Re: Cygwin participation threshold

Re: Cygwin participation threshold

[Christopher G. Faylor]

>From: Steve Morris <smorris@nexen.com>
>Date: Fri, 26 Feb 1999 11:55:50 -0500 (EST)
>
>In all this discussion something important is being lost. Cris
>bemoaned the lack of development support for cygwin and asked for
>reasons. I and others tried to explain where we think the issues
>are. Inevitably this comes out sounding negative, but at least on my
>part, this is not intended. Maybe we took Cris' questions too
>literally. Flogging Cygnus was not the intent. We were trying to offer
>legitimate feedback to a legitimate question.
>
>Let me reiterate that Cygnus is clearly one of the Good Guys. The best
>guys are the Cygnus employees (like Cris) who volunteer their own time
>to this project.
>
>Many of us are rooting for Cygnus and are hoping more companies figure
>out how to make money on free software; because they then tend to give
>back. As an example gcc and gdb have been in much better shape all
>these years since Cygnus became the official release site.

Thanks very much for this positive note.

I was hoping that my original message would generate a lot of discussion
because I thought that it would give DJ and myself an opportunity to
explain something about Cygnus/Cygwin as well as providing a venue
for everyone else to provide their outlook.

It may not be obvious but I do appreciate any and all feedback on this
issue.  It certainly helps educate me on people's perception of my
company and my product.  I do want to understand what people think
about this.

That doesn't mean that I won't argue with perceptions that I think
are wrong but I certainly respect everyone's right to an opinion.

Again, thanks to you and to the handful of other people who sent me
private supportive email.  I appreciate it.

cgf

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

mySQL, mSQL, PostgreSQL with cygwin ?

[Sebastien Barre]

I was wondering if anyone of you did successfully build/install/test one of
these products with Cygwin32 :

	- mySQL
	- mSQL
	- PostgreSQL

Thanks :)

(I already found something related to mSQL at the Cygwin Porting Projet
http://www.student.uni-koeln.de/cygwin/Distribution/Source/msql/ , but it
seems not complete). 

______________________________________________________________
Sebastien Barre                  http://www.hds.utc.fr/~barre/

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

mySQL, mSQL, PostgreSQL with cygwin ?

[Sebastien Barre]

I was wondering if anyone of you did successfully build/install/test one of
these products with Cygwin32 :

	- mySQL
	- mSQL
	- PostgreSQL

Thanks :)

(I already found something related to mSQL at the Cygwin Porting Projet
http://www.student.uni-koeln.de/cygwin/Distribution/Source/msql/ , but it
seems not complete). 

______________________________________________________________
Sebastien Barre                  http://www.hds.utc.fr/~barre/

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: mySQL, mSQL, PostgreSQL with cygwin ?

[Stipe Tolj]

> I was wondering if anyone of you did successfully build/install/test one of
> these products with Cygwin32 :
>
>         - mySQL
>         - mSQL
>         - PostgreSQL
>
> Thanks :)
>
> (I already found something related to mSQL at the Cygwin Porting Projet
> http://www.student.uni-koeln.de/cygwin/Distribution/Source/msql/ , but it
> seems not complete).

sorry, the mSQL 2.0.4.1 source port is outdated and non-stable.

A contributed stable port of mSQL 2.0.5 may be found within the camp-1.1b
(Cygwin Apache mSQL PHP3) embedded web authoring environment in binary form at

    http://www.student.uni-koeln.de/cygwin/Distribution/Binary

A contributed PostgreSQL 6.4 port in binary form is also available at this
location, but it's only supposed to run on NT.

MySQL is _not_ ported to Cygwin due to the lack of the pthreads functions used
in the multi-user multi-threaded MySQL database engine. Currently Cygwin is
getting at least experimental support for pthreads functions, but I suppose it
will take some time to get a stable port of MySQL done to Cygwin.

There is a native Win32 shareware version of MySQL 3.21.29a which works
perfectly with our Apache 1.3.4 and integrated PHP 3.0.6, check out MySQL's
official web site.

Regards,
Stipe

--
Stipe Tolj <tolj@uni-duesseldorf.de>

Cygwin Porting Project -- "We build UNIX on top of Windows"
http://www.student.uni-koeln.de/cygwin/

Department of Economical Computer Science
University of Cologne, Germany




--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: mySQL, mSQL, PostgreSQL with cygwin ?

[Sebastien Barre]

At 07:59 01/03/99 +0100, Stipe Tolj wrote:

>A contributed PostgreSQL 6.4 port in binary form is also available at this
>location, but it's only supposed to run on NT

Very good. I also downloaded the 6.4.2 source and will try to compile it.
Too bad there isn't a diff file at your location, but if I fail, I will
install the binary, for sure.

>MySQL is _not_ ported to Cygwin due to the lack of the pthreads functions used
>in the multi-user multi-threaded MySQL database engine. Currently Cygwin is
>getting at least experimental support for pthreads functions, but I suppose it
>will take some time to get a stable port of MySQL done to Cygwin.

Thanks for the information, but there is something very obscure to me, they
ARE binaries for cygwin32 :
http://www.tcx.se/download.html

"MySQL 3.22.10 clients compiled with the cygwin32 kit. This includes
'mysql' with readline editing capabilities. " and a link to
http://www.tcx.se/Downloads/Win32/mysqlclient-3.22.10b-cygwin-b20.tar.gz

I downloaded it, here is the content :

-rw-r--r-- tolj/None    723652 1998-11-08 17:59 cygwin1.dll
-rwxr-xr-x tolj/None     41472 1998-11-17 19:40 insert_test.exe
-rwxr-xr-x tolj/None    192512 1998-11-17 19:40 mysql.exe
-rwxr-xr-x tolj/None     55296 1998-11-17 19:40 mysqladmin.exe
-rwxr-xr-x tolj/None     62464 1998-11-17 19:40 mysqldump.exe
-rwxr-xr-x tolj/None     53760 1998-11-17 19:40 mysqlimport.exe
-rwxr-xr-x tolj/None     53760 1998-11-17 19:40 mysqlshow.exe
-rwxr-xr-x tolj/None     41472 1998-11-17 19:40 select_test.exe

Which seems to mean that the mySQL DLL was succesfully built (how ??), and
linked or added to the original cygwin1.dll ? If it's true, it may remove
my coolview support, am I wrong ? 

Furthermore, you have to get a license to use it officialy.

>There is a native Win32 shareware version of MySQL 3.21.29a which works
>perfectly with our Apache 1.3.4 and integrated PHP 3.0.6, check out MySQL's
>official web site.

I do not know if I'll be able to use it with Perl built for cygwin ? If I
compile a DBI module for MySQL within my cygwin perl context, the DLL
formats won't match I guess.

Regards


______________________________________________________________
Sebastien Barre                  http://www.hds.utc.fr/~barre/

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

[HELP] stat(), file permission, r/w access : i'm LOST :(

[Sebastien Barre]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 4449 bytes --]

Dear cygwin people

As some of you might have noticed (sorry for the flood), I'm still trying
to arbitrate The Fight between Perl 5.002 and cygwin32-B20.1 (the
battlefield being NT4/SP4). Well, thanks to some of you, I nearly did it
(and also ruined a couple of nights), but a strange problem lead me to some
more confusion :

Problem : although I'm working as single user on a standalone workstation,
and always logged as sysadmin ("administrateur" in french), cygwin seems to
see/create the files on my hard disk as *not* belonging to me, but to
another "being" (which is not obvious to me). 

