Re: cygwin stable and cvs snapshot - fork() bug

[Corinna Vinschen]

On Nov  1 10:58, Corinna Vinschen wrote:
> On Oct 31 14:26, Lev Bishop wrote:
> > $ cat lev.c && gcc -o lev lev.c -Wall -Wextra && CYGWIN=server ./lev
> > #include 
> > #include 
> > #include 
> > 
> > int main(void)
> > {
> > int shmid;
> > if ((shmid = shmget(IPC_PRIVATE, 100,IPC_CREAT | 0600 )) < 0 ||
> > !shmat(shmid, NULL, 0) ||
> > shmctl(shmid, IPC_RMID, NULL) < 0)
> > puts("problems with shm!");
> > fork();
> > }
> > lev.c: In function `main':
> > lev.c:13: warning: control reaches end of non-void function
> >   3 [main] lev 1924 c:\Documents and
> > Settings\Lev\Desktop\mpd-0.13.0\lev.exe: *** fatal error -
> > MapViewOfFileEx (0x3E), Win32 error 6.  Terminating.
> > 124 [main] lev 5076 fork: child 1924 - died waiting for dll
> > loading, errno 11
> 
> Thanks for the testcase.  I'm surprised that nobody experienced this
> problem before.  Sorta holiday here, so I'll look into it next week.

Ouch, ouch, ouch.  shmctl(IPC_RMID) closed the handle to the shared
memory, but neglected to remove the actual mappings as well as the
bookkeeping structure.  The result is that after a fork the child thinks
there are still mappings which have to be duplicated into its own
memory.  But the handle has already been closed in the parent, so the
MapViewOfFile call fails with "invalid handle".

Unfortunately not many applications use shmctl(IPC_RMID) before creating
a child process since usually the shared memory is meant to be... well,
shared.  That's why this didn't crop up more often, obviously.

I've applied a fix to CVS.


Thanks again for the testcase,
Corinna

-- 
Corinna Vinschen  Please, send mails regarding Cygwin to
Cygwin Project Co-Leader  cygwin AT cygwin DOT com
Red Hat

--
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/pubkey authentication and use of subst

[Corinna Vinschen]

On Nov  4 23:12, Greg M wrote:
> I'm running into this, too - albeit with a System error 59. I need to
> "interact with the desktop" too, which won't happen unless I run the sshd as
> System (although I couldn't find anything documenting this limitation beyond
> the cygrunsrv error message).

It's a windows restriction.  In Vista, you can forget the "interact with
desktop" mechanism entirely.  It has been dropped for security reasons.
You need to workaround it in your application.

>  Seems I can have any two of authorized_keys,
> desktop interaction and network file-systems, but unfortunately I need all
> three. :(

If you want your application to run with desktop interaction under Vista
as well, you will have to find another solution anyway.  There's nothing
Cygwin can do for you in this matter.

> So I'm wondering if this is just an accepted problem or something that could
> potentially be worked around in Cygwin. Or am I better off building a local
> workaround? (triggering commands via a polled database instead of ssh) It

In Cygwin's CVS is a new mechanism for getting passwordless logins(*),
but I don't know how far that helps for the network shares.  It certainly
doesn't help for desktop interaction.

What you could do is to run sshd under the user's account, so you get
pubkey auth + network share access.  For the desktop interaction you
should definitely think about another approach.  This has been discussed
a lot in relation to Vista, for instance on newsgroups and forums
related to Windows development.  The suggested approach is to start
a GUI process in the desktop session of the user and to use IPC to
communicate between the service and the GUI app.


Corinna

-- 
Corinna Vinschen  Please, send mails regarding Cygwin to
Cygwin Project Co-Leader  cygwin AT cygwin DOT com
Red Hat

--
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/

Customize setup program.

[João Pedro Oliveira]

Hi!

I am doing an installation program for my students and want to install 
cygwin as well.
I am using Nullsoft NSIS and I want to install only a minimum set (even 
smaller then the minimum recommended). In the previous setup program, I 
just copied a minimum installation from an installed version on my 
computer. I would like to go a bit further and I have this question:


Is it possible to call the setup program in quiet mode (or not quiet) 
and pass to the setup the packages we want to install? I would not like 
the user "lost" in all the steps of the cygwin setup...


How can I do this?

Best regards,
João

--
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: High CPU usage on posix_fallocate call - CVS version

[Corinna Vinschen]

On Nov  2 17:42, Rob Bosch wrote:
> I found on MSDN that Windows has the capfilealloc function in all Windows OS
> versions.  This function was designed to create a preallocated capture file
> of a specified size.  It seems like this would accomplish what cygwin needs
> with the posix_fallocate function?

Not at all.  Did you *read* what this method is about?

> -Original Message-
> From: Rob Bosch [mailto:[EMAIL PROTECTED] 

Even though this is your own mail address, please:

  http://cygwin.com/acronyms/#PCYMTNQREAIYR

  http://cygwin.com/acronyms/#TOFU

> When my compiled version of rsync is using the posix_fallocate function I’m
> getting significant CPU usage. [...]
> getting 20%-25% CPU utilization during the posix_fallocate call. [...]
> It also takes 20 minutes for the file to be created using this call.

I can't reproduce this behaviour.  I created a 40 Gigs file (haven't 77
Gigs left) on my XP box using the posix_fallocate function (which, btw.,
is basically just a single native NT call under the hood) and it takes
roughly a few milliseconds.

Did you look for http://cygwin.com/acronyms/#BLODA ?

http://cygwin.com/ml/cygwin-talk/2007-q4/msg00026.html


Corinna

-- 
Corinna Vinschen  Please, send mails regarding Cygwin to
Cygwin Project Co-Leader  cygwin AT cygwin DOT com
Red Hat

--
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: File permission hassles on Vista

[Corinna Vinschen]

On Nov  4 11:45, Joe Krahn wrote:
> I have been running Cygwin on Vista. The way that file attributes are 
> handled are causing problems. Cygwin appears to add attributes for 
> 'Everyone' and a 'None' user to handle world and group POSIX attributes. On 

'None' is the name of the primary group the user is member of in all
Windows NT versions if the user is not member in a domain.  If you don't
like it, you can change your primary group in Cygwin to, for instance,
the 'Users' group.  Just change it in your /etc/passwd.  And read
http://cygwin.com/cygwin-ug-net/ntsec.html, it might help a bit to
understand NT security.

