Re: .logout files?

[Jani tiainen]

Larry Hall wrote:

At 02:43 PM 4/17/2004, you wrote:

On Sat, Apr 17, 2004 at 11:03:00AM -0700, Karl M wrote:

On Sat, Apr 17, 2004 at 09:17:16AM +0200, Corinna Vinschen wrote:

On Apr 16 21:35, Christopher Spears wrote:

Hi! I can't seem to find a .logout file.  I found
The reason might be that there is none.
Sometimes the simple answers are just so paradoxically elusive.
And it can be so satisfying to knoy you have completely answered a
question without actually conveying any information.
Bingo.  It's one of the many perks in being a cygwin developer.
That's just plain mean!  I think all cygwin developers should be 
banished!! ...Wait a minute, that won't work... ;-)
Err... Maybe short time electricshock treatment would help a little.. =)

Well, what's this dot-logout file is used for anyway..?

--

Jani

--
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: gentoo portage and cygwin

[Thorsten Kampe]

* Sven Köhler (2004-04-19 07:58 +0100)
> has anybody ever tried to port the gentoo-portage to cygwin? not that i 
> expect many of the gentoo-ebuilds to compile, but the cygwin-people 
> could maintain their own portage. the gentoo-portage also supports 
> binary packages so that you don't have to compile everything yourself.

* portage-cygwin
  
http://sourceforge.net/project/showfiles.php?group_id=27249&package_id=19019&release_id=199935

* Gentoo/Cygwin
  https://sourceforge.net/project/showfiles.php?group_id=87984&package_id=91835

The first one is available through setup.exe the latter one has its
own Gentoo mailing list (http://www.gentoo.org/main/en/lists.xml)

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: gentoo portage and cygwin

[Sergey Okhapkin]

Sven Köhler wrote:

has anybody ever tried to port the gentoo-portage to cygwin? not that i 
expect many of the gentoo-ebuilds to compile, but the cygwin-people 
Take a look at 
http://sourceforge.net/project/showfiles.php?group_id=27249&package_id=19019

--
Sergey Okhapkin
Somerset, NJ
--
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: seg fault via cygwin1!aclcheck() ?

[Dave Korn]

> -Original Message-
> From: cygwin-owner On Behalf Of Larry Hall
> Sent: 18 April 2004 23:56

> >I am running ns on cygwin, and after my program executes for 
> a while with no
> >problems, suddenly I get a seg fault with the following 
> backtrace (in gdb):
> >
> >Program received signal SIGSEGV, Segmetation fault.
> >0x610ab27b in random () from /usr/bin/cygwin1.dll
> >(gdb) backtrace
> >#0 0x610ab27b in random () from /usr/bin/cygwin1.dll
> >#1 0x610550c6 in localtime_r () from /usr/bin/cygwin1.dll
> >#2 0x61086751 in cygwin1!aclcheck () from /usr/bin/cygwin1.dll
> >...(the rest of the trace is from my ns code)

  That would have perhaps been the more informative bit.  If there's a crash
in the .dll, it's most usually because your code has made a system call with
bad arguments, such as an invalid pointer.

> >What is wrong? It seems like a permissions problem, but my 
> code is not writing
> >to a file. Although my code reads from a file, this seg 
> fault occurs after a
> >great many lines have been already read from this file. 
> Based on the back
> >trace, the seg fault does not occur at a point where the 
> file is read anyway. 
> >
> >How can I fix this?
> 
> 
> Looks like you're going to have to break out the debugger.  
> aclcheck() 
> doesn't call localtime_r() (or anything close to it) 

  Not only does aclcheck not call localtime_r, but I'd be *really*
interested to know what reason localtime_r could *ever* have for calling
random 

> so this leads me to 
> conclude that you have a stack corruption in your program somewhere.

  It's not necessarily even that at all; it's just that a non-debug build of
the .dll only contains symbols for the exported entry points.  All the
internal functions that have no external linkage (static functions in C
terms; file-scope in C++) aren't declared.  So there are only intermittent
symbols to mark points throughout the code in the .dll.  The function names
you see in the backtrace are misleading: gdb looks backward from the
recorded value of eip to find the nearest symbol that's declared, but it's
not smart enough to know that if it's stepped backwards over a return
instruction it must have backed up into a completely different subroutine.

  Without a full debug build of the .dll, stacktracebacks are essentially
worthless.  Sorry, that's imprecise, let me correct myself:  Without a full
debug build of the .dll, the part of the traceback showing stack frames for
functions inside the .dll is worthless, and without a full debug build of
the application, the subsequent part of the traceback showing the stack
frame chain through the functions in the application is worthless.  (Or very
very close to worthless).


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/

Unix time conversion error

[jurgen . defurne]

Hello,

I have a conversion problem. When I convert UNIX timestamps
to datetime representations, they are one hour off. Does this have to
with the following mail (indeed we are now in DST) ?

http://www.cygwin.com/ml/cygwin/2003-10/msg00995.html

I first thought it was a Perl problem, but I now notice it in mySQL
too (although the mySQL server is running on Solaris).

Regards,

Jurgen

--
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: gentoo portage and cygwin

[Ralf Habacker]

On Monday 19 April 2004 12:06, Thorsten Kampe wrote:
> * Sven Köhler (2004-04-19 07:58 +0100)
>
> > has anybody ever tried to port the gentoo-portage to cygwin? not that i
> > expect many of the gentoo-ebuilds to compile, but the cygwin-people
> > could maintain their own portage. the gentoo-portage also supports
> > binary packages so that you don't have to compile everything yourself.
>
> * portage-cygwin
>  
> http://sourceforge.net/project/showfiles.php?group_id=27249&package_id=1901
>9&release_id=199935

please note that this portage release depends on python-2.2. The initial 
porter hasn't updated it to python2.3 and we cannot give support for this 
package. 


Ralf 

--
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/

Typo in Cygwin website main menu

[Zas]

Hi,

In main menu present on most pages:

http://x.cygwin.com/";>Cygwin/X 
Home
   here

Regards,

-- 

Zas

--
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: Typo in Cygwin website main menu

[Corinna Vinschen]

On Apr 19 14:42, Zas wrote:
> Hi,
> 
>   In main menu present on most pages:
> 
> http://x.cygwin.com/";>Cygwin/X 
> Home
>    here

Thanks, fixed.

Corinna

-- 
Corinna Vinschen  Please, send mails regarding Cygwin to
Cygwin Co-Project Leader  mailto:[EMAIL PROTECTED]
Red Hat, Inc.

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

Re: Cygwin setup window size, why no maximize button enabled ?

[Andrew Schulman]

> > > http://cygwin.com/setup-snapshots/
> > 
> > Hallelujah, a resizable setup program.  What a relief.
> > 
> > Now if only it would recognize my mouse wheel...
> 
> It depends on windows to send the appropriate messages. What OS and what
> mouse are you using?

Win2K SP4, with a Logitech generic mouse.  Other apps recognize my mouse 
wheel, so I don't think it's a Windows message problem.


--
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/

Wishlist additions for setup.exe

[Shish]

o) An "approximate time remaining" field in the download progress window would
be nice - you're got the total size, size done, and download speed; Adding this
field should be a single simple equation.

o) Cache mirrors.lst and setup.bz2 - if they're already there, ask me if I want
to update. Even using the "back" and "next" buttons causes it to be
re-downloaded

o) I was going to complain about the lack of resiazability, but then a mail
came in with the solution ^_^. Still, the other parts than the package
selection look a bit silly stuck in the top left, but at least the package
select is usable now :)

Also, I'm another person for whom the the package list isn't mouse-wheel
compatible (setup 2.416, windows 98, micorsoft mouse, other apps work fine)

 -- Shish




__
Do you Yahoo!?
Yahoo! Photos: High-quality 4x6 digital prints for 25¢
http://photos.yahoo.com/ph/print_splash

--
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: Wishlist additions for setup.exe

[Dave Korn]

> -Original Message-
> From: cygwin-owner  On Behalf Of Shish
> Sent: 19 April 2004 14:48

> o) Cache mirrors.lst and setup.bz2 - if they're already 
> there, ask me if I want
> to update. Even using the "back" and "next" buttons causes it to be
> re-downloaded

  Deep disagreement, about setup.bz2 at any rate.  If you ever attempt to do
a setup with an out-of-date setup.bz2 file that doesn't reflect the true
reality of what's actually there at the mirror sites, the consequences could
be pretty damn awful.  It sounds to me like an idea that would save people
ten or twenty seconds every time they updated at the cost of losing an
entire afternoon every so often when it goes wrong and they end up having to
reinstall a trashed cygwin installation.  I think that's probably a fairly
marginal trade-off.  I can also see it generating absolutely *tons* of
support requests with *really* obscure and esoteric breakages being sent to
the list.

  Oh, and BTW, the traditional cygwin list response to a wishlist is just to
reply "PTC!" :)  BWAM!

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: [OT?] make bash script wait for called program to finish?

