];
- int instsz, srcsz;
+ char line[1000], pkgname[1000], inst[1000];
+ int instsz;
if (db-gets (line, 1000))
{
@@ -80,14 +80,11 @@ packagedb::packagedb ()
while (db-gets (line, 1000))
{
int parseable;
- src[0] = '\0';
+ int ign;
pkgname[0] = '\0';
inst[0] = '\0';
- srcsz = 0;
- instsz = 0;
- sscanf (line, %s %s %d %s %d, pkgname, inst, instsz, src,
- srcsz);
+ sscanf (line, %s %s %d, pkgname, inst, ign);
if (pkgname[0] == '\0' || inst[0] == '\0')
continue;
@@ -109,7 +106,7 @@ packagedb::packagedb ()
}
packageversion binary =
- cygpackage::createInstance (pkgname, inst, instsz, f.ver,
+ cygpackage::createInstance (pkgname, inst, f.ver,
package_installed,
package_binary);
Re: Cygwin Setup: Fatal Error: Uncaught Exception
Moving to cygwin-apps, as this is likely to get technical. On Mon, 23 Jan 2006, Brian Dessent wrote: Igor Peshansky wrote: I've looked at this a bit. Here's the weird part: the error says Uncaught Exception, but all the throws of that exception appear to be properly wrapped in try/catch blocks. So a simple change exception into an mbox kind of solution won't work here. More debugging is needed. The reason for the box is that the md5 checking was changed to run in a different thread than originally designed (now in the main thread instead of the download thread IIRC) and that thread does not have any particular catch handler for that exception, only the TOPLEVEL_CATCH which punts with the generic error. Do you mean packagemeta::ScanDownloadedFiles() calling packageversion::scan(), which calls check_for_cached()? Then yes, this isn't properly wrapped in a try/catch. I'd like to verify that this is the culprit (when someone sends me the corrupt tarball), but I think I see a proper fix for this. Will submit a patch shortly. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac
Re: [PATCH] setup: Handle the package validation exception
On Mon, 23 Jan 2006, Brian Dessent wrote: Igor Peshansky wrote: Well, let's look at it this way: if we are downloading, that's exactly the behavior we want -- pretend the corrupted cached version doesn't exist and re-download it. If we're installing from the local directory, then we cannot install the corrupted packages anyway, so they might as well not exist. Sounds to me like ignoring a corrupted version is the right course of action in any case. In fact, why should the user even care that their local version is corrupted? All they should know is whether it's available to be installed (when doing a local install). And for network installs, caching should be transparent anyway. For power users, we can add some logging on corrupted packages. The problem as I see it is that if somebody is doing a install from local directory and they have a bad package, it will be silently ignored. So they will not know that it needs to be deleted, redownloaded, etc. I'm not convinced. How is this different from the package not being there in the first place? It will just be gone from the list of available packages, potentially in a confusing way, for example: I installed 'foo' but it doesn't work! -- 'bar' was a dependency of 'foo' but because it was an invalid package 'bar' was ignored. Huh? Wouldn't setup complain about that? It will check foo, and see that bar is a dependency, but bar is not selected... There's even a special screen for that, isn't there? I updated my cygwin distro but the problem persists -- because the local copy of 'foo' package had the wrong md5, so it wasn't upgraded. Updated from local directory? Or from the network? If the latter, wouldn't it be re-downloaded due to the cached check failing? If the former, again, it's the same as not having it downloaded at all. I'm looking for libfoo but it's not on the list yes it is, just select it no, I don't see it are you sure, you might have to swtich to 'full' (34 further messages exchanged trying to diagnose) Oh, my downloaded copy was the wrong size and needed to be deleted and redownloaded. In all these cases the user needs to know that there is something wrong with their local copy of the package, so that they can delete or redownload it. If we silently pretend it doesn't exist then all kinds of hard-to-diagnose scenarios come to mind. In the scenarios you've given above, if network install is selected and the locally cached copy is corrupt, it will be re-downloaded. In case of local install, the corrupt copy will result in the same behavior as not having the package on your machine. Frankly, I fail to see the problem... Perhaps I'm missing something. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac
Re: x-term font
On Mon, 23 Jan 2006, thad wrote: how can i change the font size of xterm? If you really mean xterm, then this question is off-topic for this list, and belongs on the cygwin-xfree list. I'm redirecting my reply there and setting the Reply-To: accordingly. You need to use the -fn option to pass the name/size combination for the font you want. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://x.cygwin.com/docs/ FAQ: http://x.cygwin.com/docs/faq/
Re: Cygwin Setup: Fatal Error: Uncaught Exception
On Mon, 23 Jan 2006, Yitzchak Scott-Thoennes wrote: On Sun, Jan 22, 2006 at 09:37:12PM -0800, Brian Dessent wrote: Linda Walsh wrote: Is an application crash the expected behavior when package validation fails? Maybe it should mark it unusable or delete it? Yuck. I went through this, too. It appeared to me that there were a large number (a dozen? I no longer remember) of packages uploaded more than once, without bumping the version, and if you were so fortunate as to get the earlier rather than the later copy, setup won't have updated to the later. It might be easier to blow away your local package directory than to run setup over and over, each time identifying one more package to rename/delete. At the moment, yes. It's a consequence of a change in the md5 checking. It just means you need to delete the package, as it has the wrong size or checksum. It's on the todo list to clean this up. Is there a snapshot that has this fixed? I've looked at this a bit. Here's the weird part: the error says Uncaught Exception, but all the throws of that exception appear to be properly wrapped in try/catch blocks. So a simple change exception into an mbox kind of solution won't work here. More debugging is needed. To whomever can reproduce this problem: can you please run setup under gdb (even if you don't have a debug version -- this is just to get the log output) and send me the *full* log output and the tarball that setup claims to be corrupt off-list? Once I can reproduce this, I'll try to dig further. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: which package has cyggnome-menu-0.dll
On Mon, 23 Jan 2006, Eric Blake wrote: According to Vijay Kiran Kamuju on 1/23/2006 4:43 AM: Hi, Can anyone tell me which package has cyggnome-menu-0.dll, please? $ cygcheck -p cyggnome-menu-0.dll Found 0 matches for cyggnome-menu-0.dll Or on the web: http://cygwin.com/packages/ and search for the filename. In other words, no package on cygwin.com provides this, so this mailing list won't be able to help you. Maybe you should check with the cygwin-ports mailing list instead. If you have that file on your system, the following command will tell you which installed package contains it: $ cygcheck -f /usr/bin/cyggnome-menu-0.dll libgnome-menu0-2.10.2-1 (I may have gotten the version wrong -- I don't have it installed). In any case, what Eric said above about cygwin-ports still applies. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: curses.h
On Mon, 23 Jan 2006, Bob Rossi wrote: On Sat, Jan 21, 2006 at 05:31:24PM -0500, Igor Peshansky wrote: On Sat, 21 Jan 2006, Bob Rossi wrote: I will do this on Monday. However, is there a way to tell cygwin to fix all packages? Just in case others are broken? Once you have /bin/sh, you can re-run all postinstall scripts (go to /etc/postinstall and run everything that ends in .sh.done). [snip] I did try your suggestions about for I in *.sh.done; do ./$I; done, however, it was producing lot's of errors, so I stopped it. Heh? lot's of errors (sic) isn't particularly descriptive... What errors were you getting, exactly? Note that until you've run all postinstall scripts, your installation will not be in a consistent state. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
Re: curses.h (Attn: bash and setup maintainers)
On Mon, 23 Jan 2006, Bob Rossi wrote: On Mon, Jan 23, 2006 at 11:06:09AM -0500, Igor Peshansky wrote: On Mon, 23 Jan 2006, Bob Rossi wrote: On Sat, Jan 21, 2006 at 05:31:24PM -0500, Igor Peshansky wrote: On Sat, 21 Jan 2006, Bob Rossi wrote: I will do this on Monday. However, is there a way to tell cygwin to fix all packages? Just in case others are broken? Once you have /bin/sh, you can re-run all postinstall scripts (go to /etc/postinstall and run everything that ends in .sh.done). [snip] I did try your suggestions about for I in *.sh.done; do ./$I; done, however, it was producing lot's of errors, so I stopped it. What errors were you getting, exactly? Note that until you've run all postinstall scripts, your installation