> Vista, these attributes give the file a 'shared' property, which adds a 
> shared file overlay on the GUI icons. If you try to manipulate files from 
> the Windows GUI, things become VERY slow. It can take minutes just to move 

Sorry, I can't reproduce this.  I never heard of any Windows version
which would treat the existence of a None or Everyone ACE as a marker
for a shared file or directory.  I have Vista test machines.  When I
create files or directories on Vista with Cygwin, standard umask 022,
then none of the dirs or files are treated as shared byt the OS or
Windows Explorer (besides of the fact that only directories can be
shared).  The ACL of a dir created by Cygwin looks like this:

  $ mkdir foo
  $ cacls foo
  C:\cygwin\home\corinna\foo vmvista\corinna:F
 vmvista\None:R
 Everyone:R
 CREATOR OWNER:(OI)(CI)(IO)F
 CREATOR GROUP:(OI)(CI)(IO)R
 Everyone:(OI)(CI)(IO)R

A good hint (to me at least) is also the fact that the ACL doesn't
change, regardless if the directory is shared or not.

> I am guessing that Cygwin developers just are not using Vista yet, because 
> I have seen very little discussion of this problem.

You guessed wrong.

> Until this problem is 
> resolved, an easy fix might be to avoid adding the extra attributes groups 
> as long as they have no permissions. A Vista user can set umask to 077, and 
> avoid making those attribute groups.

What about figuring out what's really wrong?  It's not Cygwin per se,
otherwise I could easily reproduce it.  At the very least I need steps
to reproduce this effect.  Maybe it has something to do with your system
settings, or you're running some strange application which interacts
badly with certain ACL settings.

> Another quick work-around would be to have a command-line tool to strip out 
> the extra permission attributes.

cacls?


Corinna

-- 
Corinna Vinschen  Please, send mails regarding Cygwin to
Cygwin Project Co-Leader  cygwin AT cygwin DOT com
Red Hat

--
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/

1.5.24 -2 cygwin python 2.5.1 problem, "import sqlite3" fail

[Vincent Huang]

when i try to import sqlite3 in python, it results:
   ...
   File "/usr/lib/python2.5/sqlite3/dbapi2.py"
   from _sqlite3 import *
  ImportError: No module name _sqlite3

Actually, i can't find any modules named _sqlite3 in
/usr/lib/python2.5/.  Did anyone meet this befor?
How can i do? thanks
 "find /usr/lib/python2.5/  -print | grep _sqlite3" - find nothing

--
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: Customize setup program.

[Hussein Patwa]

Hi,

I've never really took that close of a look at the cygwin installer, but
I do know that all of the 'major' installers (installshield, nullsoft,
etc) usually have a /Q or /S switch for silent installs.  You would then
need to define a config file with all the parameters for the install. 
I'm thinking maybe you do the install manually, find the setup.ini (or
whichever) file is created with the install log and use that as a
starting point.

Someone will probably be along with a more definitive answer, but
hopefully this is somewhat in the ball park for what you're looking for.

Cheers,

H.

On Mon, 05 Nov 2007 11:07:50 +, "João Pedro Oliveira"
<[EMAIL PROTECTED]> said:
> Hi!
> 
> I am doing an installation program for my students and want to install 
> cygwin as well.
> I am using Nullsoft NSIS and I want to install only a minimum set (even 
> smaller then the minimum recommended). In the previous setup program, I 
> just copied a minimum installation from an installed version on my 
> computer. I would like to go a bit further and I have this question:
> 
> Is it possible to call the setup program in quiet mode (or not quiet) 
> and pass to the setup the packages we want to install? I would not like 
> the user "lost" in all the steps of the cygwin setup...
> 
> How can I do this?
> 
> Best regards,
> João
> 
-- 
  Hussein Patwa
  [EMAIL PROTECTED]

-- 
http://www.fastmail.fm - Email service worth paying for. Try it for free


--
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 stable and cvs snapshot - fork() bug

[Lev Bishop]

On 11/5/07, Corinna Vinschen  wrote:
> On Nov  1 10:58, Corinna Vinschen wrote:
> > On Oct 31 14:26, Lev Bishop wrote:
> > > $ cat lev.c && gcc -o lev lev.c -Wall -Wextra && CYGWIN=server ./lev
> > > #include 
> > > #include 
> > > #include 
> > >
> > > int main(void)
> > > {
> > > int shmid;
> > > if ((shmid = shmget(IPC_PRIVATE, 100,IPC_CREAT | 0600 )) < 0 ||
> > > !shmat(shmid, NULL, 0) ||
> > > shmctl(shmid, IPC_RMID, NULL) < 0)
> > > puts("problems with shm!");
> > > fork();
> > > }
> > > lev.c: In function `main':
> > > lev.c:13: warning: control reaches end of non-void function
> > >   3 [main] lev 1924 c:\Documents and
> > > Settings\Lev\Desktop\mpd-0.13.0\lev.exe: *** fatal error -
> > > MapViewOfFileEx (0x3E), Win32 error 6.  Terminating.
> > > 124 [main] lev 5076 fork: child 1924 - died waiting for dll
> > > loading, errno 11
> >
> > Thanks for the testcase.  I'm surprised that nobody experienced this
> > problem before.  Sorta holiday here, so I'll look into it next week.
>
> Ouch, ouch, ouch.  shmctl(IPC_RMID) closed the handle to the shared
> memory, but neglected to remove the actual mappings as well as the
> bookkeeping structure.  The result is that after a fork the child thinks
> there are still mappings which have to be duplicated into its own
> memory.  But the handle has already been closed in the parent, so the
> MapViewOfFile call fails with "invalid handle".
>
> Unfortunately not many applications use shmctl(IPC_RMID) before creating
> a child process since usually the shared memory is meant to be... well,
> shared.  That's why this didn't crop up more often, obviously.

Are you sure that you're interpreting IPC_RMID correctly? My
understanding is that you can still share the memory until you
actually remove the mapping. (Sort of like how you can unlink() a temp
file immediately after you open it, and continue to use it). I assumed
this was the reason for the create-map-remove pattern used by mpd.

From the linux man page:
   IPC_RMIDMark the segment to be destroyed.  The  segment  will  only
   actually  be  destroyed  after the last process detaches it
   (i.e., when the shm_nattch member of the associated  strucâ
   ture  shmid_ds  is  zero).  The caller must be the owner or
   creator, or be privileged.  If a segment  has  been  marked
   for  destruction,  then the (non-standard) SHM_DEST flag of
   the shm_perm.mode field in the  associated  data  structure
   retrieved by IPC_STAT will be set.