[Buchbinder, Barry (NIH/NIAID)]

Actually, there may be a simpler way than cmd /c start /wait
program_being_waited_for

/c/tmp> cat t
#!/bin/sh
date
/c/windows/notepad
date
/c/tmp> t
Mon Apr 19 09:51:35 EDT 2004
Mon Apr 19 09:51:47 EDT 2004
1 /c/tmp>

Ditto with #!/bin/bash.

So (at least on XP) if you just start the program without using cygstart or
putting it in the background with &, the shell just waits for it to finish.

-Original Message-
From: Thomas L Roche [mailto:[EMAIL PROTECTED] 
Sent: Sunday, April 18, 2004 8:01 PM
To: [EMAIL PROTECTED]
Subject: Re: [OT?] make bash script wait for called program to finish?


Tom Roche Sunday, April 18, 2004 6:20 PM
>>> what I'd like my script to do is

>>> 0 wget

>>> 1 launch the extractor UI, which I will briefly twiddle

>>> 2 wait for the extractor to finish

>>> 3 check that the installer and images were created

>>> 4 run the installer on the images

>>> I know how to script everything but 2

Barry Buchbinder Sun, 18 Apr 2004 18:56:47 -0400
>> How about the following?
>> $ cmd /c start /wait program_being_waited_for

Thanks! I've never run windows' shell from bash, but that does the
job.

Larry Hall Sun, 18 Apr 2004 19:12:03 -0400
> Should we assume that 'wait' didn't work for you?

I didn't try that: doesn't 'wait' just wait for the given period
(arg=n)? If so, that would be at best a kludge.


--
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: libc

[Richard Campbell]

>Libc has as a standard catgets. But catgets and catopen are absent from the 
>cygwin implementation and gencat and nl_types are not to be found.
>
>So where can I find either the source for these or a technical spec?

Switched to the main cygwin list.

Cygwin uses newlib.

http://sources.redhat.com/newlib/

-Richard Campbell.

--
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: libc

[Dave Korn]

> -Original Message-
> From: cygwin-owner On Behalf Of Richard Campbell
> Sent: 19 April 2004 15:12

> >Libc has as a standard catgets. But catgets and catopen are 
> absent from the 
> >cygwin implementation and gencat and nl_types are not to be found.
> >
> >So where can I find either the source for these or a technical spec?
> 
> Switched to the main cygwin list.
> 
> Cygwin uses newlib.
> 
> http://sources.redhat.com/newlib/


  Also, cygwin implements a POSIX layer in compliance with the single unix
standard SUSv3.  It doesn't attempt to implement all the potential
extensions.  If you look up catgets in SUSv3, you'll see it's tagged with a
little "XSI" tag, which means:

"  The functionality described is an XSI extension. Functionality marked XSI
is also an extension to the ISO C standard. Application writers may
confidently make use of an extension on all systems supporting the X/Open
System Interfaces Extension.  "

  Since cygwin uses, as you say, newlib, you'll find that you can really
only rely on the core POSIX functions being implemented.  To find out what
functions are included in cygwin, see
http://cygwin.com/cygwin-api/cygwin-api.html .  To find out the technical
specs for them, as well as any unimplemented extensions, see the SUSv3 at
http://www.opengroup.org/onlinepubs/007904975/ .  For the sources to the
missing functions, well.. think about it  

[hint: glibc is an open source software package]


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/

How do I link libnsl.a in Cygwin? It is not available in the Cygwin installed.

[Cuckoo]

Hi All,

I am trying to port my Linux 8.0 code to Cygwin.
However my source code uses libnsl.a which is not to
be found in the package...!
Where can I download this library ?
Is there any other package that I can use  ??
Plzzz help !!


Thanks in advance,
Megha.






Chat instantly with your online friends?  Get the
FREE Yahoo! Messenger http://uk.messenger.yahoo.com/

--
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: Wishlist additions for setup.exe

[Dave Korn]

> -Original Message-
> From: Shish 
> Sent: 19 April 2004 15:21
> To: Dave Korn

[Conversation restored to the list address]

> > > o) Cache mirrors.lst and setup.bz2
> > Deep disagreement, about setup.bz2 at any rate.  If you 
> ever attempt to do
> > a setup with an out-of-date setup.bz2 file that doesn't 
> reflect the true
> > reality of what's actually there at the mirror sites, the 
> consequences could
> > be pretty damn awful.
> 
> Point taken - but can it really get dangerously outdated in 
> the second or so it
> takes me to click "back", change an option, and click "next" again?

  Well, when ever an update to a new version does happen, it only takes an
instant.  Any time that instant is in between downloading the setup file and
using its contents as a guide to what's there, a problem is possible.

  Come to think of it though, I don't know if the cygwin release process
actually tries to ensure that the entire release archive is updated
atomically, so I guess maybe it's just something that already goes wrong
once in a while.  If there's no guarantee that one of the cygwin mirror
sites won't sync itself in between someone uploading a new setup.bz2 to the
main mirror and uploading the packages to which it refers, there's no point
in worrying about a bit of further caching that might occur at the client
end.  OTOH if the cygwin release process involves putting the new packages
up and then atomically updateing setup.bz2 to reflect the new state of the
server, there could be some breakage caused by caching it locally.

> >   Oh, and BTW, the traditional cygwin list response to a 
> wishlist is just to
> > reply "PTC!" :)  BWAM!
> 
> I'm new here, what's that about? :)

OLOCA! [*]


cheers, 
  DaveK

[*] http://cygwin.com/acronyms/, in particular
http://cygwin.com/acronyms/#PTC, http://cygwin.com/acronyms/#BWAM, and of
course http://cygwin.com/acronyms/#OLOCA
-- 
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/

How do I link libnsl.a in Cygwin? It is not available in the Cygwin installed.

[Cuckoo]

 
Following is the error message i recieve :-


g++  -g -O2 -o run run.o libsim.a  ../../bfd/libbfd.a
../../opcodes/libopcodes.a 
../../libiberty/libiberty.a -lnsl  -lm 
/usr/lib/gcc-lib/i686-pc-cygwin/3.3.1/../../../../i686-pc-cygwin/bin/ld:
cannot
find -lnsl
collect2: ld returned 1 exit status
make[2]: *** [run] Error 1

Note: forwarded message attached. 






Yahoo! Messenger - Communicate instantly..."Ping" 
your friends today! Download Messenger Now 
http://uk.messenger.yahoo.com/download/index.html--- Begin Message ---
Hi All,

I am trying to port my Linux 8.0 code to Cygwin.
However my source code uses libnsl.a which is not to
be found in the package...!
Where can I download this library ?
Is there any other package that I can use  ??
Plzzz help !!


Thanks in advance,
Megha.






Chat instantly with your online friends?  Get the
FREE Yahoo! Messenger http://uk.messenger.yahoo.com/
--- End Message ---
--
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: [OT?] make bash script wait for called program to finish?

[Thomas L Roche]

Barry Buchbinder Sun, 18 Apr 2004 18:56:47 -0400
 How about the following?
 $ cmd /c start /wait program_being_waited_for

Tom Roche Sunday, April 18, 2004 8:01 PM
>>> I've never run windows' shell from bash, but that does the job.

Larry Hall 04/18/2004 07:12 PM
>> [For 'wait n',] 'n' is a proc id.  See 'man bash'.

arrgghh: I tried "{info, man} wait".

Buchbinder, Barry 04/19/2004 10:03:00 AM:
> (at least on XP) if you just start the program without using
> cygstart or putting it in the background with &, the shell just
> waits for it to finish.

I'm not sure if the problem is that

* I'm on 2k

* the program being run is java-based

but bash does not stop.


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

RE: How do I link libnsl.a in Cygwin? It is not available in the Cygwin installed.

[Dave Korn]

> -Original Message-
> From: cygwin-owner On Behalf Of Cuckoo
> Sent: 19 April 2004 15:26

> Hi All,
> 
> I am trying to port my Linux 8.0 code to Cygwin.
> However my source code uses libnsl.a which is not to
> be found in the package...!
> Where can I download this library ?

  From http://cygwin.com/cgi-bin2/package-grep.cgi?grep=libnsl :

"Cygwin Package List
Search Results
Found 0 matches for libnsl"

  Which suggests that there's no official cygwin port of it at any rate.

> Is there any other package that I can use  ??

  Well, there are plenty of library packages out there that help with
network programming, but only you actually know what your program is or what
it does and therefore which functions it needs and whether or not any other
library might help.

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


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

RE: [OT?] make bash script wait for called program to finish?

[Larry Hall]

At 10:36 AM 4/19/2004, you wrote:
>Barry Buchbinder Sun, 18 Apr 2004 18:56:47 -0400
> How about the following?
> $ cmd /c start /wait program_being_waited_for
>
>Tom Roche Sunday, April 18, 2004 8:01 PM
 I've never run windows' shell from bash, but that does the job.
>
>Larry Hall 04/18/2004 07:12 PM
>>> [For 'wait n',] 'n' is a proc id.  See 'man bash'.
>
>arrgghh: I tried "{info, man} wait".
>
>Buchbinder, Barry 04/19/2004 10:03:00 AM:
>> (at least on XP) if you just start the program without using
>> cygstart or putting it in the background with &, the shell just
>> waits for it to finish.
>
>I'm not sure if the problem is that
>
>* I'm on 2k
>
>* the program being run is java-based
>
>but bash does not stop.