I'll do my best to describe it (it's really *easy* to observe) : let's have
a look at a simple Perl example, then a short C example with stat(), and my
/etc/passwd and /etc/group. Finally I will create a simple file with
'touch', and you will notice that it will be reported as NOT belonging to
me and not being writable :((

Note : this seems specific to NT, which is not surprising regarding file
permissions.


Perl example :
----------------------

Note : I created /etc/passwd and /etc/group with 'mkpasswd -l' and 'mkgroup
-l' respectively, as reported in many FAQS.

administrateur [26] /etc$ perl -e 'print "yes" if -w "group";'

administrateur [27] /etc$ ll
total 7
-rw-r--r--   1 544 Aucun          27 Feb 26 02:45 group
-rw-r--r--   1 544 Aucun         65 Mar  1 21:08 passwd
-rw-r--r--   1 544 Aucun        9828 Dec  1 14:00 termcap

That command (-w) will display 'yes' if the file (here, /etc/group) is
writable. Apparently, it fails. And fails everywhere in my filesystem. The
'-w' command is using C<stat>, let's move on :


C example :
----------------------
#include <sys/stat.h>
#include <stdlib.h>

int main() {
    struct stat mystat;

    stat("group", &mystat);

    printf("mode : %o, uid : %u, gid : %u\n", 
           mystat.st_mode, mystat.st_uid, mystat.st_gid);

    exit(0);
}

Run :

administrateur [52] /etc$ gcc mystat.c

administrateur [53] /etc$ a.exe
mode : 100644, uid : 544, gid : 513

administrateur [55] /etc$ ll group
-rw-r--r--   1 544 Aucun          27 Feb 26 02:45 group

Which means : 
	- /etc/group belongs to user which UID is 544, and to group which GID is 513
	- it's a regular file (100000), 
	- read/write permission to owner (600), read permission to group (40) and
read to other (4).

BUT (and it drives me crazy). I CREATED that file, and I'm NOT user 544 !
(more about this strange fellow below).

administrateur [56] /etc$ whoami
administrateur

administrateur [74] /etc$ cat passwd
Administrateur::500:513:seb::/bin/sh
InvitÚ::501:513:::/bin/sh

administrateur [75] /etc$ cat group
Aucun::513:
Everyone::0:

Obviously, I'm user 500 (I changed my name to 'seb' so that to check). 
=> Therefore WHY are all files created as 544 (even with tar) ? (the group
is correct : 513).

I thought it was related to /etc/passwd, but I just do NOT UNDERSTAND the
difference between -l and -g option for 'mkpasswd'.

   -l,--local              print local accounts
   -g,--local-groups       print local group information too

administrateur [77] /etc$ mkpasswd -l
Administrateur::500:513:seb::/bin/sh
InvitÚ::501:513:::/bin/sh

administrateur [79] /etc$ mkpasswd -g
Administrateurs::544:0:::
Duplicateurs::552:0:::
InvitÚs::546:0:::
OpÚrateurs de sauvegarde::551:0:::
Utilisateurs::545:0:::
Utilisateurs avec pouvoir::547:0:::

=> who are these users ?!? these should be groups !! I'm belonging to the
"Administrateurs" group for NT (in the "Gestionnaire d'utilisateurs" ~=
"User manager"), and this has been translated to a user, I'm completely
lost :((

=> dumping 'mkpasswd -l -g' to /etc/passwd (instead of 'mkpasswd -l') did
NOT help.

Here is the same joke :

administrateur [84] /etc$ touch test

administrateur [85] /etc$ ll test
-rw-r--r--   1 544      Aucun           0 Mar  1 21:46 test

administrateur [86] /etc$ perl -e 'print "yes" if -w "test";'
administrateur [87] /etc$

WOAH : I created a file, and it's automatically assigned to someone else !
Give me my file back please :( 
And of course, although I created it, it's now NOT writable for cygwin.


I guess I might be just dumb. Any help would be really appreciated (by my
mental health).

Thanks






______________________________________________________________
Sebastien Barre                  http://www.hds.utc.fr/~barre/

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[Larry Hall (RFK Partners, Inc)]

At 10:14 PM 3/1/99 +0100, Sebastien Barre wrote:
>administrateur [27] /etc$ ll
>total 7
>-rw-r--r--   1 544 Aucun          27 Feb 26 02:45 group
>-rw-r--r--   1 544 Aucun         65 Mar  1 21:08 passwd
>-rw-r--r--   1 544 Aucun        9828 Dec  1 14:00 termcap
>
>That command (-w) will display 'yes' if the file (here, /etc/group) is
>writable. Apparently, it fails. And fails everywhere in my filesystem. The
>'-w' command is using C<stat>, let's move on :


This is a known bug.  testing for file permissions will not return "yes"
unless that permission is given to everyone.  This may be related to the
fact that you are the "Administrator" (see below).


>BUT (and it drives me crazy). I CREATED that file, and I'm NOT user 544 !
>(more about this strange fellow below).


544 is the "Administrator" group, which is used as the UID of everyone in
that group.  This has been discussed many times on this list and is an NTism.



Larry Hall                             lhall@rfk.com
RFK Partners, Inc.                     (781) 239-1053
8 Grove Street                         (781) 239-1655
Wellesley, MA, 02482-7797              http://www.rfk.com

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[Sebastien Barre]

At 16:38 01/03/99 -0500, Larry Hall (RFK Partners, Inc) wrote:

>This is a known bug.

Arghh.

>testing for file permissions will not return "yes"
>unless that permission is given to everyone.  This may be related to the
>fact that you are the "Administrator" (see below).

>544 is the "Administrator" group, which is used as the UID of everyone in
>that group.

But shall be used as GID instead.

> This has been discussed many times on this list and is an NTism.

Is there a workaround ? I mean, if this is a know bug, is there a way to
fix it in the cygwin sources ?

Thanks for your help

______________________________________________________________
Sebastien Barre                  http://www.hds.utc.fr/~barre/

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[Larry Hall (RFK Partners, Inc)]

At 11:27 PM 3/1/99 +0100, Sebastien Barre wrote:
>> This has been discussed many times on this list and is an NTism.
>
>Is there a workaround ? I mean, if this is a know bug, is there a way to
>fix it in the cygwin sources ?
>

Workarounds?  You might try removing yourself from Administrators group.
This may do the trick.  Alternatively, you can change your umask so that 
write permision is given to everyone.  That should solve it too.  

Fixing the bug in the source is also possible.  I know Corinna has been 
doing some work with making permissions track more closely with UNIX style.
I'm not sure whether his changes will help in this arena...


Larry Hall                             lhall@rfk.com
RFK Partners, Inc.                     (781) 239-1053
8 Grove Street                         (781) 239-1655
Wellesley, MA, 02482-7797              http://www.rfk.com

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[Sebastien Barre]

At 17:43 01/03/99 -0500, you wrote:

>>Is there a workaround ? I mean, if this is a know bug, is there a way to
>>fix it in the cygwin sources ?

>Workarounds?  You might try removing yourself from Administrators group.

Which seems a bit risky.

>This may do the trick.  Alternatively, you can change your umask so that 
>write permision is given to everyone.

That sounds interesting, could you please be more specific, I'm not exactly
a umask guru, and I have no man page. Thanks a lot for any help.

>Fixing the bug in the source is also possible.  I know Corinna has been 
>doing some work with making permissions track more closely with UNIX style.
>I'm not sure whether his changes will help in this arena...