Lev

--
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: High CPU usage on posix_fallocate call - CVS version

[Rob Bosch]

> I can't reproduce this behaviour.  I created a 40 Gigs file (haven't 77
> Gigs left) on my XP box using the posix_fallocate function (which, btw.,
> is basically just a single native NT call under the hood) and it takes
> roughly a few milliseconds.

I checked the dodgy app lists.none of those are running.  The machine is
running Windows Server 2003 x64 R2 and nothing else besides ssh and rsync
under cygwin. I've posted to the rsync list so the developer that created
the patch can tell me if something else is going on.  I looked at the code
and it looks like a simple posix_fallocate call but you never know.  I'm
also asking if someone gets the same issue under a native Linux
implementation.

I'm not sure if it will make a difference but the disk array is separate
from the machine in that it is a direct-attached fiber array that appears as
a SCSI drive under Windows.


--
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 corrupt...

[Jon Seidel CMC]

The problem is related to Vista security: by default, cygwin tries to put
the home directory in c:/Users/, but this is a protected directory
(analogous to the old "Documents and Settings" directory in XP). Even as
that user, you can't get to much of that data!

To get around this problem, simply define the HOME variable BEFORE you
attempt to setup cygwin, pointing it to something less restricted (such as
c:/home/) and everything works just fine.

HTH...jon

--


-Original Message-
From: Jon Seidel CMC [mailto:[EMAIL PROTECTED] 
Sent: Saturday, November 03, 2007 9:12 AM
To: cygwin@cygwin.com
Subject: Installation corrupt...

I just downloaded/installed the latest cygwin (1.5.24-2) using 
cygwin.elite-systems.org, into a Vista Home Premium system (I know - ugh!)

I added a few packages to the default (such as vim, ImageMagick), but didn't

think it was much; I removed ash..?

Anyway, the installed completed without complaints in a reasonable amount of

time (none of the delay/hang problems reported on the list).

Almost nothing works when I start the Cygwin bash shell -- help, cd, pwd for

example work, but not much else:

bash-3.2$ ls *
bash: ls: command not found
>>>???

[EMAIL PROTECTED] /etc]$man -k ls
bash: man: command not found
>>>???

[EMAIL PROTECTED] /etc]$cygcheck -s -v -r > cygcheck.out
bash: cygcheck: command not found
>>> So I was unable to get the cygcheck output that is requested for this 
>>> list

[EMAIL PROTECTED] /etc]$echo $PATH
/cygdrive/c/usr/bin:/cygdrive/c/Windows/system32:/cygdrive/c/Windows:/cygdri
ve/c/Windows/System32/Wbem:/cygdrive/c/Program 
Files/Common Files/Roxio Shared/DLLShared/:/cygdrive/c/Program Files/Common 
Files/Roxio Shared/DLLShared/:/cygdrive/c/Program Files/Common Files/Roxio 
Shared/9.0/DLLShared/:/cygdrive/c/Program Files/QuickTime/QTSystem/
>>>The PATH option looks very strange... no /usr entries...

Viewing the setup log shows that all the post-install tasks ended with an 
abnormal exit code:
 exit code=126 or exit code=127

My login user is an Administrator

/Users/carol isn't visible in bash, but it's there when I use cmd.exe from 
DOS.

I tried running setup again, but this made no changes also re-ran it 
from the local directory (as suggested in one of the posts) and no change...

Any pointers? I've got cygwin up and running on two WinXP machines and it 
works just great!

Thanks...jon





--
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 stable and cvs snapshot - fork() bug

[Corinna Vinschen]

On Nov  5 07:43, Lev Bishop wrote:
> On 11/5/07, Corinna Vinschen  wrote:
> > Ouch, ouch, ouch.  shmctl(IPC_RMID) closed the handle to the shared
> > memory, but neglected to remove the actual mappings as well as the
> > bookkeeping structure.  The result is that after a fork the child thinks
> > there are still mappings which have to be duplicated into its own
> > memory.  But the handle has already been closed in the parent, so the
> > MapViewOfFile call fails with "invalid handle".
> >
> > Unfortunately not many applications use shmctl(IPC_RMID) before creating
> > a child process since usually the shared memory is meant to be... well,
> > shared.  That's why this didn't crop up more often, obviously.
> 
> Are you sure that you're interpreting IPC_RMID correctly? My
> understanding is that you can still share the memory until you
> actually remove the mapping. (Sort of like how you can unlink() a temp
> file immediately after you open it, and continue to use it). I assumed
> this was the reason for the create-map-remove pattern used by mpd.
> 
> From the linux man page:
>IPC_RMIDMark the segment to be destroyed.  The  segment  will  only
>actually  be  destroyed  after the last process detaches it
>(i.e., when the shm_nattch member of the associated  strucâ
>ture  shmid_ds  is  zero).  The caller must be the owner or
>creator, or be privileged.  If a segment  has  been  marked
>for  destruction,  then the (non-standard) SHM_DEST flag of
>the shm_perm.mode field in the  associated  data  structure
>retrieved by IPC_STAT will be set.

Well, when I created the patch, I had read the SUSv3 man page
http://www.opengroup.org/onlinepubs/009695399/functions/shmctl.html.
It says something different:

  IPC_RMID
Remove the shared memory identifier specified by shmid from the
system and destroy the shared memory segment and shmid_ds data
structure associated with it. IPC_RMID can only be executed by a
process that has an effective user ID equal to either that of a
process with appropriate privileges or to the value of shm_perm.cuid
or shm_perm.uid in the shmid_ds data structure associated with
shmid.

OTOH, the implementation within cygserver is more along the lines of
the Linux manpage.  Oh well.  I guess I'll have a look into implementing
it more Linux-like.