Clearly, whatever your program is, it effectively runs a separate process,
otherwise you wouldn't see the problem you do (as Barry points out).  It's
not strictly related to the O/S or the programming language. 

I think we've covered this rather off-topic subject well enough.  I'd like
to suggest that we end the thread here unless something Cygwin-specific 
surfaces.



--
Larry Hall  http://www.rfk.com
RFK Partners, Inc.  (508) 893-9779 - RFK Office
838 Washington Street   (508) 893-9889 - FAX
Holliston, MA 01746 


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

CVS + SSH + Binary File = ^M

[kgrizzle]

I have a cygwin CVS server that I access remotely with the cygwin CVS client and Open 
SSH.  When I add binary files to the repository from a remote machine, they are 
corrupted by having ^M line endings appended to each line.  When I add a binary file 
to the repository from the server machine (ie - CVSROOT=/cvsroot), the file is not 
corrupted.

Both server and client machines are running cygwin in textmode rather that binmode.  
However, I mounted a test CVS repository on the server in binmode, and the files were 
still corrupted when added from the remote machine.  Here are the details of my setup:

Client:
cygwin (1.5.9-1)
cvs (1.11.5-1)
openssh (3.8p1-1)
all filesystems are mounted in textmode

Server:
cygwin (1.3.22-1)
cvs (1.11.5-1)
openssh (3.6.1p1-2)
all filesystems are mounted in textmode


Any insight on troubleshooting this problem would be much appreciated.

Thanks,
Kelly


--
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: Unix time conversion error

[Larry Hall]

At 07:18 AM 4/19/2004, you wrote:
>Hello,
>
>I have a conversion problem. When I convert UNIX timestamps
>to datetime representations, they are one hour off. Does this have to
>with the following mail (indeed we are now in DST) ?
>
>http://www.cygwin.com/ml/cygwin/2003-10/msg00995.html
>
>I first thought it was a Perl problem, but I now notice it in mySQL
>too (although the mySQL server is running on Solaris).


Maybe but I doubt that someone will be able to provide much input unless
they see exactly the same problem you do.  You've really provided no other
information on which someone else here might be able to comment.  So I'd
recommend reading:

>Problem reports:   http://cygwin.com/problems.html


Of course, if you were enquiring on the off-chance that someone might 
have already seen this problem and have a solution for you, then you 
don't need to bother doing anything else.  If someone has seen this and
has some relevant input, I'm sure they'll speak up.



--
Larry Hall  http://www.rfk.com
RFK Partners, Inc.  (508) 893-9779 - RFK Office
838 Washington Street   (508) 893-9889 - FAX
Holliston, MA 01746 


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

Where to get cygwin's gmake sources, and does it build without errors?

[Sandy Currier]

Hi all,

[hope this is an ok place to ask this question...]

I need to build the latest 3.81rc1 (beta1) version of gmake under cygwin.
However, the native gmake CVS distribution fails 6 make tests when I
build it under cygwin.  The maintainer of gmake claims that FSF does not
support make under cygwin, basically only because he has not seen the
modifications that the 'cygwin' community has made to gmake.
I have looked for the cygwin gmake sources for the cygwin distribution,
but cannot find them on the web.  (This is nominally for 3.80 which appears
to be the current version).
As mentioned, though 3.81rc1 does build under cygwin, I at least get 6 test
failures when I build it.  A couple questions:
1) Does 3.80 gmake build and test without errors?
2) Where can I get the 3.80 sources?  Have they been modified from the FSF
  native sources?
3) Should the CVS tip for the 3.81 release build and test without errors?
If the cygwin community has modified the native FSF, I would like to
volunteer to try to resolve any such modifications, incorporate them back
into the FSF sources, and build 3.81 without error.  If the cygwin community
has not modified the gmake sources, it would be great to get some pointers
as to how to successfully build gmake.
Thanks much in advance!
-sandy
--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

RE: CVS + SSH + Binary File = ^M

[Dave Korn]

> -Original Message-
> From: cygwin-owner On Behalf Of kgrizzle
> Sent: 19 April 2004 16:15

> I have a cygwin CVS server that I access remotely with the 
> cygwin CVS client and Open SSH.  When I add binary files to 
> the repository from a remote machine, they are corrupted by 
> having ^M line endings appended to each line.  When I add a 
> binary file to the repository from the server machine (ie - 
> CVSROOT=/cvsroot), the file is not corrupted.
> 
> Both server and client machines are running cygwin in 
> textmode rather that binmode.  However, I mounted a test CVS 
> repository on the server in binmode, and the files were still 
> corrupted when added from the remote machine

  At this point you should have tried mounting a directory on the *client*
in binmode, and adding a file from *that* directory.  As it is, you seem to
have shown (both by adding files locally on the server and having no
problem, and by adding them remotely but to a known binmode fs and seeing
the problem) that you should be looking for something going wrong at the
client end.  It might also be a useful experiment to try adding files from
the remote machine to the repository using the :pserver: protocol, to try
and establish if it's the SSh on the client machine that is underlying this
problem.  If the problem still occurs using :pserver: suspect a bad
interaction with the textmode filesystem.

  Basically I think it's a bad idea to mount your filesystems in textmode
and then complain "But it keeps adding \r to my line endings".  Accessing a
binary file through a textmode mountpoint is a dubious thing to expect to
work.  I personally think the most reliable strategy is to accept that we're
working with a POSIX system, standardize on Unix EOLs throughout your files,
mount your filesystems in binary mode, and only ever have Unix<=>DOS EOL
conversions by deliberately invoking d2u/u2d manually.


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: Where to get cygwin's gmake sources, and does it build without errors?

[Dave Korn]

> -Original Message-
> From: cygwin-owner On Behalf Of Sandy Currier
> Sent: 19 April 2004 16:22


> However, the native gmake CVS distribution fails 6 make tests when I
> build it under cygwin.  The maintainer of gmake claims that 
> FSF does not
> support make under cygwin, basically only because he has not seen the
> modifications that the 'cygwin' community has made to gmake.
> 
> I have looked for the cygwin gmake sources for the cygwin 
> distribution,
> but cannot find them on the web.

  You don't.  You run the cygwin setup application instead, and tick the
little box in the 'Src' column next to make.  That downloads the sources for
whatever version corresponds to the version of make you have installed,
unpacking them under /usr/src/

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/

problem installing libiconv-1.9.2

[Steve Kelem]

What's the trick to getting libiconv to install?  If I try to do "make 
install" from tcsh, I get the message:
/bin/install -c -m 644 .libs/cygiconv-2.dll $dldir/cygiconv-2.dll
/bin/install: cannot remove `/usr/lib/../bin/cygiconv-2.dll': Permission 
denied

I assume that this is because I'm using cygwin/tcsh.

The main reason I'm trying to recompile iconv is because after building 
gd-2.0.22, I cannot get the perl module GD-2.12 to link, so I'm trying 
to build a more recent copy (1.9.2) of libiconv:

gcc -shared -o  GD.dll -Wl,--out-implib=libGD.dll.a 
-Wl,--export-all-symbols -Wl,--enable-auto-import -Wl,--stack,8388608 \
-s -L/usr/local/lib GD.o  
/usr/lib/perl5/5.8.2/cygwin-thread-multi-64int/CORE/libperl.dll.a 
-L/usr/X11R6/lib -L/usr/X11/lib -L/usr/local/lib -lgd -lpng -lz 
-lfreetype -ljpeg -lm -lX11 -lXpm
Creating library file: libGD.dll.a
/bin/../lib/gcc-lib/i686-pc-cygwin/3.3.1/../../../libgd.a(gdkanji.o)(.text+0x707): 
In function `do_convert':
/usr/local/src/gd-2.0.22/gdkanji.c:380: undefined reference to 
`_libiconv_open'
/bin/../lib/gcc-lib/i686-pc-cygwin/3.3.1/../../../libgd.a(gdkanji.o)(.text+0x74c):/usr/local/src/gd-2.0.22/gdkanji.c:394: 
undefined reference to `_libiconv'
/bin/../lib/gcc-lib/i686-pc-cygwin/3.3.1/../../../libgd.a(gdkanji.o)(.text+0x757):/usr/local/src/gd-2.0.22/gdkanji.c:415: 
undefined reference to `_libiconv_close'
collect2: ld returned 1 exit status
perlld: *** system() failed to execute

Thanks for your help,
Steve Kelem
Charles Wilson wrote:

The GNU libiconv package provides an iconv() implementation, for use 
on systems which don't have one, or whose implementation cannot 
convert from/to Unicode.

These should be upgraded together with the gettext packages.  This 
release is built against the cygwin-1.5.x kernel; it will not work 
with 1.3.22.

libiconv-1.9.1-3
libiconv2-1.9.1-3
libcharset1-1.9.1-3
gettext-0.12.1-3
gettext-devel-0.12.1-3
libintl2-0.12.1-3
libgettextpo0-0.12.1-3 (*)
(*) new library package -- gettext-devel now depends on it.