I do not want to put pressure on the cygwin team if I can solve it myself
in a efficent way, they already do a great job.

Sincerely


______________________________________________________________
Sebastien Barre                  http://www.hds.utc.fr/~barre/

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[Heinz-Jürgen Oertel]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 2596 bytes --]

Sebastien Barre schrieb:
> 
> At 17:43 01/03/99 -0500, you wrote:
> 
>
> That sounds interesting, could you please be more specific, I'm not exactly
> a umask guru, and I have no man page. Thanks a lot for any help.
> 

       umask [-S] [mask]
              Display  or  set the file permission creation mask,
              or umask (see umask(2)).  If the -S option is used,
              the mask displayed or set is symbolic, otherwise it
              is an octal number.

              Symbolic masks are like those used by chmod(1):
                     [ugoa]{{=+-}{rwx}*}+[,...]
              in which the first group of characters is  the  who
              part, the second group is the op part, and the last
              group is the perm part.   The  who  part  specifies
              which  part  of  the  umask is to be modified.  The
              letters mean:
                      u      the user permissions

                      g      the group permissions

                      o      the  other  permissions   (non-user,
                             non-group)

                      a      all  permissions  (user,  group  and
                             other)

              The op part indicates how the who  permissions  are
              to be modified:

                      =      set

                      +      added to

                      -      removed from
              The perm part specifies which permissions are to be
              set, added or removed:

                      r      read permission

                      w      write permission

                      x      execute permission

              When symbolic masks are used,  they  describe  what
              permissions  may  be  made available (as opposed to
              octal masks in which a set  bit  means  the  corre­
              sponding   bit   is   to   be  cleared).   Example:
              `ug=rwx,o=' sets the mask  so  files  will  not  be
              readable,  writable  or executable by `others', and
              is equivalent (on most systems) to the  octal  mask
              `07'.



-- 
with best regards / mit freundlichen Grüßen

  Heinz-Jürgen Oertel

==========================================
Heinz-Juergen Oertel
port GmbH            phone +49 3493 743-10
Antonienstr. 3       fax   +49 3493 743-15
D-06749 Bitterfeld   mailto:service@port.de
Germany              http://www.port.de
==========================================

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[Heinz-Jürgen Oertel]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 2597 bytes --]

Sebastien Barre schrieb:
> 
> At 17:43 01/03/99 -0500, you wrote:
> 
>
> That sounds interesting, could you please be more specific, I'm not exactly
> a umask guru, and I have no man page. Thanks a lot for any help.
> 

       umask [-S] [mask]
              Display  or  set the file permission creation mask,
              or umask (see umask(2)).  If the -S option is used,
              the mask displayed or set is symbolic, otherwise it
              is an octal number.

              Symbolic masks are like those used by chmod(1):
                     [ugoa]{{=+-}{rwx}*}+[,...]
              in which the first group of characters is  the  who
              part, the second group is the op part, and the last
              group is the perm part.   The  who  part  specifies
              which  part  of  the  umask is to be modified.  The
              letters mean:
                      u      the user permissions

                      g      the group permissions

                      o      the  other  permissions   (non-user,
                             non-group)

                      a      all  permissions  (user,  group  and
                             other)

              The op part indicates how the who  permissions  are
              to be modified:

                      =      set

                      +      added to

                      -      removed from
              The perm part specifies which permissions are to be
              set, added or removed:

                      r      read permission

                      w      write permission

                      x      execute permission

              When symbolic masks are used,  they  describe  what
              permissions  may  be  made available (as opposed to
              octal masks in which a set  bit  means  the  corre­
              sponding   bit   is   to   be  cleared).   Example:
              `ug=rwx,o=' sets the mask  so  files  will  not  be
              readable,  writable  or executable by `others', and
              is equivalent (on most systems) to the  octal  mask
              `07'.



-- 
with best regards / mit freundlichen Grüßen

  Heinz-Jürgen Oertel

==========================================
Heinz-Juergen Oertel
port GmbH            phone +49 3493 743-10
Antonienstr. 3       fax   +49 3493 743-15
D-06749 Bitterfeld   mailto:service@port.de
Germany              http://www.port.de
==========================================

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[Geoffrey Noer]

On Tue, Mar 02, 1999, Sebastien Barre wrote:
[...]
> >Fixing the bug in the source is also possible.  I know Corinna has been 
> >doing some work with making permissions track more closely with UNIX style.
[...]

Note that these changes aren't being ignored.  I'm trying to stabilize
my build environment again (pesky internal compiler errors, sigh) and
then I'll be looking into/testing her changes.  Hopefully we'll be
adding them soon to the development sources...

-- 
Geoffrey Noer		Email: noer@cygnus.com
Cygnus Solutions

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[Geoffrey Noer]

On Tue, Mar 02, 1999, Sebastien Barre wrote:
[...]
> >Fixing the bug in the source is also possible.  I know Corinna has been 
> >doing some work with making permissions track more closely with UNIX style.
[...]

Note that these changes aren't being ignored.  I'm trying to stabilize
my build environment again (pesky internal compiler errors, sigh) and
then I'll be looking into/testing her changes.  Hopefully we'll be
adding them soon to the development sources...

-- 
Geoffrey Noer		Email: noer@cygnus.com
Cygnus Solutions

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[Sebastien Barre]

At 17:43 01/03/99 -0500, you wrote:

>>Is there a workaround ? I mean, if this is a know bug, is there a way to
>>fix it in the cygwin sources ?

>Workarounds?  You might try removing yourself from Administrators group.

Which seems a bit risky.

>This may do the trick.  Alternatively, you can change your umask so that 
>write permision is given to everyone.

That sounds interesting, could you please be more specific, I'm not exactly
a umask guru, and I have no man page. Thanks a lot for any help.

>Fixing the bug in the source is also possible.  I know Corinna has been 
>doing some work with making permissions track more closely with UNIX style.
>I'm not sure whether his changes will help in this arena...

I do not want to put pressure on the cygwin team if I can solve it myself
in a efficent way, they already do a great job.

Sincerely


______________________________________________________________
Sebastien Barre                  http://www.hds.utc.fr/~barre/

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[Lassi A. Tuura]

On Mon, 1 Mar 1999, Larry Hall (RFK Partners, Inc) wrote:
|> Fixing the bug in the source is also possible.  I know Corinna has been 
|> doing some work with making permissions track more closely with UNIX style.
|> I'm not sure whether his changes will help in this arena...

If somebody is going to fix this, I would strongly encourage to fix it
in a way that uses `access' to determine file permissions, not something
based on `stat'.

This would have the benefit of making things work on the AFS filesystem
as well, where using getuid and st_uid (or similar) to determine
accessibility is meaningless: AFS uses ACLs and tokens that determine
access rights, and the application has no way to know either of these
unless it links against the AFS/Kerberos libraries.  Please make the
scheme trust the operating system (or network file system deamons), and
not to build additional logic that fails with ACL-based systems.  For
example, GNU test program has this bug -- it depends on `stat' instead
of `access'. 

Presumably Win32 system calls responds like AFS with ACLs -- call the
right function (`access'?) and it will tell you whether you can access
the file or not.  Alternatively, `stat' should use the security API to
fill in the st_mode fields correctly, but I am not sure this will work
with networked file systems that implement their own security rules.