Corinna

-- 
Corinna Vinschen  Please, send mails regarding Cygwin to
Cygwin Project Co-Leader  cygwin AT cygwin DOT com
Red Hat

--
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: 1.5.24 -2 cygwin python 2.5.1 problem, "import sqlite3" fail

[Thorsten Kampe]

* Vincent Huang (Mon, 5 Nov 2007 20:06:40 +0800)
> when i try to import sqlite3 in python, it results:
>...
>File "/usr/lib/python2.5/sqlite3/dbapi2.py"
>from _sqlite3 import *
>   ImportError: No module name _sqlite3
> 
> Actually, i can't find any modules named _sqlite3 in
> /usr/lib/python2.5/.  Did anyone meet this befor?
> How can i do? thanks
>  "find /usr/lib/python2.5/  -print | grep _sqlite3" - find nothing

/usr/lib/python2.5/lib-dynload/_sqlite3.dll


--
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: setup.exe still can't uninstall bash

[Mario Frasca]

William Crosmun helped me out of this...

remove /etc/setup/bash.lst.gz (it appears to be corrupt).
rerun setup.exe

that was it.

thanks,
Mario


--
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 stable and cvs snapshot - fork() bug

[Lev Bishop]

On 11/5/07, Lev Bishop wrote:
> On 11/5/07, Corinna Vinschen  wrote:
> > On Nov  1 10:58, Corinna Vinschen wrote:
> > > On Oct 31 14:26, Lev Bishop wrote:
> > > > $ cat lev.c && gcc -o lev lev.c -Wall -Wextra && CYGWIN=server ./lev
> > > > #include 
> > > > #include 
> > > > #include 
> > > >
> > > > int main(void)
> > > > {
> > > > int shmid;
> > > > if ((shmid = shmget(IPC_PRIVATE, 100,IPC_CREAT | 0600 )) < 0 ||
> > > > !shmat(shmid, NULL, 0) ||
> > > > shmctl(shmid, IPC_RMID, NULL) < 0)
> > > > puts("problems with shm!");
> > > > fork();
> > > > }
> > > > lev.c: In function `main':
> > > > lev.c:13: warning: control reaches end of non-void function
> > > >   3 [main] lev 1924 c:\Documents and
> > > > Settings\Lev\Desktop\mpd-0.13.0\lev.exe: *** fatal error -
> > > > MapViewOfFileEx (0x3E), Win32 error 6.  Terminating.
> > > > 124 [main] lev 5076 fork: child 1924 - died waiting for dll
> > > > loading, errno 11
> > >
> > > Thanks for the testcase.  I'm surprised that nobody experienced this
> > > problem before.  Sorta holiday here, so I'll look into it next week.
> >
> > Ouch, ouch, ouch.  shmctl(IPC_RMID) closed the handle to the shared
> > memory, but neglected to remove the actual mappings as well as the
> > bookkeeping structure.  The result is that after a fork the child thinks
> > there are still mappings which have to be duplicated into its own
> > memory.  But the handle has already been closed in the parent, so the
> > MapViewOfFile call fails with "invalid handle".
> >
> > Unfortunately not many applications use shmctl(IPC_RMID) before creating
> > a child process since usually the shared memory is meant to be... well,
> > shared.  That's why this didn't crop up more often, obviously.
>
> Are you sure that you're interpreting IPC_RMID correctly? My
> understanding is that you can still share the memory until you
> actually remove the mapping. (Sort of like how you can unlink() a temp
> file immediately after you open it, and continue to use it). I assumed
> this was the reason for the create-map-remove pattern used by mpd.
>
> From the linux man page:
>IPC_RMIDMark the segment to be destroyed.  The  segment  will  only
>actually  be  destroyed  after the last process detaches it
>(i.e., when the shm_nattch member of the associated  strucâ
>ture  shmid_ds  is  zero).  The caller must be the owner or
>creator, or be privileged.  If a segment  has  been  marked
>for  destruction,  then the (non-standard) SHM_DEST flag of
>the shm_perm.mode field in the  associated  data  structure
>retrieved by IPC_STAT will be set.

It indeed seems this is behaviour not described in SuSv3. But several
unices support (some variant of) this behaviour. At least linux,
freebsd, hp-ux, solaris 10 mention it in their man pages, and openbsd
and netbsd seem to implement it that way even though they don't
describe it in the man pages.

FreeBSD:
 IPC_RMID Removes the segment from the system.  The removal will not
  take effect until all processes having attached the segment
  have exited; however, once the IPC_RMID operation has taken
  place, no further processes will be allowed to attach the
  segment.  For the operation to succeed, the calling
  process's effective uid must match shm_perm.uid or
  shm_perm.cuid, or the process must have superuser privi-
  leges.

HP-UX:
IPC_RMID
Remove the shared memory identifier specified by shmid from the system
and destroy the shared memory segment and data structure associated
with it. If the segment is attached to one or more processes, then the
segment key is changed to IPC_PRIVATE and the segment is marked
removed. The segment disappears when the last attached process
detaches it. This cmd can only be executed by a process that has an
effective user ID equal to either that of a user with appropriate
privileges or to the value of either shm_perm.uid or shm_perm.cuid in
the data structure associated with shmid.

Solaris 10:
IPC_RMID
Remove the shared memory identifier specified by shmid from the
system. The segment referenced by the identifier will be destroyed
when all processes with the segment attached have either detached the
segment or exited. If the segment is not attached to any process when
IPC_RMID is invoked, it will be destroyed immediately. This command
can be executed only by a process that has appropriate privileges or
an effective user ID equal to the value of shm_perm.cuid or
shm_perm.uid in the data structure associated with shmid.

A further linux extension: In addition to all the above, Linux goes
even further and still allows you to attach the segment even after
marking it for deletion. Linux ma

Re: cygwin compliation crash when uninstalling the old version of bash

[Mario Frasca]

William Crosmun wrote:

Mario - I've had the same problem for some time now. I finally determined to
track it down. I'll spare you the details, but I found that the file /etc/setup
/bash.lst.gz was corrupted. Using Windows Explorer, I could open the .gz file,
but I got an error opening the contained the .lst file contained therein. I
deleted /etc/setup/bash.lst.gz and was able to upgrade bash successfully.

Maybe this will work for you.

---
indeed it worked for me too.

[EMAIL PROTECTED]:~/Bureaublad$ tar tvzf /etc/setup/bash.lst.gz 

gzip: stdin: unexpected end of file
tar: Child returned status 1
tar: Error exit delayed from previous errors
[EMAIL PROTECTED]:~/Bureaublad$ rm !$
rm /etc/setup/bash.lst.gz

after this I could run Setup.exe without problems.

thanks,
Mario


--
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 compliation crash when uninstalling the old version of bash

[Dave Korn]

On 05 November 2007 14:48, Mario Frasca wrote:

> William Crosmun wrote:
> 
> Mario - I've had the same problem for some time now. I finally determined to
> track it down. I'll spare you the details, but I found that the file
> /etc/setup /bash.lst.gz was corrupted. Using Windows Explorer, I could open
> the .gz file, but I got an error opening the contained the .lst file
> contained therein. I deleted /etc/setup/bash.lst.gz and was able to upgrade
> bash successfully. 
> 
> Maybe this will work for you.
> 
> ---
> indeed it worked for me too.
> 
> [EMAIL PROTECTED]:~/Bureaublad$ tar tvzf /etc/setup/bash.lst.gz
> 
> gzip: stdin: unexpected end of file
> tar: Child returned status 1
> tar: Error exit delayed from previous errors
> [EMAIL PROTECTED]:~/Bureaublad$ rm !$
> rm /etc/setup/bash.lst.gz
> 
> after this I could run Setup.exe without problems.