CHANGES:

* Both libiconv and gettext packages were COMPLETELY reorganized
  between (libiconv: 1.8 -> 1.9.1, gettext: 0.11.5 -> 0.12.1).
* compiled against cygwin-1.5.2 kernel
* documentation moved to /usr/share/*


--
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: [OT?] make bash script wait for called program to finish?

[Buchbinder, Barry (NIH/NIAID)]

I was actually surprised that bash stopped when I did the experiment.

It wouldn't surprise me if it would depend on which program one was trying
to wait for, in addition to which version of windows.  MS products do so
much behind one's back -- making life easier for the novice but more
difficult if one half-knows what one is doing -- that it is difficult to
rely on consistency.

If you are working in a mixed cmd.exe/command.com environment, you might
want to try one of the following

/c/tmp> $(cygpath -u "${COMSPEC}") /c start /wait /program_to_wait_for

or

/c/tmp> eval $(cygpath -u "${COMSPEC}") /c start /wait /program_to_wait_for

(I've yet to figure out when eval is appropriate, but both seem to work
here, at least for me.)

Or use command.com instead of cmd.exe -- I think all system have it.  (At
least XP Pro does.)

-Original Message-
From: Thomas L Roche [mailto:[EMAIL PROTECTED] 
Sent: Monday, April 19, 2004 10:36 AM
To: [EMAIL PROTECTED]
Subject: RE: [OT?] make bash script wait for called program to finish?


Barry Buchbinder Sun, 18 Apr 2004 18:56:47 -0400
 How about the following?
 $ cmd /c start /wait program_being_waited_for

Tom Roche Sunday, April 18, 2004 8:01 PM
>>> I've never run windows' shell from bash, but that does the job.

Larry Hall 04/18/2004 07:12 PM
>> [For 'wait n',] 'n' is a proc id.  See 'man bash'.

arrgghh: I tried "{info, man} wait".

Buchbinder, Barry 04/19/2004 10:03:00 AM:
> (at least on XP) if you just start the program without using
> cygstart or putting it in the background with &, the shell just
> waits for it to finish.

I'm not sure if the problem is that

* I'm on 2k

* the program being run is java-based

but bash does not stop.


--
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: problem installing libiconv-1.9.2

[Dave Korn]

> -Original Message-
> From: cygwin-owner On Behalf Of Steve Kelem
> Sent: 19 April 2004 16:42

> What's the trick to getting libiconv to install?  If I try to 
> do "make 
> install" from tcsh, I get the message:
>  /bin/install -c -m 644 .libs/cygiconv-2.dll $dldir/cygiconv-2.dll
> /bin/install: cannot remove `/usr/lib/../bin/cygiconv-2.dll': 
> Permission 
> denied

  That would suggest that the dll cannot be deleted because it is still
currently in use by some application.  But it could just be that the perms
on it are wrong.  "make install" often requires admin (root) rights to run
completely, depending on the particular package you're building and the .
 
> I assume that this is because I'm using cygwin/tcsh.

  Well, does tcsh use cygiconv-2.dll?  That would really explain the problem
if it did.  But it turns out not to, at least not directly:

[EMAIL PROTECTED] /usr/bin> cygcheck tcsh.exe
Found: .\tcsh.exe
Found: C:\cygwin\bin\tcsh.exe
tcsh.exe
  .\cygncurses7.dll
.\cygwin1.dll
  C:\WINDOWS\System32\ADVAPI32.DLL
C:\WINDOWS\System32\ntdll.dll
C:\WINDOWS\System32\KERNEL32.dll
C:\WINDOWS\System32\RPCRT4.dll
[EMAIL PROTECTED] /usr/bin>

  So 1) either the permissions on it are wrong for the username you're
running as at "make install" time, or 2) some *other* cygwin application
(got any cygrunsrv services installed by any chance?) is still keeping it
open.  If you get really desparate, use something like Process explorer or
Handle from sysinternals.com to find out what process is keeping a handle to
the dll open.

> The main reason I'm trying to recompile iconv is because 
> after building 
> gd-2.0.22, I cannot get the perl module GD-2.12 to link, so 
> I'm trying 
> to build a more recent copy (1.9.2) of libiconv:
> 
> gcc -shared -o  GD.dll -Wl,--out-implib=libGD.dll.a 
> -Wl,--export-all-symbols -Wl,--enable-auto-import 
> -Wl,--stack,8388608 \
> -s -L/usr/local/lib GD.o  
> /usr/lib/perl5/5.8.2/cygwin-thread-multi-64int/CORE/libperl.dll.a 
> -L/usr/X11R6/lib -L/usr/X11/lib -L/usr/local/lib -lgd -lpng -lz 
> -lfreetype -ljpeg -lm -lX11 -lXpm

  I don't think a more recent version of libiconv is going to help your
basic problem here, which is that you haven't actually told gcc to link
libiconv into the application!  I'd expect to see a "-lcygiconv-2" option
somewhere in that command if you expect it to actually work..!

> Creating library file: libGD.dll.a
> /bin/../lib/gcc-lib/i686-pc-cygwin/3.3.1/../../../libgd.a(gdka
> nji.o)(.text+0x707): 
> In function `do_convert':
> /usr/local/src/gd-2.0.22/gdkanji.c:380: undefined reference to 
> `_libiconv_open'
> /bin/../lib/gcc-lib/i686-pc-cygwin/3.3.1/../../../libgd.a(gdka
> nji.o)(.text+0x74c):/usr/local/src/gd-2.0.22/gdkanji.c:394: 
> undefined reference to `_libiconv'
> /bin/../lib/gcc-lib/i686-pc-cygwin/3.3.1/../../../libgd.a(gdka
> nji.o)(.text+0x757):/usr/local/src/gd-2.0.22/gdkanji.c:415: 
> undefined reference to `_libiconv_close'
> collect2: ld returned 1 exit status
> perlld: *** system() failed to execute

  When you're missing fairly fundamental entry points, such as libiconv_open
or libiconv_close in this case, you should probably be able to deduce that
'open' and 'close' aren't advanced features that only recently got added to
libiconv, which in turn should have given you the clue that maybe you
weren't actually linking against it at all..


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: [OT?] make bash script wait for called program to finish?

[Dave Korn]

> -Original Message-
> From: cygwin-owner On Behalf Of Buchbinder, Barry (NIH/NIAID)

> I was actually surprised that bash stopped when I did the experiment.
> 
> It wouldn't surprise me if it would depend on which program 
> one was trying
> to wait for, in addition to which version of windows. 

  Yep, it certainly seems that some windows gui apps have the behaviour of
detaching from the console, while others remain attached and so don't return
control to the shell until they're finished.  I noticed this in particular
when editing text files at an interactive bash prompt: if I use notepad.exe,
my shell is locked up until I exit, whereas if I use write.exe (wordpad), it
detaches and the shell prompt returns immediately.

  I also note as a data point that when run from a cmd.exe shell, both
notepad and wordpad detach.  I can only guess that wordpad has some
functionality that notepad doesn't to detach itself, and that the cmd.exe
shell also has functionality that detects gui-based programs and detaches
them on launch (regardless of whether they're capable of detaching
themselves).

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/

1.57 on Win2k or WinXP. Not more than 16 com ports. Differences between //./comX and /dev/comX

[Lutz Hörl]

Hello,

I want to use (Windows-) COM port numbers greater
than 16, but when I use open() to get a file
descriptor for the devices I get the behaviour:

---1.case-

errno = 0;
fd = open("/dev/com8", O_RDWR | O_NONBLOCK);
-> fd = 3, errno = 0
-> everything is OK,
   I can use the opened COM port

---2.case-

errno = 0;
fd = open("\\.\com8", O_RDWR | O_NONBLOCK);

-> fd = 3, errno = 20 = ERROR_BAD_UNIT
   = "The system cannot find the device specified"
-> I can not use the opened COM port. The port
   number 1...16 versus 17...255 does NOT matter.

---3.case-

errno = 0;
fd = open("/dev/com23", O_RDWR | O_NONBLOCK);

-> fd = -1, errno = 2 = ERROR_FILE_NOT_FOUND
   = "The system cannot find the file specified"
-> I can not use the opened COM port. The port
   number 1...16 (OK) 17...255 (not OK) DOES matter.

---end cases

Is there a workaround?

Lutz Hoerl

Thorlabs GmbH
Gauss-Strasse 11
D-85757 Karlsfeld
Germany

Tel.+49 (0)8131 5956 44
FAX +49 (0)8131 5956 99
M A I Llhoerl at thorlabs dot com

--
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: [OT?] make bash script wait for called program to finish?

[Buchbinder, Barry (NIH/NIAID)]

For me (XP Pro) from cmd.exe, both notepad and write detach but from
command.com only write detaches, notepad does not.

MS must want us all to be like the scientist would wouldn't predict what
would happen unless s/he had done the experiment.

-Original Message-
From: Dave Korn
Sent: Monday, April 19, 2004 12:12 PM
To: [EMAIL PROTECTED]
Subject: RE: [OT?] make bash script wait for called program to finish?