Cheers,
//lat
--
With sufficient thrust, pigs fly just fine.  However, this is not
necessarily a good idea.  It is hard to be sure where they are going
to land, and it could be dangerous sitting under them as they fly
overhead.  --RFC1925, "The Twelve Networking Truths"


--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[Lassi A. Tuura]

On Mon, 1 Mar 1999, Larry Hall (RFK Partners, Inc) wrote:
|> Fixing the bug in the source is also possible.  I know Corinna has been 
|> doing some work with making permissions track more closely with UNIX style.
|> I'm not sure whether his changes will help in this arena...

If somebody is going to fix this, I would strongly encourage to fix it
in a way that uses `access' to determine file permissions, not something
based on `stat'.

This would have the benefit of making things work on the AFS filesystem
as well, where using getuid and st_uid (or similar) to determine
accessibility is meaningless: AFS uses ACLs and tokens that determine
access rights, and the application has no way to know either of these
unless it links against the AFS/Kerberos libraries.  Please make the
scheme trust the operating system (or network file system deamons), and
not to build additional logic that fails with ACL-based systems.  For
example, GNU test program has this bug -- it depends on `stat' instead
of `access'. 

Presumably Win32 system calls responds like AFS with ACLs -- call the
right function (`access'?) and it will tell you whether you can access
the file or not.  Alternatively, `stat' should use the security API to
fill in the st_mode fields correctly, but I am not sure this will work
with networked file systems that implement their own security rules.

Cheers,
//lat
--
With sufficient thrust, pigs fly just fine.  However, this is not
necessarily a good idea.  It is hard to be sure where they are going
to land, and it could be dangerous sitting under them as they fly
overhead.  --RFC1925, "The Twelve Networking Truths"


--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[Sebastien Barre]

At 17:43 01/03/99 -0500, Larry Hall (RFK Partners, Inc) wrote:
>At 11:27 PM 3/1/99 +0100, Sebastien Barre wrote:
>>> This has been discussed many times on this list and is an NTism.
>>
>>Is there a workaround ? I mean, if this is a know bug, is there a way to
>>fix it in the cygwin sources ?

>Workarounds?  You might try removing yourself from Administrators group.
>This may do the trick.  Alternatively, you can change your umask so that 
>write permision is given to everyone.  That should solve it too.  

Well, write permission was *already* given to anyone :

administrateur [21] ~$ umask
000
administrateur [22] ~$ umask -S
u=rwx,g=rwx,o=rwx

and this had no influence :

administrateur [23] ~$ touch foo
administrateur [24] ~$ ll foo
-rw-r--r--   1 544      Aucun           0 Mar  5 21:46 foo
administrateur [26] ~$ perl -e 'print "yes" if -w "foo";'
administrateur [27] ~$



______________________________________________________________
Sebastien Barre                  http://www.hds.utc.fr/~barre/

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[Larry Hall]

At 09:47 PM 3/5/99 +0100, Sebastien Barre wrote:
>At 17:43 01/03/99 -0500, Larry Hall (RFK Partners, Inc) wrote:
>>At 11:27 PM 3/1/99 +0100, Sebastien Barre wrote:
>>>> This has been discussed many times on this list and is an NTism.
>>>
>>>Is there a workaround ? I mean, if this is a know bug, is there a way to
>>>fix it in the cygwin sources ?
>
>>Workarounds?  You might try removing yourself from Administrators group.
>>This may do the trick.  Alternatively, you can change your umask so that 
>>write permision is given to everyone.  That should solve it too.  
>
>Well, write permission was *already* given to anyone :
>
>administrateur [21] ~$ umask
>000
>administrateur [22] ~$ umask -S
>u=rwx,g=rwx,o=rwx
>
>and this had no influence :
>
>administrateur [23] ~$ touch foo
>administrateur [24] ~$ ll foo
>-rw-r--r--   1 544      Aucun           0 Mar  5 21:46 foo
>administrateur [26] ~$ perl -e 'print "yes" if -w "foo";'
>administrateur [27] ~$
>
>


I think I mentioned this to you before but I believe I initially left out
that you need to set "ntea" in your CYGWIN environment variable for the 
permissions to explicitly show up via ls and friends.  Be sure to check the
FAQ/archives about the issues with using "ntea" and FAT partitions however...




Larry Hall                              lhall@rfk.com
RFK Partners, Inc.                      (781) 239-1053
8 Grove Street                          (781) 239-1655 - FAX
Wellesley, MA  02482-7797               http://www.rfk.com

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[Larry Hall]

At 09:47 PM 3/5/99 +0100, Sebastien Barre wrote:
>At 17:43 01/03/99 -0500, Larry Hall (RFK Partners, Inc) wrote:
>>At 11:27 PM 3/1/99 +0100, Sebastien Barre wrote:
>>>> This has been discussed many times on this list and is an NTism.
>>>
>>>Is there a workaround ? I mean, if this is a know bug, is there a way to
>>>fix it in the cygwin sources ?
>
>>Workarounds?  You might try removing yourself from Administrators group.
>>This may do the trick.  Alternatively, you can change your umask so that 
>>write permision is given to everyone.  That should solve it too.  
>
>Well, write permission was *already* given to anyone :
>
>administrateur [21] ~$ umask
>000
>administrateur [22] ~$ umask -S
>u=rwx,g=rwx,o=rwx
>
>and this had no influence :
>
>administrateur [23] ~$ touch foo
>administrateur [24] ~$ ll foo
>-rw-r--r--   1 544      Aucun           0 Mar  5 21:46 foo
>administrateur [26] ~$ perl -e 'print "yes" if -w "foo";'
>administrateur [27] ~$
>
>


I think I mentioned this to you before but I believe I initially left out
that you need to set "ntea" in your CYGWIN environment variable for the 
permissions to explicitly show up via ls and friends.  Be sure to check the
FAQ/archives about the issues with using "ntea" and FAT partitions however...




Larry Hall                              lhall@rfk.com
RFK Partners, Inc.                      (781) 239-1053
8 Grove Street                          (781) 239-1655 - FAX
Wellesley, MA  02482-7797               http://www.rfk.com

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[Sebastien Barre]

At 17:43 01/03/99 -0500, Larry Hall (RFK Partners, Inc) wrote:
>At 11:27 PM 3/1/99 +0100, Sebastien Barre wrote:
>>> This has been discussed many times on this list and is an NTism.
>>
>>Is there a workaround ? I mean, if this is a know bug, is there a way to
>>fix it in the cygwin sources ?

>Workarounds?  You might try removing yourself from Administrators group.
>This may do the trick.  Alternatively, you can change your umask so that 
>write permision is given to everyone.  That should solve it too.  

Well, write permission was *already* given to anyone :

administrateur [21] ~$ umask
000
administrateur [22] ~$ umask -S
u=rwx,g=rwx,o=rwx

and this had no influence :

administrateur [23] ~$ touch foo
administrateur [24] ~$ ll foo
-rw-r--r--   1 544      Aucun           0 Mar  5 21:46 foo
administrateur [26] ~$ perl -e 'print "yes" if -w "foo";'
administrateur [27] ~$



______________________________________________________________
Sebastien Barre                  http://www.hds.utc.fr/~barre/

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[Larry Hall (RFK Partners, Inc)]

At 11:27 PM 3/1/99 +0100, Sebastien Barre wrote:
>> This has been discussed many times on this list and is an NTism.
>
>Is there a workaround ? I mean, if this is a know bug, is there a way to
>fix it in the cygwin sources ?
>

Workarounds?  You might try removing yourself from Administrators group.
This may do the trick.  Alternatively, you can change your umask so that 
write permision is given to everyone.  That should solve it too.  