  Augh!  There goes the last chance of finding out what happened!

  Next time, could somebody hang on to one of those files and perhaps offer it
on the list to anyone who wants to try and figure out in what way it got
corrupted?

cheers,
  DaveK
-- 
Can't think of a witty .sigline today


--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Re: cygwin stable and cvs snapshot - fork() bug

[Corinna Vinschen]

On Nov  5 09:49, Lev Bishop wrote:
> It indeed seems this is behaviour not described in SuSv3. But several
> unices support (some variant of) this behaviour. At least linux,
> freebsd, hp-ux, solaris 10 mention it in their man pages, and openbsd
> and netbsd seem to implement it that way even though they don't
> describe it in the man pages.

Yeah, we're using the FreeBSD code so the behaviour is already as
in Linux, as I mentioned in my previous mail.

> A further linux extension: In addition to all the above, Linux goes
> even further and still allows you to attach the segment even after
> marking it for deletion.
> [...]
> Freebsd (since version 5.2) has a sysctl kern.ipc.shm_allow_removed
> which seems to allow you to force the linux behaviour on this issue.
> Openbsd automatically does it (only) when running linux binaries via
> compat_linux(8).

Since we're using FreeBSD code, there's a variable shm_allow_removed in
the code already which allows this behaviour.  There's just no way right
now to set it.  It's always zero.  It would be quite easy to add a
cygserver.conf setting for this, though.

> If you do implement the behaviour of not destroying the segment until
> shm_nattach==0, you'll want to make sure that the shared memory key
> can be reused immediately after the old segment has been IPC_RMIDed,
> even though the old mapping may still be around. The other OS's which
> implement it seem to do this by having shmctl(IPC_RMID) change the key
> of the segment to be IPC_PRIVATE.

Since we're using FreeBSD code...

> Don't you just love standards

The best thing with standards is that we have so many of them.


Corinna

-- 
Corinna Vinschen  Please, send mails regarding Cygwin to
Cygwin Project Co-Leader  cygwin AT cygwin DOT com
Red Hat

--
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: 1.5.24 -2 cygwin python 2.5.1 problem, "import sqlite3" fail

[Vincent Huang]

thans to get your reply, but i can't found the _sqlite3.dll in the
/usr/lib/python2.5/lib-dynload/.  This is the packet list from:
http://cygwin.com/packages/python/python-2.5.1-2

Fri May 18 21:02:55 2007  10752
usr/lib/python2.5/lib-dynload/_random.dll
Fri May 18 21:02:55 2007  49664
usr/lib/python2.5/lib-dynload/_socket.dll
Fri May 18 21:02:55 2007  13824
usr/lib/python2.5/lib-dynload/_ssl.dll
Fri May 18 21:02:55 2007  21504
usr/lib/python2.5/lib-dynload/_struct.dll
Fri May 18 21:02:55 2007  19968
usr/lib/python2.5/lib-dynload/_testcapi.dll
Fri May 18 21:02:55 2007  36864
usr/lib/python2.5/lib-dynload/_tkinter.dll
Fri May 18 21:02:55 2007   6144
usr/lib/python2.5/lib-dynload/_weakref.dll


On 11/5/07, Thorsten Kampe <[EMAIL PROTECTED]> wrote:
> * Vincent Huang (Mon, 5 Nov 2007 20:06:40 +0800)
> > when i try to import sqlite3 in python, it results:
> >...
> >File "/usr/lib/python2.5/sqlite3/dbapi2.py"
> >from _sqlite3 import *
> >   ImportError: No module name _sqlite3
> >
> > Actually, i can't find any modules named _sqlite3 in
> > /usr/lib/python2.5/.  Did anyone meet this befor?
> > How can i do? thanks
> >  "find /usr/lib/python2.5/  -print | grep _sqlite3" - find nothing
>
> /usr/lib/python2.5/lib-dynload/_sqlite3.dll
>
>
> --
> 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/
>
>

--
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 stable and cvs snapshot - fork() bug

[Corinna Vinschen]

On Nov  5 16:17, Corinna Vinschen wrote:
> On Nov  5 09:49, Lev Bishop wrote:
> > It indeed seems this is behaviour not described in SuSv3. But several
> > unices support (some variant of) this behaviour. At least linux,
> > freebsd, hp-ux, solaris 10 mention it in their man pages, and openbsd
> > and netbsd seem to implement it that way even though they don't
> > describe it in the man pages.
> 
> Yeah, we're using the FreeBSD code so the behaviour is already as
> in Linux, as I mentioned in my previous mail.
> 
> > A further linux extension: In addition to all the above, Linux goes
> > even further and still allows you to attach the segment even after
> > marking it for deletion.
> > [...]
> > Freebsd (since version 5.2) has a sysctl kern.ipc.shm_allow_removed
> > which seems to allow you to force the linux behaviour on this issue.
> > Openbsd automatically does it (only) when running linux binaries via
> > compat_linux(8).
> 
> Since we're using FreeBSD code, there's a variable shm_allow_removed in
> the code already which allows this behaviour.  There's just no way right
> now to set it.  It's always zero.  It would be quite easy to add a
> cygserver.conf setting for this, though.

I've applied two patches which change shmctl(IPC_RMID) to the BSD/Linux
behaviour as well as adding a kern.ipc.shm_allow_removed option to the
/etc/cygserver.conf file.  Please give it a try.


Corinna

-- 
Corinna Vinschen  Please, send mails regarding Cygwin to
Cygwin Project Co-Leader  cygwin AT cygwin DOT com
Red Hat

--
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: Customize setup program.

[Tony Richardson]

João Pedro Oliveira  lx.it.pt> writes:
> Is it possible to call the setup program in quiet mode (or not quiet) 
> and pass to the setup the packages we want to install? I would not like 
> the user "lost" in all the steps of the cygwin setup...

You can define your own package in the setup.ini file that is
in the Base category and have it "require" the packages that 
you want automatically selected.