> -Original Message-
> From: cygwin-owner On Behalf Of Buchbinder, Barry (NIH/NIAID)

> I was actually surprised that bash stopped when I did the experiment.
> 
> It wouldn't surprise me if it would depend on which program 
> one was trying
> to wait for, in addition to which version of windows. 

  Yep, it certainly seems that some windows gui apps have the behaviour of
detaching from the console, while others remain attached and so don't return
control to the shell until they're finished.  I noticed this in particular
when editing text files at an interactive bash prompt: if I use notepad.exe,
my shell is locked up until I exit, whereas if I use write.exe (wordpad), it
detaches and the shell prompt returns immediately.

  I also note as a data point that when run from a cmd.exe shell, both
notepad and wordpad detach.  I can only guess that wordpad has some
functionality that notepad doesn't to detach itself, and that the cmd.exe
shell also has functionality that detects gui-based programs and detaches
them on launch (regardless of whether they're capable of detaching
themselves).

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: 1.57 on Win2k or WinXP. Not more than 16 com ports. Differences between //./comX and /dev/comX

[Dave Korn]

> -Original Message-
> From: cygwin-owner On Behalf Of Lutz Hörl
> Sent: 19 April 2004 17:29

> Hello,
> 
> I want to use (Windows-) COM port numbers greater
> than 16, but when I use open() to get a file
> descriptor for the devices I get the behaviour:
> 
> ---1.case-
> 
> errno = 0;
> fd = open("/dev/com8", O_RDWR | O_NONBLOCK);
> -> fd = 3, errno = 0
> -> everything is OK,
>I can use the opened COM port
>
> ---2.case-
> 
> errno = 0;
> fd = open("\\.\com8", O_RDWR | O_NONBLOCK);
> 
> -> fd = 3, errno = 20 = ERROR_BAD_UNIT
>= "The system cannot find the device specified"
> -> I can not use the opened COM port. The port
>number 1...16 versus 17...255 does NOT matter.

  You got this one wrong.  The backslashes inside the quotes need to be
escaped, otherwise the first two will be parsed as a single escaped
backslash, and the second as '\c', which isn't a valid control char in C, so
you're likely to end up actually trying to open a device called "\.com8".
You must have ignored a warning message from the compiler when you tried
this.  WHY???!?

[EMAIL PROTECTED] /davek/test> cat testslash.c

#include 

int main (int argc, const char **argv)
{
  printf ("\\.\com8\n");
  return 0;
}

[EMAIL PROTECTED] /davek/test> gcc -o testslash testslash.c
testslash.c:6:11: warning: unknown escape sequence '\c'
[EMAIL PROTECTED] /davek/test> ./testslash.exe
\.com8
[EMAIL PROTECTED] /davek/test>

> ---3.case-
> 
> errno = 0;
> fd = open("/dev/com23", O_RDWR | O_NONBLOCK);
> 
> -> fd = -1, errno = 2 = ERROR_FILE_NOT_FOUND
>= "The system cannot find the file specified"
> -> I can not use the opened COM port. The port
>number 1...16 (OK) 17...255 (not OK) DOES matter.
> 
> ---end cases

  It's a dos/windoze limitation that only the low few com ports actually
have dos devices created for them.  8 is the upper limit, IIRC.  So you can
just open com1com8 from an ordinary DOS program, and so cygwin creates 8
com devices under /dev, because that's all it finds windoze has listed.  If
you want to use the others, you have to use the \\.\comX notation.
 
> Is there a workaround?

  Yep.  Use the \\.\comX notation and get it correct.  That should work.

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: Where to get cygwin's gmake sources, and does it build without errors?

[Christopher Faylor]

On Mon, Apr 19, 2004 at 11:21:36AM -0400, Sandy Currier wrote:
>I need to build the latest 3.81rc1 (beta1) version of gmake under
>cygwin.  However, the native gmake CVS distribution fails 6 make tests
>when I build it under cygwin.  The maintainer of gmake claims that FSF
>does not support make under cygwin, basically only because he has not
>seen the modifications that the 'cygwin' community has made to gmake.

The cygwin patches were sent to the make maintainer years ago but, at
the time, there was no reponse.

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

Re: Where to get cygwin's gmake sources, and does it build without errors?

[Larry Hall]

At 11:21 AM 4/19/2004, you wrote:
>Hi all,
>
>[hope this is an ok place to ask this question...]
>
>I need to build the latest 3.81rc1 (beta1) version of gmake under cygwin.
>However, the native gmake CVS distribution fails 6 make tests when I
>build it under cygwin.  The maintainer of gmake claims that FSF does not
>support make under cygwin, basically only because he has not seen the
>modifications that the 'cygwin' community has made to gmake.
>
>I have looked for the cygwin gmake sources for the cygwin distribution,
>but cannot find them on the web.  (This is nominally for 3.80 which appears
>to be the current version).




So run setup and download the source if you like.

Also see 


--
Larry Hall  http://www.rfk.com
RFK Partners, Inc.  (508) 893-9779 - RFK Office
838 Washington Street   (508) 893-9889 - FAX
Holliston, MA 01746 


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

Re: gentoo portage and cygwin

[Sven Köhler]

has anybody ever tried to port the gentoo-portage to cygwin? not that 
i expect many of the gentoo-ebuilds to compile, but the cygwin-people 
Take a look at 
http://sourceforge.net/project/showfiles.php?group_id=27249&package_id=19019 
the portage available there seems to work with python 2.2 only.
here's another portage that should run with python 2.3
http://gentoocygwin.sourceforge.net/

--
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: .logout files?

[Larry Hall]

At 03:11 AM 4/19/2004, you wrote:
>Larry Hall wrote:
>
>>At 02:43 PM 4/17/2004, you wrote:
>>
>>>On Sat, Apr 17, 2004 at 11:03:00AM -0700, Karl M wrote:
>>>
>On Sat, Apr 17, 2004 at 09:17:16AM +0200, Corinna Vinschen wrote:
>
>>On Apr 16 21:35, Christopher Spears wrote:
>>
>>>Hi! I can't seem to find a .logout file.  I found
>>
>>The reason might be that there is none.
>
>Sometimes the simple answers are just so paradoxically elusive.

And it can be so satisfying to knoy you have completely answered a
question without actually conveying any information.
>>>
>>>Bingo.  It's one of the many perks in being a cygwin developer.
>>That's just plain mean!  I think all cygwin developers should be banished!! ...Wait 
>>a minute, that won't work... ;-)
>
>Err... Maybe short time electricshock treatment would help a little.. =)


My computer tries that every once in a while with a static-electric 
discharge.  It hasn't done anything for me yet but since my computer 
is continuing the treatment, I can only assume the benefits require 
repeated exposure. 


>Well, what's this dot-logout file is used for anyway..?


Must be a script that is invoked on logout.  Nah, that's too simple.
Has to be something more complicated and sinister!


--
Larry Hall  http://www.rfk.com
RFK Partners, Inc.  (508) 893-9779 - RFK Office
838 Washington Street   (508) 893-9889 - FAX
Holliston, MA 01746 


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

Re: 1.57 on Win2k or WinXP. Not more than 16 com ports. Differences between //./comX and /dev/comX

[Larry Hall]

At 12:29 PM 4/19/2004, you wrote:
>Hello,
>
>I want to use (Windows-) COM port numbers greater
>than 16, but when I use open() to get a file
>descriptor for the devices I get the behaviour:
>
>---1.case-
>
>errno = 0;
>fd = open("/dev/com8", O_RDWR | O_NONBLOCK);
>-> fd = 3, errno = 0
>-> everything is OK,
>   I can use the opened COM port
>
>---2.case-
>
>errno = 0;
>fd = open("\\.\com8", O_RDWR | O_NONBLOCK);

^
You didn't escape the backslashes.  "//./com8" should work too.


>-> fd = 3, errno = 20 = ERROR_BAD_UNIT
>   = "The system cannot find the device specified"
>-> I can not use the opened COM port. The port
>   number 1...16 versus 17...255 does NOT matter.
>
>---3.case-
>
>errno = 0;
>fd = open("/dev/com23", O_RDWR | O_NONBLOCK);
>
>-> fd = -1, errno = 2 = ERROR_FILE_NOT_FOUND
>   = "The system cannot find the file specified"
>-> I can not use the opened COM port. The port
>   number 1...16 (OK) 17...255 (not OK) DOES matter.


Right.  Cygwin only maps the first 16 com ports to this
POSIX syntax.  The Win32 syntax as amended above should
work for you though.




--
Larry Hall  http://www.rfk.com
RFK Partners, Inc.  (508) 893-9779 - RFK Office
838 Washington Street   (508) 893-9889 - FAX
Holliston, MA 01746 


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

Cygwin make thinks a statement can be neither true nor false....