Fixing the bug in the source is also possible.  I know Corinna has been 
doing some work with making permissions track more closely with UNIX style.
I'm not sure whether his changes will help in this arena...


Larry Hall                             lhall@rfk.com
RFK Partners, Inc.                     (781) 239-1053
8 Grove Street                         (781) 239-1655
Wellesley, MA, 02482-7797              http://www.rfk.com

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[Sebastien Barre]

At 16:38 01/03/99 -0500, Larry Hall (RFK Partners, Inc) wrote:

>This is a known bug.

Arghh.

>testing for file permissions will not return "yes"
>unless that permission is given to everyone.  This may be related to the
>fact that you are the "Administrator" (see below).

>544 is the "Administrator" group, which is used as the UID of everyone in
>that group.

But shall be used as GID instead.

> This has been discussed many times on this list and is an NTism.

Is there a workaround ? I mean, if this is a know bug, is there a way to
fix it in the cygwin sources ?

Thanks for your help

______________________________________________________________
Sebastien Barre                  http://www.hds.utc.fr/~barre/

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[Larry Hall (RFK Partners, Inc)]

At 10:14 PM 3/1/99 +0100, Sebastien Barre wrote:
>administrateur [27] /etc$ ll
>total 7
>-rw-r--r--   1 544 Aucun          27 Feb 26 02:45 group
>-rw-r--r--   1 544 Aucun         65 Mar  1 21:08 passwd
>-rw-r--r--   1 544 Aucun        9828 Dec  1 14:00 termcap
>
>That command (-w) will display 'yes' if the file (here, /etc/group) is
>writable. Apparently, it fails. And fails everywhere in my filesystem. The
>'-w' command is using C<stat>, let's move on :


This is a known bug.  testing for file permissions will not return "yes"
unless that permission is given to everyone.  This may be related to the
fact that you are the "Administrator" (see below).


>BUT (and it drives me crazy). I CREATED that file, and I'm NOT user 544 !
>(more about this strange fellow below).


544 is the "Administrator" group, which is used as the UID of everyone in
that group.  This has been discussed many times on this list and is an NTism.



Larry Hall                             lhall@rfk.com
RFK Partners, Inc.                     (781) 239-1053
8 Grove Street                         (781) 239-1655
Wellesley, MA, 02482-7797              http://www.rfk.com

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[John Fortin]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 5297 bytes --]

I think I said this before...  I had the same problem.  It has to do ( it seems ) with
being part of the administrator group.  NT assigns the owner of the file to 544 if you
are part of the administrator group.

Try this if you like.  Create a new user and set it up like your normal on EXCEPT don't
make it part of the adminstrator group.  You may need to set up some different
directories to mirror what you have now but owned by the new userid.  Try your build
there.

Sebastien Barre wrote:

> Dear cygwin people
>
> As some of you might have noticed (sorry for the flood), I'm still trying
> to arbitrate The Fight between Perl 5.002 and cygwin32-B20.1 (the
> battlefield being NT4/SP4). Well, thanks to some of you, I nearly did it
> (and also ruined a couple of nights), but a strange problem lead me to some
> more confusion :
>
> Problem : although I'm working as single user on a standalone workstation,
> and always logged as sysadmin ("administrateur" in french), cygwin seems to
> see/create the files on my hard disk as *not* belonging to me, but to
> another "being" (which is not obvious to me).
>
> I'll do my best to describe it (it's really *easy* to observe) : let's have
> a look at a simple Perl example, then a short C example with stat(), and my
> /etc/passwd and /etc/group. Finally I will create a simple file with
> 'touch', and you will notice that it will be reported as NOT belonging to
> me and not being writable :((
>
> Note : this seems specific to NT, which is not surprising regarding file
> permissions.
>
> Perl example :
> ----------------------
>
> Note : I created /etc/passwd and /etc/group with 'mkpasswd -l' and 'mkgroup
> -l' respectively, as reported in many FAQS.
>
> administrateur [26] /etc$ perl -e 'print "yes" if -w "group";'
>
> administrateur [27] /etc$ ll
> total 7
> -rw-r--r--   1 544 Aucun          27 Feb 26 02:45 group
> -rw-r--r--   1 544 Aucun         65 Mar  1 21:08 passwd
> -rw-r--r--   1 544 Aucun        9828 Dec  1 14:00 termcap
>
> That command (-w) will display 'yes' if the file (here, /etc/group) is
> writable. Apparently, it fails. And fails everywhere in my filesystem. The
> '-w' command is using C<stat>, let's move on :
>
> C example :
> ----------------------
> #include <sys/stat.h>
> #include <stdlib.h>
>
> int main() {
>     struct stat mystat;
>
>     stat("group", &mystat);
>
>     printf("mode : %o, uid : %u, gid : %u\n",
>            mystat.st_mode, mystat.st_uid, mystat.st_gid);
>
>     exit(0);
> }
>
> Run :
>
> administrateur [52] /etc$ gcc mystat.c
>
> administrateur [53] /etc$ a.exe
> mode : 100644, uid : 544, gid : 513
>
> administrateur [55] /etc$ ll group
> -rw-r--r--   1 544 Aucun          27 Feb 26 02:45 group
>
> Which means :
>         - /etc/group belongs to user which UID is 544, and to group which GID is 513
>         - it's a regular file (100000),
>         - read/write permission to owner (600), read permission to group (40) and
> read to other (4).
>
> BUT (and it drives me crazy). I CREATED that file, and I'm NOT user 544 !
> (more about this strange fellow below).
>
> administrateur [56] /etc$ whoami
> administrateur
>
> administrateur [74] /etc$ cat passwd
> Administrateur::500:513:seb::/bin/sh
> InvitÚ::501:513:::/bin/sh
>
> administrateur [75] /etc$ cat group
> Aucun::513:
> Everyone::0:
>
> Obviously, I'm user 500 (I changed my name to 'seb' so that to check).
> => Therefore WHY are all files created as 544 (even with tar) ? (the group
> is correct : 513).
>
> I thought it was related to /etc/passwd, but I just do NOT UNDERSTAND the
> difference between -l and -g option for 'mkpasswd'.
>
>    -l,--local              print local accounts
>    -g,--local-groups       print local group information too
>
> administrateur [77] /etc$ mkpasswd -l
> Administrateur::500:513:seb::/bin/sh
> InvitÚ::501:513:::/bin/sh
>
> administrateur [79] /etc$ mkpasswd -g
> Administrateurs::544:0:::
> Duplicateurs::552:0:::
> InvitÚs::546:0:::
> OpÚrateurs de sauvegarde::551:0:::
> Utilisateurs::545:0:::
> Utilisateurs avec pouvoir::547:0:::
>
> => who are these users ?!? these should be groups !! I'm belonging to the
> "Administrateurs" group for NT (in the "Gestionnaire d'utilisateurs" ~=
> "User manager"), and this has been translated to a user, I'm completely
> lost :((
>
> => dumping 'mkpasswd -l -g' to /etc/passwd (instead of 'mkpasswd -l') did
> NOT help.
>
> Here is the same joke :
>
> administrateur [84] /etc$ touch test
>
> administrateur [85] /etc$ ll test
> -rw-r--r--   1 544      Aucun           0 Mar  1 21:46 test
>
> administrateur [86] /etc$ perl -e 'print "yes" if -w "test";'
> administrateur [87] /etc$
>
> WOAH : I created a file, and it's automatically assigned to someone else !
> Give me my file back please :(
> And of course, although I created it, it's now NOT writable for cygwin.
>
> I guess I might be just dumb. Any help would be really appreciated (by my
> mental health).
>
> Thanks
>
> ______________________________________________________________
> Sebastien Barre                  http://www.hds.utc.fr/~barre/
>
> --
> Want to unsubscribe from this list?
> Send a message to cygwin-unsubscribe@sourceware.cygnus.com




--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[Sebastien Barre]

At 16:49 01/03/99 -0500, you wrote:

>I think I said this before...

