RE: scp exits often with -1

[Jörg Schaible]

Reini Urban wrote on Wednesday, November 17, 2004 7:21 PM:

> Hi Jörg :)
> 
>> So, if you can't or won't help, I prefer no answer.
> 
> Downgrading to cygwin-1.3.11 should help.

Hmmm. My old cygwin.dll was not *that* old. 1.5.9 or so (unfortunately I 
deleted my download cache to be sure to gain the proper X packages). 
Downgrading may break more than just ssh, cygwin dll's are only proposed to be 
*backward* compatible.

> or if you like living on the edge there's a new experimental
> "noparent" branch for cygwin1.dll

If Chris explicitly don't want people to use it, I dunno.

I would really appreciate to isolate the problem.

- Jörg

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

Re: Perl-tk?

[Brian Dessent]

dwycoff wrote:

> I tried using a perl-tk script in cygwin that I wrote more than
> a year ago.  Whenever I run it it dies at:
> 
>   Can't locate Tk.pm in @INC (@INC contains: /usr/lib/perl5/5...
> 
> I did a search and couldn't find Tk.pm.  So I looked to the cygwin
> website, but didn't find any mention of it there, either.
> 
> Has perl-tk made it to cygwin?

Perl modules are installed from CPAN, not through Cygwin setup (with the
exception of libwin32 which does not compile without patches.)  Have you
tried installing the module with CPAN?

Brian

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

Perl-tk?

[dwycoff]

I tried using a perl-tk script in cygwin that I wrote more than
a year ago.  Whenever I run it it dies at:
 Can't locate Tk.pm in @INC (@INC contains: /usr/lib/perl5/5...
I did a search and couldn't find Tk.pm.  So I looked to the cygwin
website, but didn't find any mention of it there, either.
Has perl-tk made it to cygwin?
-d
--
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: m4-1.4.2-1

[Eric Blake]

m4 has been updated to version 1.4.2

NEWS

This release patches a security hole in m4 1.4: the maketemp macro now 
creates the temporary file, rather than just returning an unused filename. 
Additionally, the documentation has been moved to the /usr/share/info directory.

DESCRIPTION
===
m4 is an implementation of the traditional Unix macro processor. It is 
mostly SVR4 compatible although it has some extensions (for example, 
handling more than 9 positional parameters to macros). GNU m4 also has 
built-in functions for including files, running shell commands, doing 
arithmetic, etc.

UPDATE
==
To update your installation, click on the "Install Cygwin now" link on
the http://cygwin.com/ web page.  This downloads setup.exe to your
system.  Save it and run setup, answer the questions and pick up
'm4' from the 'Interpreters' category.

DOWNLOAD:
=
Note that downloads from sources.redhat.com (aka cygwin.com) aren't
allowed due to bandwidth limitations.  This means that you will need
to find a mirror which has this update, please choose the one nearest
to you: http://cygwin.com/mirrors.html

QUESTIONS:
==
If you want to make a point or ask a question the Cygwin mailing
list is the appropriate place.

--
Eric Blake


CYGWIN-ANNOUNCE UNSUBSCRIBE INFO:
=
To unsubscribe to the cygwin-announce mailing list, look at the
"List-Unsubscribe: " tag in the email header of this message.  Send
email to the address specified there.  It will be in the format:

[EMAIL PROTECTED]

If you need more information on unsubscribing, start reading here:

http://sources.redhat.com/lists.html#unsubscribe-simple

Please read *all* of the information on unsubscribing that is available
starting at this URL.