[Dave Korn]

  Haven't got any time to investigate further tonight, and I'm not sure if
this is a real bug or some misunderstanding of mine, but it doesn't appear
to be how the documentation ("info make") suggests things should work.
Here's a cut-down testcase:

-snip-
[EMAIL PROTECTED] /test/mk-test/test2> ls -lart
total 1
drwxr-xr-x+   4 dk   Domain U0 Apr 19 18:47 ..
-rw-r--r--1 dk   Domain U  402 Apr 19 18:49 makefile
drwxr-xr-x+   2 dk   Domain U0 Apr 19 18:49 .
[EMAIL PROTECTED] /test/mk-test/test2> cat makefile

ifdef ANYTHING
$(error: 1 ANYTHING is defined )
endif
ifndef ANYTHING
$(error: 1 ANYTHING is ***NOT*** defined )
endif

$(error: PLEASE give me an error.  What do I have to do, for god's sake?
Beg? )

$(warning: How about an error, or is even that asking too much of your
mighty in
tellect? )

.PHONY: all force
all: force

force:
echo Does make not use aristotelian two-valued logic then? > force

[EMAIL PROTECTED] /test/mk-test/test2>
[EMAIL PROTECTED] /test/mk-test/test2> make
echo Does make not use aristotelian two-valued logic then? > force
[EMAIL PROTECTED] /test/mk-test/test2>
-snip-

  Now, shouldn't at least one of those error messages be displayed?  Or have
I failed to understand that they don't get parsed because they're not
required to make sense of the targets and dependencies?



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 make thinks a statement can be neither true nor false....

[Robb, Sam]

Dave,

  "error:" != "error".  You want to use the following
syntax:

  $(error 1 ANYTHING is defined )

  Possibly a bug in make, as I'd expect it to complain
about an undefined function named "error:".  Similar
constructs are also silently ignored:

  $(foo This isn't a valid make function)
  $(bar Neither is this)

-Samrobb

> -Original Message-
> From: Dave Korn [mailto:[EMAIL PROTECTED]
> Sent: Monday, April 19, 2004 1:53 PM
> To: [EMAIL PROTECTED]
> Subject: Cygwin make thinks a statement can be neither true nor
> false
> 
> 
> 
>   Haven't got any time to investigate further tonight, and 
> I'm not sure if
> this is a real bug or some misunderstanding of mine, but it 
> doesn't appear
> to be how the documentation ("info make") suggests things should work.
> Here's a cut-down testcase:
> 
> -snip-
> [EMAIL PROTECTED] /test/mk-test/test2> ls -lart
> total 1
> drwxr-xr-x+   4 dk   Domain U0 Apr 19 18:47 ..
> -rw-r--r--1 dk   Domain U  402 Apr 19 18:49 makefile
> drwxr-xr-x+   2 dk   Domain U0 Apr 19 18:49 .
> [EMAIL PROTECTED] /test/mk-test/test2> cat makefile
> 
> ifdef ANYTHING
> $(error: 1 ANYTHING is defined )
> endif
> ifndef ANYTHING
> $(error: 1 ANYTHING is ***NOT*** defined )
> endif
> 
> $(error: PLEASE give me an error.  What do I have to do, for 
> god's sake?
> Beg? )
> 
> $(warning: How about an error, or is even that asking too much of your
> mighty in
> tellect? )
> 
> .PHONY: all force
> all: force
> 
> force:
>   echo Does make not use aristotelian two-valued logic 
> then? > force
> 
> [EMAIL PROTECTED] /test/mk-test/test2>
> [EMAIL PROTECTED] /test/mk-test/test2> make
> echo Does make not use aristotelian two-valued logic then? > force
> [EMAIL PROTECTED] /test/mk-test/test2>
> -snip-
> 
>   Now, shouldn't at least one of those error messages be 
> displayed?  Or have
> I failed to understand that they don't get parsed because they're not
> required to make sense of the targets and dependencies?
> 
> 
> 
> 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/
> 
> 

--
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/

2nd Try 1.57 on Win2k or WinXP. Not more than 16 com ports. Differences between //./comX and /dev/comX

[Lutz Hörl]

Hello,

thank you Dave, thank you Larry.

But unfortunately your advice to use "//./devX" did NOT solve the problem.

Dave:
Cygwin uses upto 16 COM ports, not only 8. I did not compile with
open("\\.\com8",...) but with open(buf,...) so I did not get
warnings from the compiler.

I corrected the examples below and got the result described.
What makes me nervous ist the errno = 20.

I must add that I use Win32-API to set baudrate and to read the modem
lines (DSR, DCD). I use the Win32 functions
GetCommModemStatus()
EscapeCommFunction()
GetCommState()
SetCommState()
in my program. That works fine as long as I use "/dev/comX" with X=1...16.
But it does not work with "//./comX".




I want to use (Windows-) COM port numbers greater
than 16, but when I use open() to get a file
descriptor for the devices I get the behaviour:

---1.case-

errno = 0;
fd = open("/dev/com8", O_RDWR | O_NONBLOCK);
-> fd = 3, errno = 0
-> everything is OK,
   I can use the opened COM port

---2.case-

errno = 0;
fd = open("//./com8", O_RDWR | O_NONBLOCK);

-> fd = 3, errno = 20 = ERROR_BAD_UNIT
   = "The system cannot find the device specified"
-> I can not use the opened COM port. The port
   number 1...16 versus 17...255 does NOT matter.

---3.case-

errno = 0;
fd = open("/dev/com23", O_RDWR | O_NONBLOCK);

-> fd = -1, errno = 2 = ERROR_FILE_NOT_FOUND
   = "The system cannot find the file specified"
-> I can not use the opened COM port. The port
   number 1...16 (OK) 17...255 (not OK) DOES matter.

---end cases

Is there a workaround?

Lutz Hoerl

Thorlabs GmbH
Gauss-Strasse 11
D-85757 Karlsfeld
Germany

Tel.+49 (0)8131 5956 44
FAX +49 (0)8131 5956 99
M A I Llhoerl at thorlabs dot com

--
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 make thinks a statement can be neither true nor false....

[Dave Korn]

> -Original Message-
> From: cygwin-owner On Behalf Of Robb, Sam
> Sent: 19 April 2004 18:57

> 
> Dave,
> 
>   "error:" != "error".  You want to use the following
> syntax:
> 
>   $(error 1 ANYTHING is defined )
> 

  Ach, thanks.  I had a space between the 'error' and the ':' everywhere
else I've been using it like that, but I forgot this time, and suddenly
couldn't work out what was going wrong.

>   Possibly a bug in make, as I'd expect it to complain
> about an undefined function named "error:".  Similar
> constructs are also silently ignored:
> 
>   $(foo This isn't a valid make function)
>   $(bar Neither is this)
> 
> -Samrobb

  Yeh.  And when it can't find an include file, I'd like it to complain,
rather than just silently add it to a list of missing dependencies, then get
to the end of the file and only complain about the first missing dependency.
As it stands I can't see any difference between using "-include" and
"include".

   gmake needs a LOT of work in the
emission-of-useful-and-accurate-diagnostics department.  All too often it
just silently does nothing without explaining why, and the only choice you
get is that or a million tons of debugging output which doesn't usually
contain any useful information whatsoever about what make is actually doing
and why.  Gah!



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/

Need copy of /etc/profile

[LarrysPCRemedies]

I very stupidly overwrote my /etc/profile file. Can someone send me (directly or 
through the mailing list) a pristine copy of it, or is there a way to get just that 
file from CYGWIN's setup?

(very embarrassed)
Larry

--
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: 2nd Try 1.57 on Win2k or WinXP. Not more than 16 com ports. Differences between //./comX and /dev/comX

[Corinna Vinschen]

On Apr 19 20:14, Lutz H?rl wrote:
> Hello,
> 
> thank you Dave, thank you Larry.
> 
> But unfortunately your advice to use "//./devX" did NOT solve the problem.

No, it can't.  When using the Win32 device names, the files are not
recognized as devices by Cygwin but instead they are treated like 
normal files.  If you want Cygwin support, you must use POSIX device
names.  If you want to access the serial interfaces using Win32 functions
anyway, I'd suggest to use Win32 functions throughout and not rely on
Cygwin.  This already begins with using CreateFile instead of open.

Corinna

-- 
Corinna Vinschen  Please, send mails regarding Cygwin to
Cygwin Co-Project Leader  mailto:[EMAIL PROTECTED]
Red Hat, Inc.

--
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: rsync question

[Peter Kok]

> At 12:08 AM 4/17/2004 +0200, Corinna Vinschen wrote:
> >On Apr 16 15:44, Peter Kok wrote:
> 
> >> Q2: Could nontsec work with public key authentication?  I have 
granted 
> >> the account with several local user rights, "create token object, 
> >> logon 
> >> as a service' and 'replace a process level token'
> >
> >Did you give the SYSTEM account the right to read your ~/.ssh 
directory
> >and the files in it?  Does the service know about nontsec (set CYGWIN
> >in global windows environment or through cygrunsrv)?  Is StrictModes 
set
> >to no in /etc/sshd_config?
> 
> >From Peter's question it's not clear if his sshd is running as 
SYSTEM.
> If it is, then granting the privileges to the user should not be
> necessary, but that doesn't explain the problem.
> 
> I can reproduce on an NT system, with sshd running as SYSTEM,
> but I can't explain it. Part of the debug output of ssh is given
> below, with and without ntsec. The difference is in the last few
> lines.
> 
> Pierre
> 
>It's a problem with the ntsec specific test in OpenSSH itself.  The
>test requires ntsec to be turned on for switching user context w/o
>password.  This isn't required anymore for a while but the test in
>OpenSSH still insists on ntsec for pubkey auth.
>
>I've send a patch to the portable OpenSSH developers list which 
>hopefully
>makes it into 3.8.1p1, which is due RSN.
>
>Corinna

Thank you Corinna for your quick response.  I just saw that OpenSSH has 
release 3.81pl on April 19, but unfortunately, it didn't include the 
new patch.  Can you please show me tips/website onto how I could 
compile the new modified bsd-cygwin_util.c to be used by cygwin?