will not be in a consistent state. Here's the command I ran, along with output. for I in *.sh.done; do ./$I; done ~/out.txt 21 Bob, Most of the messages are diagnostics to track script progress. The only exceptions are the lines below: ./00bash.sh.done: line 12: ./01bash.bat: No such file or directory cp: cannot stat `./01bash.bat.t': No such file or directory rm: cannot remove `./01bash.bat.t': No such file or directory and ./XFree86-bin-icons.sh.done: line 8: /usr/X11R6/bin/XFree86-bin-icons.sh: No such file or directory The first set is from 00bash.sh (it should probably have a test for the existence the .bat file). Also, in-place edits should work just fine (using 'sed -i -e /^echo/d; s,REM BASHPATH,$bashpath, 01bash.bat'). The latter message is actually because of a bug in setup -- when it uninstalls a package, it doesn't take into account the fact that files in /etc/postinstall were renamed to .done, and thus leaves stale postinstall scripts behind (like the one from XFree86-bin-icons, which is an obsolete package). [Brian, can we put this on the TODO list to fix in setup?] In any case, looks like all the postinstall scripts ran for you, so you should be good to go. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Prompt issue within cygwin
On Mon, 23 Jan 2006, Zach Gelnett wrote: Here is what i am seeing: [EMAIL PROTECTED] ~ the duplicate is the issue. Here is my PS1: \[\033]61;[EMAIL PROTECTED]@\H \W as you can see, only a single [snip] You have a bug in your PS1 value. Bash uses \[ and \] to delimit non-printable sequences in the prompt. So, your PS1 should at least be \[\033]61;[EMAIL PROTECTED]@\H \W ^^ Note the closing '\]'. Also, ESC] doesn't look like a valid ANSI escape sequence What exactly are you trying to accomplish with your prompt? Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Prompt issue within cygwin
Ugh, top-posting... Reformatted. On Mon, 23 Jan 2006, Zach Gelnett wrote: On 1/23/06, Igor Peshansky [EMAIL PROTECTED] wrote: http://cygwin.com/acronyms/#PCYMTNQREAIYR. Thanks. On Mon, 23 Jan 2006, Zach Gelnett wrote: Here is what i am seeing: [EMAIL PROTECTED] ~ the duplicate is the issue. Here is my PS1: \[\033]61;[EMAIL PROTECTED]@\H \W as you can see, only a single [snip] You have a bug in your PS1 value. Bash uses \[ and \] to delimit non-printable sequences in the prompt. So, your PS1 should at least be \[\033]61;[EMAIL PROTECTED]@\H \W ^^ Note the closing '\]'. Also, ESC] doesn't look like a valid ANSI escape sequence. What exactly are you trying to accomplish with your prompt? Igor That was precicesly my issue. Thank you very much, both issues (tab complete and dup entries) are resolved! FYI, I'm just trying to have a prompt that says: [EMAIL PROTECTED] workingdir this is because i login to multiple machines and use tabbed windows (available within mrxvt) to do this. i find having this prompt makes it quite easy to keep track of what env i'm logged into and where i'm at. If that was all you wanted, you could have set PS1 to '[EMAIL PROTECTED] \W'. Your prompt tries to do more than that. I'm still unsure of what the initial sequence of non-printing characters in your prompt is for. BTW, forgot to mention: none of this is Cygwin-specific. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Prompt issue within cygwin
On Mon, 23 Jan 2006, Eric Blake wrote: \[\033]61;[EMAIL PROTECTED]@\H \W as you can see, only a single [snip] You have a bug in your PS1 value. Bash uses \[ and \] to delimit non-printable sequences in the prompt. So, your PS1 should at least be \[\033]61;[EMAIL PROTECTED]@\H \W ^^ Note the closing '\]'. Also, ESC] doesn't look like a valid ANSI escape sequence. What exactly are you trying to accomplish with your prompt? ESC] starts a sequence for sending a message to the terminal. Ouch. That's what I get for firing off emails in a hurry. I meant ESC]61; (which I'm glad you share my confusion about). In the standard bash prompt from the base-files package, the sequence: \[\e]0;\w\a\] sets up part of the bash prompt \[non-printing chars\], with a message to the terminal to change its title \e]0;text\a to the working directory \w. I have no idea what \e]61; in the OP's prompt is trying to tell the terminal to do, though; I've only ever seen \e]n; for n=0, 1, or 2, never for n=61. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Prompt issue within cygwin