Sorry, I'm new, and the search feature is down :(

>I had the same problem.  It has to do ( it 
>seems ) with being part of the administrator group.  NT assigns the owner
of the file to 
>544 if you are part of the administrator group.

I see. Thanks to Microsoft ?

>Try this if you like.  Create a new user and set it up like your normal on 
>EXCEPT don't
>make it part of the adminstrator group.  You may need to set up some different
>directories to mirror what you have now but owned by the new userid.  Try 
>your build
>there.

All right. But in fact the build is no more the problem : the problem is
the behaviour of perl regarding this bug in "everyday life."

Thanx for your help

______________________________________________________________
Sebastien Barre                  http://www.hds.utc.fr/~barre/

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[DJ Delorie]

> Sorry, I'm new, and the search feature is down :(

I've arranged to have the search link point to
http://www.delorie.com/archives/ until the sourceware engine can be
fixed.


--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[DJ Delorie]

> Sorry, I'm new, and the search feature is down :(

I've arranged to have the search link point to
http://www.delorie.com/archives/ until the sourceware engine can be
fixed.


--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[Sebastien Barre]

At 16:49 01/03/99 -0500, you wrote:

>I think I said this before...

Sorry, I'm new, and the search feature is down :(

>I had the same problem.  It has to do ( it 
>seems ) with being part of the administrator group.  NT assigns the owner
of the file to 
>544 if you are part of the administrator group.

I see. Thanks to Microsoft ?

>Try this if you like.  Create a new user and set it up like your normal on 
>EXCEPT don't
>make it part of the adminstrator group.  You may need to set up some different
>directories to mirror what you have now but owned by the new userid.  Try 
>your build
>there.

All right. But in fact the build is no more the problem : the problem is
the behaviour of perl regarding this bug in "everyday life."

Thanx for your help

______________________________________________________________
Sebastien Barre                  http://www.hds.utc.fr/~barre/

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: [HELP] stat(), file permission, r/w access : i'm LOST :(

[John Fortin]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 5298 bytes --]

I think I said this before...  I had the same problem.  It has to do ( it seems ) with
being part of the administrator group.  NT assigns the owner of the file to 544 if you
are part of the administrator group.

Try this if you like.  Create a new user and set it up like your normal on EXCEPT don't
make it part of the adminstrator group.  You may need to set up some different
directories to mirror what you have now but owned by the new userid.  Try your build
there.

Sebastien Barre wrote:

> Dear cygwin people
>
> As some of you might have noticed (sorry for the flood), I'm still trying
> to arbitrate The Fight between Perl 5.002 and cygwin32-B20.1 (the
> battlefield being NT4/SP4). Well, thanks to some of you, I nearly did it
> (and also ruined a couple of nights), but a strange problem lead me to some
> more confusion :
>
> Problem : although I'm working as single user on a standalone workstation,
> and always logged as sysadmin ("administrateur" in french), cygwin seems to
> see/create the files on my hard disk as *not* belonging to me, but to
> another "being" (which is not obvious to me).
>
> I'll do my best to describe it (it's really *easy* to observe) : let's have
> a look at a simple Perl example, then a short C example with stat(), and my
> /etc/passwd and /etc/group. Finally I will create a simple file with
> 'touch', and you will notice that it will be reported as NOT belonging to
> me and not being writable :((
>
> Note : this seems specific to NT, which is not surprising regarding file
> permissions.
>
> Perl example :
> ----------------------
>
> Note : I created /etc/passwd and /etc/group with 'mkpasswd -l' and 'mkgroup
> -l' respectively, as reported in many FAQS.
>
> administrateur [26] /etc$ perl -e 'print "yes" if -w "group";'
>
> administrateur [27] /etc$ ll
> total 7
> -rw-r--r--   1 544 Aucun          27 Feb 26 02:45 group
> -rw-r--r--   1 544 Aucun         65 Mar  1 21:08 passwd
> -rw-r--r--   1 544 Aucun        9828 Dec  1 14:00 termcap
>
> That command (-w) will display 'yes' if the file (here, /etc/group) is
> writable. Apparently, it fails. And fails everywhere in my filesystem. The
> '-w' command is using C<stat>, let's move on :
>
> C example :
> ----------------------
> #include <sys/stat.h>
> #include <stdlib.h>
>
> int main() {
>     struct stat mystat;
>
>     stat("group", &mystat);
>
>     printf("mode : %o, uid : %u, gid : %u\n",
>            mystat.st_mode, mystat.st_uid, mystat.st_gid);
>
>     exit(0);
> }
>
> Run :
>
> administrateur [52] /etc$ gcc mystat.c
>
> administrateur [53] /etc$ a.exe
> mode : 100644, uid : 544, gid : 513
>
> administrateur [55] /etc$ ll group
> -rw-r--r--   1 544 Aucun          27 Feb 26 02:45 group
>
> Which means :
>         - /etc/group belongs to user which UID is 544, and to group which GID is 513
>         - it's a regular file (100000),
>         - read/write permission to owner (600), read permission to group (40) and
> read to other (4).
>
> BUT (and it drives me crazy). I CREATED that file, and I'm NOT user 544 !
> (more about this strange fellow below).
>
> administrateur [56] /etc$ whoami
> administrateur
>
> administrateur [74] /etc$ cat passwd
> Administrateur::500:513:seb::/bin/sh
> InvitÚ::501:513:::/bin/sh
>
> administrateur [75] /etc$ cat group
> Aucun::513:
> Everyone::0:
>
> Obviously, I'm user 500 (I changed my name to 'seb' so that to check).
> => Therefore WHY are all files created as 544 (even with tar) ? (the group
> is correct : 513).
>
> I thought it was related to /etc/passwd, but I just do NOT UNDERSTAND the
> difference between -l and -g option for 'mkpasswd'.
>
>    -l,--local              print local accounts
>    -g,--local-groups       print local group information too
>
> administrateur [77] /etc$ mkpasswd -l
> Administrateur::500:513:seb::/bin/sh
> InvitÚ::501:513:::/bin/sh
>
> administrateur [79] /etc$ mkpasswd -g
> Administrateurs::544:0:::
> Duplicateurs::552:0:::
> InvitÚs::546:0:::
> OpÚrateurs de sauvegarde::551:0:::
> Utilisateurs::545:0:::
> Utilisateurs avec pouvoir::547:0:::
>
> => who are these users ?!? these should be groups !! I'm belonging to the
> "Administrateurs" group for NT (in the "Gestionnaire d'utilisateurs" ~=
> "User manager"), and this has been translated to a user, I'm completely
> lost :((
>
> => dumping 'mkpasswd -l -g' to /etc/passwd (instead of 'mkpasswd -l') did
> NOT help.
>
> Here is the same joke :
>
> administrateur [84] /etc$ touch test
>
> administrateur [85] /etc$ ll test
> -rw-r--r--   1 544      Aucun           0 Mar  1 21:46 test
>
> administrateur [86] /etc$ perl -e 'print "yes" if -w "test";'
> administrateur [87] /etc$
>
> WOAH : I created a file, and it's automatically assigned to someone else !
> Give me my file back please :(
> And of course, although I created it, it's now NOT writable for cygwin.
>
> I guess I might be just dumb. Any help would be really appreciated (by my
> mental health).
>
> Thanks
>
> ______________________________________________________________
> Sebastien Barre                  http://www.hds.utc.fr/~barre/
>
> --
> Want to unsubscribe from this list?
> Send a message to cygwin-unsubscribe@sourceware.cygnus.com