Thanks in advance,
Peter


--
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: rsync question

[Corinna Vinschen]

On Apr 19 15:46, Peter Kok wrote:
> > At 12:08 AM 4/17/2004 +0200, Corinna Vinschen wrote:
> >I've send a patch to the portable OpenSSH developers list which 
> >hopefully
> >makes it into 3.8.1p1, which is due RSN.
> >
> >Corinna
> 
> Thank you Corinna for your quick response.  I just saw that OpenSSH has 
> release 3.81pl on April 19, but unfortunately, it didn't include the 
> new patch.

It did.  I've just uploaded 3.8.1p1.

Corinna

-- 
Corinna Vinschen  Please, send mails regarding Cygwin to
Cygwin Co-Project Leader  mailto:[EMAIL PROTECTED]
Red Hat, Inc.

--
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: Unix time conversion error

[Mark Thornton]

[EMAIL PROTECTED] wrote:

Hello,

I have a conversion problem. When I convert UNIX timestamps
to datetime representations, they are one hour off. Does this have to
with the following mail (indeed we are now in DST) ?
http://www.cygwin.com/ml/cygwin/2003-10/msg00995.html
 

If you mean that timestamps for date/times in winter (i.e. not DST) are
shown incorrectly, then yes this is a problem that afflicts many
applications. From the referenced MS document:
"By rewriting the affected application(s) to use the appropriate Win32
API calls, this problem can be corrected. Under NTFS, the Win32 API
GetFileTime"
Anything which uses the standard C calls unfortunately gets an adjusted
value, and this is what most applications ported from Unix do, resulting
in this problem. Microsoft did this to minimise the surprise to people
using FAT file systems which were most common at the time. The decision
to use the offset for today on all timestamps including those applying
to a date when the offset would have (or will be) different was
deliberate (Microsoft refuse to consider it a bug).
I suppose we could ask whether cygwin should replace the relevant calls
with better alternatives. Otherwise implement your own or live with the
pain.
Mark Thornton



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

RE: CVS + SSH + Binary File = ^M

[Hannu E K Nevalainen]

> From: Dave Korn
> Sent: Monday, April 19, 2004 5:29 PM

Commenting a message from kgrizzle
> > Sent: 19 April 2004 16:15

> Basically I think it's a bad idea to mount your filesystems in textmode
> and then complain "But it keeps adding \r to my line endings".
> Accessing a binary file through a textmode mountpoint is a dubious
> thing to expect to work.  I personally think the most reliable strategy
> is to accept that we're working with a POSIX system, standardize on
> Unix EOLs throughout your files, mount your filesystems in binary mode,
> and only ever have Unix<=>DOS EOL conversions by deliberately invoking
> d2u/u2d manually.

I can only agree with the above; I've been using this strategy for a couple
of decades - yet to be bitten hard.

My situation for a long time:
 AmigaOS (LF line endings)
   <= versus =>
 MS DOS and whatever came after it (The well known CRLF line endings)


/Hannu E K Nevalainen, B.Sc. EE - 59+16.37'N, 17+12.60'E

** on a mailing list; please keep replies on that particular list **

-- printf("LocalTime: UTC+%02d\n",(DST)? 2:1); --
--END OF MESSAGE--


--
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: Need copy of /etc/profile

[Ross Boulet]

> -Original Message-
> From: cygwin-owner On Behalf
> Of [EMAIL PROTECTED]
> 
> I very stupidly overwrote my /etc/profile file. Can someone send me
> (directly or through the mailing list) a pristine copy of it, or is there
> a way to get just that file from CYGWIN's setup?
> 
> (very embarrassed)
> Larry
> 

Try looking for /etc/defaults/etc/profile.

HTH



--
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/

RFI: Emulating hard links on FAT et al.

[A . Alper Atici]

Hello,

I've been pondering over the prospects of emulating hard links for
some time. List archives don't show much about it, and I have not come
across any similar open implementation on the net. 

My rudimentary idea of emulating hard links is based on employing a
new type of windows shortcut which will be regarded as a hardlinking
file, rather than a symlink, by Cygwin. For this, I hope to figure out
a possible combination in the magic bitvector byte(word?) in shortcut
header. Any comments? How about 0x1c?


--
A. Alper Atici   OpenPGP KeyID: 0xB824F550


--
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/

Cygwin (december 2003 version) and strcat

[Martin Johansen]

Hi.

I am experiencing problems using strcat and strncat.

For some reason, these functions do not appent a zero.
I am using sprintf for now, but this is not an elegant solution.

How can I fix cygwin to handle these simple functions?

Thanks!


--
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 make thinks a statement can be neither true nor false....

[Ross Ridge]

>  Possibly a bug in make, as I'd expect it to complain
>about an undefined function named "error:". 

I would've expected it to complain about a bad substition reference,
ie. it's missing an "=".

> Similar constructs are also silently ignored:
>
>  $(foo This isn't a valid make function)
>  $(bar Neither is this)

Since "foo" and "bar" aren't functions supported by GNU Make these are
just simple variable references.  Eg:

foo This isn't a valid make function=one
bar Neither is this=two

test:
echo $(foo This isn't a valid make function)
echo $(bar Neither is this)

Ross Ridge

-- 
 l/  //   Ross Ridge -- The Great HTMU
[oo][oo]  [EMAIL PROTECTED]
-()-/()/  http://www.csclub.uwaterloo.ca/u/rridge/ 
 db  //   

--
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 make thinks a statement can be neither true nor false....

[Robb, Sam]

Interesting - I'll have to go back and reread the make docs to
fix this in my mind.
 
Thanks for teaching me something new today :-)
 
-Samrobb

-Original Message- 
From: Ross Ridge [mailto:[EMAIL PROTECTED] 
Sent: Mon 4/19/2004 9:40 PM 
To: [EMAIL PROTECTED] 
Cc: 
Subject: RE: Cygwin make thinks a statement can be neither true nor false



>  Possibly a bug in make, as I'd expect it to complain
>about an undefined function named "error:".

I would've expected it to complain about a bad substition reference,
ie. it's missing an "=".

> Similar constructs are also silently ignored:
>
>  $(foo This isn't a valid make function)
>  $(bar Neither is this)

Since "foo" and "bar" aren't functions supported by GNU Make these are
just simple variable references.  Eg:

foo This isn't a valid make function=one
bar Neither is this=two

test:
echo $(foo This isn't a valid make function)
echo $(bar Neither is this)

Ross Ridge

--
 l/  //   Ross Ridge -- The Great HTMU
[oo][oo]  [EMAIL PROTECTED]
-()-/()/  http://www.csclub.uwaterloo.ca/u/rridge/
 db  //  

--
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: Wishlist additions for setup.exe

[GregMo]

Since the topic is a patch to setup.exe, let me restate an earlier oversight
with it that I mentioned, and add a new one I've discovered through hours of
trial and error...

The first of these is an issue with changing the install options from
"default" on 'All' to "Install".  When you do this, setup.exe poorly decides
that you don't need "base-files" or "aspell-dev", and only those two, which
results in a broken install.

Secondly, and this one took me hours on end, and still not completely
resolved, when you install Cygwin you're given the option to install for
"All Users" or "Just Me".  Off hand I don't know what the default is, but
apparently when I installed it "Just Me" was selected.  I know this, because
when you do this it hoses up the permissions to /var/log so that sshd can't
write to it's log file at all, and hence will not start.

Being on 'XP Home' I still can't get sshd to run as a service, but at least
crond will and I can use it to make sure that sshd is running.  I think I
know a fix for that too, though, just haven't yet had the chance or pressing
desire to test it, especially now that I can run it through crond.

Cheers,
Greg



---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.659 / Virus Database: 423 - Release Date: 4/15/2004


--
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/

incorrect dos version

[Daniel Senderowicz]

Hi,

I'm new to this list. I read the FAQ but I couldn't find an answer
nor a fix for my problem. I just installed the latest cygwin version
(1.5.9) on a PC running windows NT-4.0. Everything went OK, and I
can run programs from the console. However when I try to 'startx'
it comes back with the message "Incorrect DOS version". I was
wondering if this is something that can be fixed or it happens with
all NT systems. Many thanks in advance.

Dan

--
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: Wishlist additions for setup.exe

[Larry Hall]

At 10:47 PM 4/19/2004, you wrote:
>Since the topic is a patch to setup.exe, let me restate an earlier oversight
>with it that I mentioned, and add a new one I've discovered through hours of
>trial and error...
>
>The first of these is an issue with changing the install options from
>"default" on 'All' to "Install".  When you do this, setup.exe poorly decides
>that you don't need "base-files" or "aspell-dev", and only those two, which
>results in a broken install.
>
>Secondly, and this one took me hours on end, and still not completely
>resolved, when you install Cygwin you're given the option to install for
>"All Users" or "Just Me".  Off hand I don't know what the default is, but
>apparently when I installed it "Just Me" was selected.  I know this, because
>when you do this it hoses up the permissions to /var/log so that sshd can't
>write to it's log file at all, and hence will not start.