Ugh, top-posting... Reformatted. Two more notes: first, the Cygwin lists don't accept HTML email, as you no doubt found out from the rejection notice, and second, please make sure your mailer respects the Reply-To: header -- I set it for a reason. On Mon, 23 Jan 2006, Zach Gelnett wrote: On 1/23/06, Igor Peshansky [EMAIL PROTECTED] wrote: Again, http://cygwin.com/acronyms/#PCYMTNQREAIYR. Thanks. Ugh, top-posting... Reformatted. On Mon, 23 Jan 2006, Zach Gelnett wrote: On 1/23/06, Igor Peshansky [EMAIL PROTECTED] wrote: http://cygwin.com/acronyms/#PCYMTNQREAIYR. Thanks. On Mon, 23 Jan 2006, Zach Gelnett wrote: Here is what i am seeing: [EMAIL PROTECTED] ~ the duplicate is the issue. Here is my PS1: \[\033]61;[EMAIL PROTECTED]@\H \W as you can see, only a single [snip] [snip] Also, ESC] doesn't look like a valid ANSI escape sequence. What exactly are you trying to accomplish with your prompt? [snip] If that was all you wanted, you could have set PS1 to '[EMAIL PROTECTED] \W'. Your prompt tries to do more than that. I'm still unsure of what the initial sequence of non-printing characters in your prompt is for. Those characters work with mrxvt to cause the tab name to have the appropriate [EMAIL PROTECTED] value. Ah. Ok, that explains it. Good to know. Thanks. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Cygwin Setup: Fatal Error: Uncaught Exception
Moving to cygwin-apps, as this is likely to get technical. On Mon, 23 Jan 2006, Brian Dessent wrote: Igor Peshansky wrote: I've looked at this a bit. Here's the weird part: the error says Uncaught Exception, but all the throws of that exception appear to be properly wrapped in try/catch blocks. So a simple change exception into an mbox kind of solution won't work here. More debugging is needed. The reason for the box is that the md5 checking was changed to run in a different thread than originally designed (now in the main thread instead of the download thread IIRC) and that thread does not have any particular catch handler for that exception, only the TOPLEVEL_CATCH which punts with the generic error. Do you mean packagemeta::ScanDownloadedFiles() calling packageversion::scan(), which calls check_for_cached()? Then yes, this isn't properly wrapped in a try/catch. I'd like to verify that this is the culprit (when someone sends me the corrupt tarball), but I think I see a proper fix for this. Will submit a patch shortly. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: Prompt issue within cygwin
On Mon, 23 Jan 2006, Zach Gelnett wrote: On 1/23/06, Zach Gelnett [EMAIL PROTECTED] wrote: On 1/23/06, Igor Peshansky [EMAIL PROTECTED] wrote: Again, http://cygwin.com/acronyms/#PCYMTNQREAIYR. Let's not feed the spammers. [snip] I do have one final question. If I set PS1 like this: PS1=\[\033]61;[EMAIL PROTECTED]:\w\\$ Everything works great, no issues at all, what is expected is what is displayed. Now, when i make only a slight modification as such: PS1=\[\033]61;[EMAIL PROTECTED]@\h:\w\\$ all the sudden i have a duplicate space after my prompt as in: [EMAIL PROTECTED]:~$ ls db2Mgr. that goes away when i hit the tab complete button to show me my options: [EMAIL PROTECTED]:~$ ls db2Mgr. db2Mgr.C db2Mgr.h [EMAIL PROTECTED]:~$ ls db2Mgr. I'm trying to get this prompt to work: PS1=\[\033]61;[EMAIL PROTECTED]@\H \W but the issue there is that the is duplicated (just like the space above, but much more noticable). Any ideas as to why making the title modification to use [EMAIL PROTECTED] instead of \w is causing these issues? There is a prompt bug in bash that causes it to miscount the number of displayed characters. One workaround was to append '\[\]' to PS1. Also, a good habit to get into is to use single quotes in the shell when some value contains backslashes. However, I can't reproduce your problem on my system (I only see one '' with the above PS1 setting). I don't believe you've posted the information requested in the problem reporting guidelines at http://cygwin.com/problems.html in this thread... Please attach the output of cygcheck -svr so that we can find out, among other things, what version of bash you are using, as well as what else is different in our setups. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: x-term font
On Mon, 23 Jan 2006, thad wrote: how can i change the font size of xterm? If you really mean xterm, then this question is off-topic for this list, and belongs on the cygwin-xfree list. I'm redirecting my reply there and setting the Reply-To: accordingly. You need to use the -fn option to pass the name/size combination for the font you want. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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: [Patch] g-b-s: Fix description of accept-patch.