For example, I have a small file that looks like this:

@ UEBASE
sdesc: "Required packages for UE programming courses."
ldesc: "Required packages for UE programming courses."
category: Base
requires: a2ps gcc gcc-core gcc-g++ gcc-g77 gcc-gdc gcc-gpc gcc-java ...
version: 1.0.0-0

The requires field actually contains about 100 entries.  I just
append this file to the setup.ini file before running setup

   cat ue_setup.ini >> setup.ini

Then run setup and all of your desired packages are automatically
selected for installation.  It is not quite an unattended install,
but it is just a "click-thru" install.

You can get the initial list of required packages by setting up
one machine with the standard setup and then getting the list
of installed packages from the /etc/setup/installed.db file.

I tweak this from semester to semester to add and remove packages
before we re-image our lab machines.  More importantly, I provide
it to students to ensure that they have all of the necessary software
on their personal machines for software development.

Tony Richardson




--
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: 1.5.24 -2 cygwin python 2.5.1 problem, "import sqlite3" fail

[Thorsten Kampe]

* Vincent Huang (Mon, 5 Nov 2007 23:17:41 +0800)
> thans to get your reply, but i can't found the _sqlite3.dll in the
> /usr/lib/python2.5/lib-dynload/.  This is the packet list from:
> http://cygwin.com/packages/python/python-2.5.1-2

% cygcheck -f /usr/lib/python2.5/lib-dynload/_sqlite3.dll
python-2.5.1-3


--
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: 1.5.24 -2 cygwin python 2.5.1 problem, "import sqlite3" fail

[Dave Korn]

On 05 November 2007 16:21, Thorsten Kampe wrote:

> * Vincent Huang (Mon, 5 Nov 2007 23:17:41 +0800)
>> thans to get your reply, but i can't found the _sqlite3.dll in the
>> /usr/lib/python2.5/lib-dynload/.  This is the packet list from:
>> http://cygwin.com/packages/python/python-2.5.1-2
> 
> % cygcheck -f /usr/lib/python2.5/lib-dynload/_sqlite3.dll
> python-2.5.1-3

  N'existe-pas!

http://cygwin.com/packages/python/python-2.5.1-3
The requested URL /packages/python/python-2.5.1-3 was not found on this server.



cheers,
  DaveK
-- 
Can't think of a witty .sigline today


--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Re: 1.5.24 -2 cygwin python 2.5.1 problem, "import sqlite3" fail

[Jason Tishler]

Vincent,

On Mon, Nov 05, 2007 at 08:06:40PM +0800, Vincent Huang wrote:
> when i try to import sqlite3 in python, it results:
>...
>File "/usr/lib/python2.5/sqlite3/dbapi2.py"
>from _sqlite3 import *
>   ImportError: No module name _sqlite3

AFAICT, Python's _sqlite3 extension module is dependent on SQLite
 which is not part of the standard Cygwin
distribution.  This is why Cygwin Python does not contain _sqlite3.dll.

Jason

-- 
PGP/GPG Key: http://www.tishler.net/jason/pubkey.asc or key servers
Fingerprint: 7A73 1405 7F2B E669 C19D  8784 1AFD E4CC ECF4 8EF6

--
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: 1.5.24 -2 cygwin python 2.5.1 problem, "import sqlite3" fail

[Brian Dessent]

Dave Korn wrote:

> > % cygcheck -f /usr/lib/python2.5/lib-dynload/_sqlite3.dll
> > python-2.5.1-3
> 
>   N'existe-pas!
> 
> http://cygwin.com/packages/python/python-2.5.1-3
> The requested URL /packages/python/python-2.5.1-3 was not found on this 
> server.

Translation: He's using the python package from a third party (Cygwin
ports), which obviously doesn't do any good in the context of this
mailing list since it's off-topic, other than to say "you could try
using Cygwin ports", except that's not what he said, in which case it's
a pretty useless reply.

Brian

--
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/

Connect browser via ssh tunnel to local server ?

[Tony Benham]

Hi,
I have a question re connecting a browser over ssh to a local apache server?
I have an internal apache server used by local network users. This is not
exposed on the internet at all. I have a couple of outside linux users who have
ssh accounts on the apache server machine, so they can open a shell on that
machine. Is it possible for these users to connect via ssh and run a browser
from their machine and connect the apache server via the ssh connection in any
way ? (and ideally only connect to the apache server ?) If so canyone give me
some pointers as to how this could be done ?

Tony


--
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: Connect browser via ssh tunnel to local server ?

[Brian Dessent]

Tony Benham wrote:

> I have a question re connecting a browser over ssh to a local apache server?
> I have an internal apache server used by local network users. This is not
> exposed on the internet at all. I have a couple of outside linux users who 
> have
> ssh accounts on the apache server machine, so they can open a shell on that
> machine. Is it possible for these users to connect via ssh and run a browser
> from their machine and connect the apache server via the ssh connection in any
> way ? (and ideally only connect to the apache server ?) If so canyone give me
> some pointers as to how this could be done ?

Sure, just "ssh -D  server" and then configure the browser to use a
socks proxy on localhost:, where  is some available local port
number.  Now all browser traffic goes through the tunnel, and the
internal site can be accessed in whatever way it would if you were
sitting at the console of the server.

But what's this got to do with Cygwin?

Brian

--
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: Connect browser via ssh tunnel to local server ?

[Tony Benham]

Tony Benham  benhamhouse.co.uk> writes:
> Hi,
> I have a question re connecting a browser over ssh to a local apache
> server? I have an internal apache server used by local network users.
> This is not exposed on the internet at all. I have a couple of outside
> linux users who have ssh accounts on the apache server machine, so
> they can open a shell on that machine. 
..
I forgot to say the apache server, and ssh under cygwin run on a windows
machine. 





--
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: Connect browser via ssh tunnel to local server ?

[Tony Benham]

Brian Dessent  dessent.net> writes:

> Sure, just "ssh -D  server" and then configure the browser to use a
> socks proxy on localhost:, where  is some available local port
> number.  Now all browser traffic goes through the tunnel, and the
> internal site can be accessed in whatever way it would if you were
> sitting at the console of the server.
> 
> But what's this got to do with Cygwin?
> 
> Brian
> 
> 
I forgot to say the machine that hosts the apache server is a windows machine,
that runs ssh under cygwin, so I use the cygwin/ssh to accept the remote
connections from the linux machines.   