--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

[HELP] stat(), file permission, r/w access : i'm LOST :(

[Sebastien Barre]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 4450 bytes --]

Dear cygwin people

As some of you might have noticed (sorry for the flood), I'm still trying
to arbitrate The Fight between Perl 5.002 and cygwin32-B20.1 (the
battlefield being NT4/SP4). Well, thanks to some of you, I nearly did it
(and also ruined a couple of nights), but a strange problem lead me to some
more confusion :

Problem : although I'm working as single user on a standalone workstation,
and always logged as sysadmin ("administrateur" in french), cygwin seems to
see/create the files on my hard disk as *not* belonging to me, but to
another "being" (which is not obvious to me). 

I'll do my best to describe it (it's really *easy* to observe) : let's have
a look at a simple Perl example, then a short C example with stat(), and my
/etc/passwd and /etc/group. Finally I will create a simple file with
'touch', and you will notice that it will be reported as NOT belonging to
me and not being writable :((

Note : this seems specific to NT, which is not surprising regarding file
permissions.


Perl example :
----------------------

Note : I created /etc/passwd and /etc/group with 'mkpasswd -l' and 'mkgroup
-l' respectively, as reported in many FAQS.

administrateur [26] /etc$ perl -e 'print "yes" if -w "group";'

administrateur [27] /etc$ ll
total 7
-rw-r--r--   1 544 Aucun          27 Feb 26 02:45 group
-rw-r--r--   1 544 Aucun         65 Mar  1 21:08 passwd
-rw-r--r--   1 544 Aucun        9828 Dec  1 14:00 termcap

That command (-w) will display 'yes' if the file (here, /etc/group) is
writable. Apparently, it fails. And fails everywhere in my filesystem. The
'-w' command is using C<stat>, let's move on :


C example :
----------------------
#include <sys/stat.h>
#include <stdlib.h>

int main() {
    struct stat mystat;

    stat("group", &mystat);

    printf("mode : %o, uid : %u, gid : %u\n", 
           mystat.st_mode, mystat.st_uid, mystat.st_gid);

    exit(0);
}

Run :

administrateur [52] /etc$ gcc mystat.c

administrateur [53] /etc$ a.exe
mode : 100644, uid : 544, gid : 513

administrateur [55] /etc$ ll group
-rw-r--r--   1 544 Aucun          27 Feb 26 02:45 group

Which means : 
	- /etc/group belongs to user which UID is 544, and to group which GID is 513
	- it's a regular file (100000), 
	- read/write permission to owner (600), read permission to group (40) and
read to other (4).

BUT (and it drives me crazy). I CREATED that file, and I'm NOT user 544 !
(more about this strange fellow below).

administrateur [56] /etc$ whoami
administrateur

administrateur [74] /etc$ cat passwd
Administrateur::500:513:seb::/bin/sh
InvitÚ::501:513:::/bin/sh

administrateur [75] /etc$ cat group
Aucun::513:
Everyone::0:

Obviously, I'm user 500 (I changed my name to 'seb' so that to check). 
=> Therefore WHY are all files created as 544 (even with tar) ? (the group
is correct : 513).

I thought it was related to /etc/passwd, but I just do NOT UNDERSTAND the
difference between -l and -g option for 'mkpasswd'.

   -l,--local              print local accounts
   -g,--local-groups       print local group information too

administrateur [77] /etc$ mkpasswd -l
Administrateur::500:513:seb::/bin/sh
InvitÚ::501:513:::/bin/sh

administrateur [79] /etc$ mkpasswd -g
Administrateurs::544:0:::
Duplicateurs::552:0:::
InvitÚs::546:0:::
OpÚrateurs de sauvegarde::551:0:::
Utilisateurs::545:0:::
Utilisateurs avec pouvoir::547:0:::

=> who are these users ?!? these should be groups !! I'm belonging to the
"Administrateurs" group for NT (in the "Gestionnaire d'utilisateurs" ~=
"User manager"), and this has been translated to a user, I'm completely
lost :((

=> dumping 'mkpasswd -l -g' to /etc/passwd (instead of 'mkpasswd -l') did
NOT help.

Here is the same joke :

administrateur [84] /etc$ touch test

administrateur [85] /etc$ ll test
-rw-r--r--   1 544      Aucun           0 Mar  1 21:46 test

administrateur [86] /etc$ perl -e 'print "yes" if -w "test";'
administrateur [87] /etc$

WOAH : I created a file, and it's automatically assigned to someone else !
Give me my file back please :( 
And of course, although I created it, it's now NOT writable for cygwin.


I guess I might be just dumb. Any help would be really appreciated (by my
mental health).

Thanks






______________________________________________________________
Sebastien Barre                  http://www.hds.utc.fr/~barre/

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: mySQL, mSQL, PostgreSQL with cygwin ?

[Stipe Tolj]

> Thanks for the information, but there is something very obscure to me, they
> ARE binaries for cygwin32 :
> http://www.tcx.se/download.html
>
> "MySQL 3.22.10 clients compiled with the cygwin32 kit. This includes
> 'mysql' with readline editing capabilities. " and a link to
> http://www.tcx.se/Downloads/Win32/mysqlclient-3.22.10b-cygwin-b20.tar.gz
>
> I downloaded it, here is the content :
>
> -rw-r--r-- tolj/None    723652 1998-11-08 17:59 cygwin1.dll
> -rwxr-xr-x tolj/None     41472 1998-11-17 19:40 insert_test.exe
> -rwxr-xr-x tolj/None    192512 1998-11-17 19:40 mysql.exe
> -rwxr-xr-x tolj/None     55296 1998-11-17 19:40 mysqladmin.exe
> -rwxr-xr-x tolj/None     62464 1998-11-17 19:40 mysqldump.exe
> -rwxr-xr-x tolj/None     53760 1998-11-17 19:40 mysqlimport.exe
> -rwxr-xr-x tolj/None     53760 1998-11-17 19:40 mysqlshow.exe
> -rwxr-xr-x tolj/None     41472 1998-11-17 19:40 select_test.exe

these cygwin b20 mysql client files have been provided by me to Monti from TcX.

The libmysqlclient.a lib and all client files have been ported to cygwin by us, but
the mysqld daemon uses pthreads which are currently not or at leasr experimentacly
supported within the latest cygwin snapshot.

So you will have to wait for a cygwin port of mysql until pthreads are running or
get a license for the native Win32 port provided from TcX.

Regards,
Stipe

--
Stipe Tolj <tolj@uni-duesseldorf.de>

Cygwin Porting Project -- "We build UNIX on top of Windows"
http://www.student.uni-koeln.de/cygwin/

Department of Economical Computer Science
University of Cologne, Germany





--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: mySQL, mSQL, PostgreSQL with cygwin ?