On Mon, 23 Jan 2006, Buzz wrote: Op Sat, 21 Jan 2006 23:40:16 -0500 (EST) schreef Igor Peshansky in Pine.GSO.4.63.0601212337010.18707ataccess1.cims.nyu.edu: : On Sun, 22 Jan 2006, Buzz wrote: [fix help-text for accept-patch.] : I seem to have been derelict in my duties as maintainer. Applied, : though frankly, I don't find that description much more enlightening : than the original... Thanks anyway, and sorry for the delay. At least it is true. Thanks for checking this in. Any suggestion for a better text? Should ``maintainer-only''- functionality be flagged/grouped together in the help-text? That would be good. In fact, there were suggestions for moving maintainer-only features into a separate script, or leaving them in but only enabling them with the --maintainer-mode flag or something. I'm going through a clean-up of the logging functionality right now; when I check it in, I'll try to group features into maintainer-mode and user-mode, and we can go from there. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac
Re: RFC: [PATCH] Simplify installed.db format
On Sun, 22 Jan 2006, Igor Peshansky wrote:
The attached patch fixes the parsing of installed.db package lines.
Comments welcome.
Oops. That patch was from an outdated diff, please use the attached patch
instead. Sorry for the confusion. The updated ChangeLog is below.
Igor
==
2006-01-23 Igor Peshansky [EMAIL PROTECTED]
* cygpackage.h (filesize): Remove unused field.
(cygpackage::createInstance): Ditto.
* cygpackage.cc (cygpackage::createInstance): Ditto.
* package_db.cc (packagedb::packagedb): Remove unused variables.
Simplify parsing of installed.db.
--
http://cs.nyu.edu/~pechtcha/
|\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED]
ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!)
|,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski
'---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow!
Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte.
But no -- you are no fool; you call yourself a fool, there's proof enough in
that! -- Rostand, Cyrano de BergeracIndex: setup/cygpackage.cc
===
RCS file: /cvs/cygwin-apps/setup/cygpackage.cc,v
retrieving revision 2.15
diff -u -p -r2.15 cygpackage.cc
--- setup/cygpackage.cc 14 May 2005 15:57:29 - 2.15
+++ setup/cygpackage.cc 22 Jan 2006 05:18:25 -
@@ -66,7 +66,7 @@ cygpackage::createInstance (const String
packageversion
cygpackage::createInstance (String const pkgname, String const filename,
- size_t const fs, String const version,
+ String const version,
package_status_t const newstatus,
package_type_t const newtype)
{
@@ -75,7 +75,6 @@ cygpackage::createInstance (String const
temp-fn = filename;
temp-status = newstatus;
temp-type = newtype;
- temp-filesize = fs;
temp-setCanonicalVersion (version);
return packageversion(temp);
}
Index: setup/cygpackage.h
===
RCS file: /cvs/cygwin-apps/setup/cygpackage.h,v
retrieving revision 2.15
diff -u -p -r2.15 cygpackage.h
--- setup/cygpackage.h 14 May 2005 15:57:29 - 2.15
+++ setup/cygpackage.h 23 Jan 2006 04:23:13 -
@@ -72,7 +72,7 @@ public:
const package_type_t type);
static packageversion createInstance (String const , String const ,
- size_t const, String const ,
+ String const ,
package_status_t const,
package_type_t const);
@@ -92,7 +92,6 @@ private:
package_type_t type;
io_stream *listdata, *listfile;
- size_t filesize;
};
#endif /* SETUP_CYGPACKAGE_H */
Index: setup/package_db.cc
===
RCS file: /cvs/cygwin-apps/setup/package_db.cc,v
retrieving revision 2.30
diff -u -p -r2.30 package_db.cc
--- setup/package_db.cc 14 Oct 2005 04:10:26 - 2.30
+++ setup/package_db.cc 22 Jan 2006 05:18:26 -
@@ -57,8 +57,8 @@ packagedb::packagedb ()
if (!db)
return;
/* flush_local_db_package_data */
- char line[1000], pkgname[1000], inst[1000], src[1000];
- int instsz, srcsz;
+ char line[1000], pkgname[1000], inst[1000];
+ int instsz;
if (db-gets (line, 1000))
{
@@ -80,14 +80,10 @@ packagedb::packagedb ()
while (db-gets (line, 1000))
{
int parseable;
- src[0] = '\0';
pkgname[0] = '\0';
inst[0] = '\0';
- srcsz = 0;
- instsz = 0;
- sscanf (line, %s %s %d %s %d, pkgname, inst, instsz, src,
- srcsz);
+ sscanf (line, %s %s, pkgname, inst);
if (pkgname[0] == '\0' || inst[0] == '\0')
continue;
@@ -109,7 +105,7 @@ packagedb::packagedb ()
}
packageversion binary =
- cygpackage::createInstance (pkgname, inst, instsz, f.ver,
+ cygpackage::createInstance (pkgname, inst, f.ver,
package_installed,
package_binary);
Re: [Patch] g-b-s: Fix description of accept-patch.