--
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/

setup.exe version issues

[Dat Head]

if you run setup.exe and it finds a .ini file that doesn't match it asks
that you go to http://cygwin.com/setup and download a newer one,
however, the version on there is 2.510.2.2 and if you go to main
cygwin.com page and get setup.exe it is 2.573.2.2

also has any thought ever been given to renaming the file cygwin-setup.exe
or does it need to adhere to 8.3 naming convention?

p.s. also note when subscribing to cygwin (main) maillist it does not anywhere
in the body tell you to use cygwin@cygwin.com to submit messages

--
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: setup.exe version issues

[Dave Korn]

On 05 November 2007 17:34, Dat Head wrote:

> p.s. also note when subscribing to cygwin (main) maillist it does not
> anywhere in the body tell you to use [EMAIL PROTECTED] to submit
> messages


  It does describe the list over and over and over again as "the cygwin (AT)
cygwin (DOT) com" mailing list.  Do you *really* need a paragraph added that
says

"To mail the cygwin (AT) cygwin (DOT) com mailing list, send mail to cygwin
(AT) cygwin (DOT) com"

... ?


cheers,
  DaveK
-- 
Can't think of a witty .sigline today


--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Wish Setup would accept my Perl

[Michael Kairys]

I've been using ActiveState Perl and Cygwin together for years and have only 
one complaint: Setup keeps hassling me about dependencies and I have to make 
sure always to uncheck Perl in the setu list (and again in the dependency 
check) or I end up with two Perl installations, which I don't want.

I wish Setup could recognize that I already have a Perl installation...


--
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: Wish Setup would accept my Perl

[Brian Mathis]

On 11/5/07, Michael Kairys <[EMAIL PROTECTED]> wrote:
> I've been using ActiveState Perl and Cygwin together for years and have only
> one complaint: Setup keeps hassling me about dependencies and I have to make
> sure always to uncheck Perl in the setu list (and again in the dependency
> check) or I end up with two Perl installations, which I don't want.
>
> I wish Setup could recognize that I already have a Perl installation...
>

The answer is "Don't do that".  Perl on cygwin and ActiveState Perl
are completely different things.  You've been using them together for
years [without problems], [except for this one big problem].

ActiveState Perl is a Windows app and should only be accessed through
Windows, such as the cmd.exe window or by using the .pl file
extension.  If you think the cmd.exe window sucks, try console2 on
sourceforge.  It's just lipstick on a pig, but at least it looks
better.

If you want to interoperate with other utilities that are provided by
cygwin, use the cygwin perl and not ActiveState.

Having both on your machine should work just fine, as long as you
don't try to share anything between them.  Treat them as completely
different things (which they are).

--
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: 1.5.24 -2 cygwin python 2.5.1 problem, "import sqlite3" fail

[Thorsten Kampe]

* Brian Dessent (Mon, 05 Nov 2007 08:44:46 -0800)
> Dave Korn wrote:
> > > % cygcheck -f /usr/lib/python2.5/lib-dynload/_sqlite3.dll
> > > python-2.5.1-3
> > 
> >   N'existe-pas!
> > 
> > http://cygwin.com/packages/python/python-2.5.1-3
> > The requested URL /packages/python/python-2.5.1-3 was not found on this 
> > server.
> 
> Translation: He's using the python package from a third party (Cygwin
> ports), which obviously doesn't do any good in the context of this
> mailing list since it's off-topic, other than to say "you could try
> using Cygwin ports", except that's not what he said, in which case it's
> a pretty useless reply.

Actually, I forgot that Cygwin Ports forcefully installs a new Python.


--
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: Wish Setup would accept my Perl

[Warren Young]

Michael Kairys wrote:

I wish Setup could recognize that I already have a Perl installation...


The two may be compatible at the Perl source level, for the most 
part[*], but I doubt they're binary compatible.  A lot of Perl modules 
contain parts compiled from C and linked into the interpreter 
dynamically at need.


Unless you can mix these modules, I don't think you can expect Cygwin to 
consider ActiveState's distribution compatible.  If a Cygwin package 
needs Perl, it's fair for setup.exe to demand a Cygwinnish Perl to 
support it.  For all setup.exe knows, it's an absolute requirement.


[*] I also doubt they're 100% compatible at the source level, what with 
the various extensions each could be enabling that the other doesn't.


--
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: 1.5.24 -2 cygwin python 2.5.1 problem, "import sqlite3" fail

[Thorsten Kampe]

* Jason Tishler (Mon, 05 Nov 2007 11:19:13 -0500)
> On Mon, Nov 05, 2007 at 08:06:40PM +0800, Vincent Huang wrote:
> > when i try to import sqlite3 in python, it results:
> >...
> >File "/usr/lib/python2.5/sqlite3/dbapi2.py"
> >from _sqlite3 import *
> >   ImportError: No module name _sqlite3
> 
> AFAICT, Python's _sqlite3 extension module is dependent on SQLite
>  which is not part of the standard Cygwin
> distribution.  This is why Cygwin Python does not contain _sqlite3.dll.

Well, but it should because it's a part of a standard Python 
installation. See "New, Improved, and Removed Modules":
"If you're compiling the Python source yourself, note that the source 
tree doesn't include the SQLite code, only the wrapper module. You'll 
need to have the SQLite libraries and headers installed before 
compiling Python, and the build process will compile the module when 
the necessary headers are available."

Thorsten


--
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: 1.5.24 -2 cygwin python 2.5.1 problem, "import sqlite3" fail

[Jason Tishler]

On Mon, Nov 05, 2007 at 08:06:37PM -, Thorsten Kampe wrote:
> * Jason Tishler (Mon, 05 Nov 2007 11:19:13 -0500)
> > [snip]
> > AFAICT, Python's _sqlite3 extension module is dependent on SQLite
> >  which is not part of the standard Cygwin
> > distribution.  This is why Cygwin Python does not contain
> > _sqlite3.dll.
> 
> Well, but it should because it's a part of a standard Python
> installation. See "New, Improved, and Removed Modules": "If you're
> compiling the Python source yourself, note that the source tree
> doesn't include the SQLite code, only the wrapper module. You'll need
> to have the SQLite libraries and headers installed before compiling
> Python, and the build process will compile the module when the
> necessary headers are available."

Understood, but as I indicated above, SQLite is not part of the standard
Cygwin distribution.  If someone contributes SQLite, then I will release
Cygwin Python with SQLite support enabled.