"All Users" creates all the necessary default mount points for all users.
"Just Me" creates them just for the user that installs Cygwin.  Without 
the mount points for all users ('mount -s'), services generally won't start
because they don't have access to those necessary mounts.  This doesn't
directly affect the permissions on particular directories though.


>Being on 'XP Home' I still can't get sshd to run as a service, but at least
>crond will and I can use it to make sure that sshd is running.  I think I
>know a fix for that too, though, just haven't yet had the chance or pressing
>desire to test it, especially now that I can run it through crond.


Sounds like a sshd configuration problem.


--
Larry Hall  http://www.rfk.com
RFK Partners, Inc.  (508) 893-9779 - RFK Office
838 Washington Street   (508) 893-9889 - FAX
Holliston, MA 01746 


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

Re: Wishlist additions for setup.exe

[GregMo]

From: "Larry Hall" <[EMAIL PROTECTED]>
Subject: Re: Wishlist additions for setup.exe


> "All Users" creates all the necessary default mount points for all users.
> "Just Me" creates them just for the user that installs Cygwin.  Without
> the mount points for all users ('mount -s'), services generally won't
start
> because they don't have access to those necessary mounts.  This doesn't
> directly affect the permissions on particular directories though.

Services won't start, precisely the point.  The requested change here is
that when "Just Me" is selected that a warning would be displayed indicating
this fact.

> >Being on 'XP Home' I still can't get sshd to run as a service, but at
least
> >crond will and I can use it to make sure that sshd is running.  I think I
> >know a fix for that too, though, just haven't yet had the chance or
pressing
> >desire to test it, especially now that I can run it through crond.
>
>
> Sounds like a sshd configuration problem.

Search the net...  It's not a configuration problem, or if it is, it's
affected many many users without any apparent resolution available.  Have a
glance at the results returned by Google for these search texts:

"win32 error 1062"
cygwin "Could not load host key"

You'll see, as I have, that it's all but assuredly not a configuration
issue.  The likely issue here is one with rights (Not owned by 'SYSTEM' and
dir attributes).  The two directories that this is entails are /var/log and
/var/empty.  As I said prior, I believe I know of a way of resolving the
issue, but it entails rebooting to safe mode and other details which isn't
important to someone that's gotten a fair kludge working.

Cheers,
Greg



---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.659 / Virus Database: 423 - Release Date: 4/15/2004


--
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/

Bugs in sed and uudecode? md5sum too?

[GregMo]

If you will, have a look at these two commands performed on both a cygwin
install and a redhat install, on the exact same file
(http://216.40.230.24/part.1)

  CygWin
[EMAIL PROTECTED]: ~/.brag/all-servers/unfinished/1082426201]$ cat part.1
|sed -e {/^$/d} |uudecode

[EMAIL PROTECTED]: ~/.brag/all-servers/unfinished/1082426201]$ md5sum part.1
915e506f04e8618aff6ab0e9f37a80d9 *part.1


  RedHat
[EMAIL PROTECTED]: ~/tmp]$ cat part.1 |sed -e {/^$/d} |uudecode
sed: -e expression #1, char 7: Extra characters after command
uudecode: stdin: No `begin' line

[EMAIL PROTECTED]: ~/tmp]$ md5sum part.1
915e506f04e8618aff6ab0e9f37a80d9  part.1

The project here is porting a script used daily on redhat to cygwin.  It
turns out there are quite a few glitches, and this is the biggest so far, so
thought we'd report it.  The two problems here are both with error
reporting.  You'll note that sed on a linux machine returns an error when
this command is tried, whereas cygwin doesn't.  With sed being issued before
uudecode, on the linux machine, uudecode isn't getting the output so it
returns no begin line.  When the output is piped directly to uudecode on
linux an error is returned about the 'end' line, or lack thereof, but no
error on cygwin.  This is nightmarish for debugging.

Additionally, have a look at the md5sum output from cygwin.  I happen to
know that an '*' before a file name in a checksum doesn't hinder using those
sums on other systems, but that doesn't make it not a bug, does it?  

Cheers,
Greg


---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.659 / Virus Database: 423 - Release Date: 4/15/2004


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

[ANNOUNCEMENT] Updated: exim-4.32-1

[Pierre A. Humblet]

I have updated Exim, the Mail Transfer Agent, to exim-4.32-1.
It is the first stable release since the current 4.31.

News from the official announcement:

---
This release fixes a serious recipient callout caching bug that was introduced
in 4.31. Also, I have reversed the change in 4.31 that caused Exim to use the
real sender when doing recipient callouts (which is where the bug was). The
default behaviour is now what it was in 4.30. However, there is a new option
to request the altered behaviour.

---
For more information see
  
  
  and /usr/share/doc/exim-4.32/NewStuff


Notes:
 This port requires Cygwin 1.5.6 or more recent.
 If you update from an Exim release earlier than 4.22, it will be necessary
 to delete all files in /var/spool/exim/db. This causes no permanent loss of
 data.

To update your installation, click on the "Install Cygwin now" link on
the http://www.cygwin.com web page.  This downloads setup.exe
to your system.  Then run setup and answer all of the questions.
Exim is in the Mail category.

If you have questions or comments, please send them to the Cygwin
mailing list: [EMAIL PROTECTED], mentioning exim in the subject line.

Pierre


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

Re: Updated: exim-4.32-1

[Jason S. Reingold]

I will be out of the country from March 30th till May 4th.  If you need to
reach me by email, please use [EMAIL PROTECTED]  If you need to reach me by
phone, please contact medical student affairs for my emergency contact
information.

Sincerely,
Jason Reingold


--
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: Wishlist additions for setup.exe

[Rodrigo Medina]

Hi all,

My whishes for SETUP:

1- In the list of packages include its size. It is very important
for people with slow connexions, so they can easily program
what to download and in which order. For example you may decide
not to download an updated 16MB fonts package with minor changes.
Right now you discover that the package has 16MB only after the
downloading of that package has started. Then you have to abort
setup and start again. (This is related to point 2-).

2- When you are downloading into a temporal directory, SETUP
downloads the packages even if they are already in the temporal
directory but are still not installed. I think this is a real  nuisance.
 Consider the case that for some reason the downloading
is interrupted (this is likely to happen if you have a slow connexion).
Say you have already downloaded 90% of the packages (that could take
hours). SETUP starts again from the beginning. This can lead to a
never ending process.
You have two choices:
a) To install what was already downloaded before starting a new
setup-downloading session. This in some cases could be dangerous
when you are installing packages that require some other package
that you have not still downloaded.
b) To look at the files in the temporal directory to check what
files are already downloaded and mark those files KEEP (or SKIP)
in the package list. Should not be SETUP an automatic process?

I think that to solve these two problems require minor changes
of SETUP and would really improve its performance.

bye,

Rodrigo Medina



--
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: incorrect dos version

[Olof Lagerkvist]

Daniel Senderowicz wrote:

Hi,

I'm new to this list. I read the FAQ but I couldn't find an answer
nor a fix for my problem. I just installed the latest cygwin version
(1.5.9) on a PC running windows NT-4.0. Everything went OK, and I
can run programs from the console. However when I try to 'startx'
it comes back with the message "Incorrect DOS version". I was
Check if you have any old DOS command line tools in your path with names
like any cygwin executables, i.e. grep, cut, tar, gzip etc.
wondering if this is something that can be fixed or it happens with
all NT systems. Many thanks in advance.
It has most likely nothing to do with NT.

--
Olof Lagerkvist
ICQ: 724451
Web page: http://here.is/olof


--
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: problem installing libiconv-1.9.2

[Charles Wilson]

Steve Kelem wrote:
What's the trick to getting libiconv to install?  If I try to do "make 
install" from tcsh, I get the message:
/bin/install -c -m 644 .libs/cygiconv-2.dll $dldir/cygiconv-2.dll
/bin/install: cannot remove `/usr/lib/../bin/cygiconv-2.dll': Permission 
denied

I assume that this is because I'm using cygwin/tcsh.
probably, the DLL is "in use".  What I do is a
   'make install DESTDIR=...'
and install iconv into a tmp directory.  From there I can (a) close all 
cygwin apps and manually copy the files to the right places, or more 
typically (b) create a setup-compatible installation tarball and then 
use setup.exe to install it (setup handles scheduling of 
finish-install-after-reboot, etc.)

Aint MSWindows grand?

The main reason I'm trying to recompile iconv is because after building 
gd-2.0.22, I cannot get the perl module GD-2.12 to link, so I'm trying 
to build a more recent copy (1.9.2) of libiconv:
Ah, the gentle nudge...

Any updates will have to wait until I'm not working 80+ hours a week. 
Which means late May, it seems.

--
Chuck
--
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/