On Sun, 22 Jan 2006, Buzz wrote: Op Tue, 6 Dec 2005 05:55:01 +0100 (MET) schreef Buzz in n2m-g.dn38qb.3vv5ill.1atbuzzy-box.bavag: : Op Tue, 22 Nov 2005 02:00:50 +0100 (MET) schreef Bas van Gompel [...] ::* templates/generic-build-script (help): fix help-text for ::accept-patch. : : Ping. Tracert. Destination host was unreachable. :-) I seem to have been derelict in my duties as maintainer. Applied, though frankly, I don't find that description much more enlightening than the original... Thanks anyway, and sorry for the delay. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac
RFC: [PATCH] Simplify installed.db format
Hi,
The current format for installed.db is (except the first line)
PKG-NAME TARBALL-NAME 0
which setup tries to parse as
package-name filename filesize src-filename srcsize
(with src-filename and srcsize never used, and the filesize field never
accessed). I propose to remove the unused fields, with the intention of
later reusing the (always 0) for something else.
The attached patch fixes the parsing of installed.db package lines.
Comments welcome.
Igor
==
ChangeLog:
2006-01-22 Igor Peshansky [EMAIL PROTECTED]
* cygpackage.h (filesize): Remove unused field.
* cygpackage.cc (cygpackage::createInstance): Ditto.
* package_db.cc (packagedb::packagedb): Remove unused variables.
Simplify parsing of installed.db.
--
http://cs.nyu.edu/~pechtcha/
|\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED]
ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!)
|,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski
'---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow!
Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte.
But no -- you are no fool; you call yourself a fool, there's proof enough in
that! -- Rostand, Cyrano de BergeracIndex: setup/cygpackage.cc
===
RCS file: /cvs/cygwin-apps/setup/cygpackage.cc,v
retrieving revision 2.15
diff -u -p -r2.15 cygpackage.cc
--- setup/cygpackage.cc 14 May 2005 15:57:29 - 2.15
+++ setup/cygpackage.cc 22 Jan 2006 05:18:25 -
@@ -66,7 +66,7 @@ cygpackage::createInstance (const String
packageversion
cygpackage::createInstance (String const pkgname, String const filename,
- size_t const fs, String const version,
+ String const version,
package_status_t const newstatus,
package_type_t const newtype)
{
@@ -75,7 +75,6 @@ cygpackage::createInstance (String const
temp-fn = filename;
temp-status = newstatus;
temp-type = newtype;
- temp-filesize = fs;
temp-setCanonicalVersion (version);
return packageversion(temp);
}
Index: setup/cygpackage.h
===
RCS file: /cvs/cygwin-apps/setup/cygpackage.h,v
retrieving revision 2.15
diff -u -p -r2.15 cygpackage.h
--- setup/cygpackage.h 14 May 2005 15:57:29 - 2.15
+++ setup/cygpackage.h 22 Jan 2006 05:18:25 -
@@ -92,7 +92,6 @@ private:
package_type_t type;
io_stream *listdata, *listfile;
- size_t filesize;
};
#endif /* SETUP_CYGPACKAGE_H */
Index: setup/package_db.cc
===
RCS file: /cvs/cygwin-apps/setup/package_db.cc,v
retrieving revision 2.30
diff -u -p -r2.30 package_db.cc
--- setup/package_db.cc 14 Oct 2005 04:10:26 - 2.30
+++ setup/package_db.cc 22 Jan 2006 05:18:26 -
@@ -57,8 +57,8 @@ packagedb::packagedb ()
if (!db)
return;
/* flush_local_db_package_data */
- char line[1000], pkgname[1000], inst[1000], src[1000];