Jason

-- 
PGP/GPG Key: http://www.tishler.net/jason/pubkey.asc or key servers
Fingerprint: 7A73 1405 7F2B E669 C19D  8784 1AFD E4CC ECF4 8EF6

--
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: 1.5.24 -2 cygwin python 2.5.1 problem, "import sqlite3" fail

[Thorsten Kampe]

* Jason Tishler (Mon, 05 Nov 2007 16:16:17 -0500)
> On Mon, Nov 05, 2007 at 08:06:37PM -, Thorsten Kampe wrote:
> > * Jason Tishler (Mon, 05 Nov 2007 11:19:13 -0500)
> > > [snip]
> > > AFAICT, Python's _sqlite3 extension module is dependent on SQLite
> > >  which is not part of the standard Cygwin
> > > distribution.  This is why Cygwin Python does not contain
> > > _sqlite3.dll.
> > 
> > Well, but it should because it's a part of a standard Python
> > installation. See "New, Improved, and Removed Modules": "If you're
> > compiling the Python source yourself, note that the source tree
> > doesn't include the SQLite code, only the wrapper module. You'll need
> > to have the SQLite libraries and headers installed before compiling
> > Python, and the build process will compile the module when the
> > necessary headers are available."
> 
> Understood, but as I indicated above, SQLite is not part of the standard
> Cygwin distribution.  If someone contributes SQLite, then I will release
> Cygwin Python with SQLite support enabled.

ftp://sunsite.dk/projects/cygwinports/release/sqlite3/ ?


--
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: 1.5.24 -2 cygwin python 2.5.1 problem, "import sqlite3" fail

[Brian Dessent]

Thorsten Kampe wrote:

> ftp://sunsite.dk/projects/cygwinports/release/sqlite3/ ?

There are zillions of things that are in Cygwin Ports that aren't in
Cygwin.  What's that got to do with anything?  It's irrelevant, until
Yaakov or someone else contributes and maintains it for Cygwin proper.

Brian

--
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: 1.5.24 -2 cygwin python 2.5.1 problem, "import sqlite3" fail

[Thorsten Kampe]

* Brian Dessent (Mon, 05 Nov 2007 13:51:46 -0800)
> Thorsten Kampe wrote:
> > ftp://sunsite.dk/projects/cygwinports/release/sqlite3/ ?
> 
> There are zillions of things that are in Cygwin Ports that aren't in
> Cygwin.  What's that got to do with anything?  It's irrelevant, until
> Yaakov or someone else contributes and maintains it for Cygwin proper.

As far as I understood Jason, he just needs the SQLite headers and 
libraries to /build/ Python with SQLite support, and he doesn't need 
them to be "in Cygwin".

Thorsten


--
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: 1.5.24 -2 cygwin python 2.5.1 problem, "import sqlite3" fail

[Brian Dessent]

Thorsten Kampe wrote:

> As far as I understood Jason, he just needs the SQLite headers and
> libraries to /build/ Python with SQLite support, and he doesn't need
> them to be "in Cygwin".

That would be silly, because it would result in a totally broken python
that would continue to throw ImportError on "import sqlite3" because the
user would have no cygsqlite3-0.dll on their system, with no way to get
it unless they installed third party packages that aren't part of
Cygwin.  Besides, requiring packages not in the distro in order to build
a package is not kosher.  The distro must be self-consistent.

Brian

--
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: 1.5.24 -2 cygwin python 2.5.1 problem, "import sqlite3" fail

[Thorsten Kampe]

* Brian Dessent (Mon, 05 Nov 2007 15:07:57 -0800)
> Thorsten Kampe wrote:
> > As far as I understood Jason, he just needs the SQLite headers and
> > libraries to /build/ Python with SQLite support, and he doesn't need
> > them to be "in Cygwin".
> 
> That would be silly, because it would result in a totally broken python
> that would continue to throw ImportError on "import sqlite3" because the
> user would have no cygsqlite3-0.dll on their system, with no way to get
> it unless they installed third party packages that aren't part of
> Cygwin.

If the Windows Python installation comes with more-or-less third party 
sqlite3.dll, then Cygwin Python coming with a third party cygsqlite3-
0.dll (from Cygwin Ports) would be fine - at least with me...

Anyway: no one ever noticed the lack of SQLite for quite some time and 
the other ones probably found workarounds...

Thorsten


--
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: 1.5.24 -2 cygwin python 2.5.1 problem, "import sqlite3" fail

[Christopher Faylor]

On Mon, Nov 05, 2007 at 03:07:57PM -0800, Brian Dessent wrote:
>Thorsten Kampe wrote:
>>As far as I understood Jason, he just needs the SQLite headers and
>>libraries to /build/ Python with SQLite support, and he doesn't need
>>them to be "in Cygwin".
>
>That would be silly, because it would result in a totally broken python
>that would continue to throw ImportError on "import sqlite3" because
>the user would have no cygsqlite3-0.dll on their system, with no way to
>get it unless they installed third party packages that aren't part of
>Cygwin.  Besides, requiring packages not in the distro in order to
>build a package is not kosher.  The distro must be self-consistent.

Yes.  What he said.  Jason understands what it means to add this support.
Thorsten does not.  Unless someone wants to support SQLite3 (which would
be a welcome addition to Cygwin), I think this thread has gone on for
about six messages longer than it should have.

cgf

--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Re: 1.5.24 -2 cygwin python 2.5.1 problem, "import sqlite3" fail

[Christopher Faylor]

On Mon, Nov 05, 2007 at 11:44:27PM -, Thorsten Kampe wrote:
>* Brian Dessent (Mon, 05 Nov 2007 15:07:57 -0800)
>> Thorsten Kampe wrote:
>> > As far as I understood Jason, he just needs the SQLite headers and
>> > libraries to /build/ Python with SQLite support, and he doesn't need
>> > them to be "in Cygwin".
>> 
>> That would be silly, because it would result in a totally broken python
>> that would continue to throw ImportError on "import sqlite3" because the
>> user would have no cygsqlite3-0.dll on their system, with no way to get
>> it unless they installed third party packages that aren't part of
>> Cygwin.
>
>If the Windows Python installation comes with more-or-less third party 
>sqlite3.dll, then Cygwin Python coming with a third party cygsqlite3-
>0.dll (from Cygwin Ports) would be fine - at least with me...
>
>Anyway: no one ever noticed the lack of SQLite for quite some time and 
>the other ones probably found workarounds...

Thorsten:  You can stop digging now.  The hole is deep enough.

--
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/