[Stipe Tolj]

> Thanks for the information, but there is something very obscure to me, they
> ARE binaries for cygwin32 :
> http://www.tcx.se/download.html
>
> "MySQL 3.22.10 clients compiled with the cygwin32 kit. This includes
> 'mysql' with readline editing capabilities. " and a link to
> http://www.tcx.se/Downloads/Win32/mysqlclient-3.22.10b-cygwin-b20.tar.gz
>
> I downloaded it, here is the content :
>
> -rw-r--r-- tolj/None    723652 1998-11-08 17:59 cygwin1.dll
> -rwxr-xr-x tolj/None     41472 1998-11-17 19:40 insert_test.exe
> -rwxr-xr-x tolj/None    192512 1998-11-17 19:40 mysql.exe
> -rwxr-xr-x tolj/None     55296 1998-11-17 19:40 mysqladmin.exe
> -rwxr-xr-x tolj/None     62464 1998-11-17 19:40 mysqldump.exe
> -rwxr-xr-x tolj/None     53760 1998-11-17 19:40 mysqlimport.exe
> -rwxr-xr-x tolj/None     53760 1998-11-17 19:40 mysqlshow.exe
> -rwxr-xr-x tolj/None     41472 1998-11-17 19:40 select_test.exe

these cygwin b20 mysql client files have been provided by me to Monti from TcX.

The libmysqlclient.a lib and all client files have been ported to cygwin by us, but
the mysqld daemon uses pthreads which are currently not or at leasr experimentacly
supported within the latest cygwin snapshot.

So you will have to wait for a cygwin port of mysql until pthreads are running or
get a license for the native Win32 port provided from TcX.

Regards,
Stipe

--
Stipe Tolj <tolj@uni-duesseldorf.de>

Cygwin Porting Project -- "We build UNIX on top of Windows"
http://www.student.uni-koeln.de/cygwin/

Department of Economical Computer Science
University of Cologne, Germany





--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: mySQL, mSQL, PostgreSQL with cygwin ?

[Sebastien Barre]

At 07:59 01/03/99 +0100, Stipe Tolj wrote:

>A contributed PostgreSQL 6.4 port in binary form is also available at this
>location, but it's only supposed to run on NT

Very good. I also downloaded the 6.4.2 source and will try to compile it.
Too bad there isn't a diff file at your location, but if I fail, I will
install the binary, for sure.

>MySQL is _not_ ported to Cygwin due to the lack of the pthreads functions used
>in the multi-user multi-threaded MySQL database engine. Currently Cygwin is
>getting at least experimental support for pthreads functions, but I suppose it
>will take some time to get a stable port of MySQL done to Cygwin.

Thanks for the information, but there is something very obscure to me, they
ARE binaries for cygwin32 :
http://www.tcx.se/download.html

"MySQL 3.22.10 clients compiled with the cygwin32 kit. This includes
'mysql' with readline editing capabilities. " and a link to
http://www.tcx.se/Downloads/Win32/mysqlclient-3.22.10b-cygwin-b20.tar.gz

I downloaded it, here is the content :

-rw-r--r-- tolj/None    723652 1998-11-08 17:59 cygwin1.dll
-rwxr-xr-x tolj/None     41472 1998-11-17 19:40 insert_test.exe
-rwxr-xr-x tolj/None    192512 1998-11-17 19:40 mysql.exe
-rwxr-xr-x tolj/None     55296 1998-11-17 19:40 mysqladmin.exe
-rwxr-xr-x tolj/None     62464 1998-11-17 19:40 mysqldump.exe
-rwxr-xr-x tolj/None     53760 1998-11-17 19:40 mysqlimport.exe
-rwxr-xr-x tolj/None     53760 1998-11-17 19:40 mysqlshow.exe
-rwxr-xr-x tolj/None     41472 1998-11-17 19:40 select_test.exe

Which seems to mean that the mySQL DLL was succesfully built (how ??), and
linked or added to the original cygwin1.dll ? If it's true, it may remove
my coolview support, am I wrong ? 

Furthermore, you have to get a license to use it officialy.

>There is a native Win32 shareware version of MySQL 3.21.29a which works
>perfectly with our Apache 1.3.4 and integrated PHP 3.0.6, check out MySQL's
>official web site.

I do not know if I'll be able to use it with Perl built for cygwin ? If I
compile a DBI module for MySQL within my cygwin perl context, the DLL
formats won't match I guess.

Regards


______________________________________________________________
Sebastien Barre                  http://www.hds.utc.fr/~barre/

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

Re: Cygwin participation threshold

[Christopher G. Faylor]

>From: Steve Morris <smorris@nexen.com>
>Date: Fri, 26 Feb 1999 11:55:50 -0500 (EST)
>
>In all this discussion something important is being lost. Cris
>bemoaned the lack of development support for cygwin and asked for
>reasons. I and others tried to explain where we think the issues
>are. Inevitably this comes out sounding negative, but at least on my
>part, this is not intended. Maybe we took Cris' questions too
>literally. Flogging Cygnus was not the intent. We were trying to offer
>legitimate feedback to a legitimate question.
>
>Let me reiterate that Cygnus is clearly one of the Good Guys. The best
>guys are the Cygnus employees (like Cris) who volunteer their own time
>to this project.
>
>Many of us are rooting for Cygnus and are hoping more companies figure
>out how to make money on free software; because they then tend to give
>back. As an example gcc and gdb have been in much better shape all
>these years since Cygnus became the official release site.

Thanks very much for this positive note.

I was hoping that my original message would generate a lot of discussion
because I thought that it would give DJ and myself an opportunity to
explain something about Cygnus/Cygwin as well as providing a venue
for everyone else to provide their outlook.

It may not be obvious but I do appreciate any and all feedback on this
issue.  It certainly helps educate me on people's perception of my
company and my product.  I do want to understand what people think
about this.

That doesn't mean that I won't argue with perceptions that I think
are wrong but I certainly respect everyone's right to an opinion.

Again, thanks to you and to the handful of other people who sent me
private supportive email.  I appreciate it.

cgf

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

RE: mySQL, mSQL, PostgreSQL with cygwin ?

[Horak Daniel]

> >A contributed PostgreSQL 6.4 port in binary form is also 
> available at this
> >location, but it's only supposed to run on NT
> 
> Very good. I also downloaded the 6.4.2 source and will try to 
> compile it.
> Too bad there isn't a diff file at your location, but if I 
> fail, I will
> install the binary, for sure.

Patches for 6.4.2 are at http://www.askesis.nl/AskesisPostgresIndex.html ,
they are included in the development snapshots
( ftp://ftp.postgresql.org/pub/postgresql.snapshot.tar.gz )

			Dan

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com

RE: mySQL, mSQL, PostgreSQL with cygwin ?

[Horak Daniel]

> >A contributed PostgreSQL 6.4 port in binary form is also 
> available at this
> >location, but it's only supposed to run on NT
> 
> Very good. I also downloaded the 6.4.2 source and will try to 
> compile it.
> Too bad there isn't a diff file at your location, but if I 
> fail, I will
> install the binary, for sure.

Patches for 6.4.2 are at http://www.askesis.nl/AskesisPostgresIndex.html ,
they are included in the development snapshots
( ftp://ftp.postgresql.org/pub/postgresql.snapshot.tar.gz )

			Dan

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com