- int instsz, srcsz;
+ char line[1000], pkgname[1000], inst[1000];
+ int instsz;
if (db-gets (line, 1000))
{
@@ -80,14 +80,10 @@ packagedb::packagedb ()
while (db-gets (line, 1000))
{
int parseable;
- src[0] = '\0';
pkgname[0] = '\0';
inst[0] = '\0';
- srcsz = 0;
- instsz = 0;
- sscanf (line, %s %s %d %s %d, pkgname, inst, instsz, src,
- srcsz);
+ sscanf (line, %s %s, pkgname, inst);
if (pkgname[0] == '\0' || inst[0] == '\0')
continue;
@@ -109,7 +105,7 @@ packagedb::packagedb ()
}
packageversion binary =
- cygpackage::createInstance (pkgname, inst, instsz, f.ver,
+ cygpackage::createInstance (pkgname, inst, f.ver,
package_installed,
package_binary);
Re: [PATCH] setup: initially expand categories starting with .
On Thu, 17 Nov 2005, Igor Pechtchanski wrote: (see http://cygwin.com/ml/cygwin-apps/2005-11/msg00141.html) 2005-11-16 Igor Peshansky [EMAIL PROTECTED] * PickView.cc (PickView::setViewMode): Auto-expand category if its name starts with a dot. Ping. Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac
Re: Downloading Cygwin under Linux ....
On Fri, 20 Jan 2006, Joshua Daniel Franklin wrote: On 1/20/06, Jan Nieuwenhuizen wrote: $ cyg-apt install lilypond to install: libXft1 libXft2 ghostscript-base libreadline6 findutils xorg-x11-bin pcre pango-runtime libncurses8 glib2-runtime pcre-doc libintl fontconfig xorg-x11-base xorg-x11-fnts openssl097 python xterm X-startup-scripts xorg-x11-libs-data libncurses7 pango libpng12 xorg-x11-etc ghostscript libguile12 freetype2 xorg-x11-bin-lndir zlib libfreetype26 terminfo libfontconfig1 termcap pcre-devel tar expat xorg-x11-fenc texinfo xorg-x11-xwin mktemp bzip2 crypt gettext coreutils libgdbm3 tcltk libintl1 libintl2 libintl3 run libbz2_1 libpcre0 xorg-x11-bin-dlls libiconv2 cygwin bash libXft diffutils openssl lilypond libdb4.2 _update-info-dir gzip glib2 base-passwd ... Jan, cyg-apt appears to be able to install packages on Cygwin as well as Linux. Do you know if it has regularly been used in such a manner? The biggest concern would be that if running under Cygwin it would not be able to install cygwin1.dll. While setup.exe is the only supported way of installing packages on Cygwin, I would like to add this information to the FAQ as another option for users who cannot use setup.exe to select packages (linux or blindness have come up several times). I haven't looked at it in detail: does cyg-apt properly run preremove and postinstall scripts? Mount/shortcut creation shouldn't be a problem if it's only used for updates. Besides, it can use the same replace-on-reboot mechanism as setup uses for in-use files (like python and cygwin1.dll). I know perl-libwin32 contains the API for accessing this mechanism in Windows -- is there a similar API for python? Igor -- http://cs.nyu.edu/~pechtcha/ |\ _,,,---,,_[EMAIL PROTECTED] | [EMAIL PROTECTED] ZZZzz /,`.-'`'-. ;-;;,_Igor Peshansky, Ph.D. (name changed!) |,4- ) )-,_. ,\ ( `'-' old name: Igor Pechtchanski '---''(_/--' `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-. Meow! Las! je suis sot... -Mais non, tu ne l'es pas, puisque tu t'en rends compte. But no -- you are no fool; you call yourself a fool, there's proof enough in that! -- Rostand, Cyrano de Bergerac -- 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/