--
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 with setup

[Bobby McNulty]

Christopher Faylor wrote:
On Thu, Nov 18, 2004 at 11:09:26AM +1100, Luke Kendall wrote:
 

Sorry to be causing you such hassles.  I'm hoping to get a chance to
try the latest Cygwin:
   

You can try the latest cygwin.  Just install everything in two or
three steps.
This is not the final solution, of course, but this problem doesn't
seem like a total showstopper to me.  Bobby McNulty can probably provide
more details.  In fact, I think he's already done so.
cgf
--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/
 

First install all of the main cygwin without the gnome stuff, then 
install the x stuff with gnome. Go to Sourceforge and get KDE on Cygwin.
rebaseall.
startx
and you shound have the KDE desktop on X with cygwin.
Neat, huh?
It works for me.
Now, I'm trying to find a good MIDI sequencer that runs under KDE and 
Cygwin.
Rosegarden has too many problems I don't want to deal with.
Bobby

--
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 with setup

[Brian Dessent]

Luke Kendall wrote:

> > > I can confirm this.  We took a snapshot of October 26th
> > > (setup-timestamp: 1098732614), and in that version of setup,
> >
> > That number you quoted identifies the version of the setup.ini datafile, not
> > the setup.exe program.
> 
> Sorry, how's this?
> $  bunzip2 -c setup.bz2 | head -5
> # This file is automatically generated.  If you edit it, your
> # edits will be discarded next time the file is generated.
> # See http://cygwin.com/setup.html for details.
> #
> setup-timestamp: 1098732614

That's still just telling you the timestamp when setup.ini was created. 
It has no bearing on the version of setup.exe, which is displayed on the
first panel, e.g. "Setup.exe version 2.431".

Brian

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

Re: Problem with setup

[Christopher Faylor]

On Thu, Nov 18, 2004 at 11:09:26AM +1100, Luke Kendall wrote:
>Sorry to be causing you such hassles.  I'm hoping to get a chance to
>try the latest Cygwin:

You can try the latest cygwin.  Just install everything in two or
three steps.

This is not the final solution, of course, but this problem doesn't
seem like a total showstopper to me.  Bobby McNulty can probably provide
more details.  In fact, I think he's already done so.

cgf

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

Re: Problem with setup

[Luke Kendall]

On 17 Nov, Max Bowsher wrote:
>  Luke Kendall wrote:
> > I should also add that after a setup.exe crash at 9:38am after being
> > left running overnight, later attempts to install that day (with the VM
> > size steadily increasing, and setup.exe appearing to "stick" at the
> > stage where it is installing the zip package), that setup.log was not
> > updated by the later installation attempts.
>  >
> > Shouldn't setup.log be getting opened in append mode?  We examined it
> > at various times during a later installation attempt, and the contents
> > didn't change and the file modification time stayed at 9:38am.
>  
>  At the moment, the log files are written to disc in one bulk write when 
>  setup.exe completes. That's probably not a great design feature.

Ah.  Absolutely.  If setup crashes, you have no debug information!
I usually fflush after writing an entry to a log file - it's more
important to have accurate debug info than to save a few seconds on
execution time, I reckon.

And in other emails:

> > I can confirm this.  We took a snapshot of October 26th
> > (setup-timestamp: 1098732614), and in that version of setup,
> 
> That number you quoted identifies the version of the setup.ini datafile, not 
> the setup.exe program.

Sorry, how's this?
$  bunzip2 -c setup.bz2 | head -5
# This file is automatically generated.  If you edit it, your
# edits will be discarded next time the file is generated.
# See http://cygwin.com/setup.html for details.
#
setup-timestamp: 1098732614

> > none of
> > the file handles for any of the temp files of the form (where c:\temp
> > is your temporary package directory) c:\temp\http%.\package.tar.bz2
> > is ever released.  The system starts page faulting and the number of
> > handles just grows and grows, and eventually setup dies like this:
> >
> > Microsoft Visual C++ Runtime Error
> > Runtime Error!
> > Program \\samba\install\win32\cygwin\setup.exe
> > This application has requested the Runtime to terminate it in an
> > unusual way.  Please contact the application's support team for
> > more information.
> 
> 
> That's nasty. I wonder why we haven't seen more reports of this.
> 
> Is there any reason why your machine should be more susceptible to this? 
> (Low memory, any kind of resource limits?)

Not that I'm aware of, no.  One machine was dual processor, 1GB of
memory.  Interestingly, looking at Task Manager during setup.exe's run,
even though there seemed to be plenty of physical memory, it changed to
display page fault activity rather than RAM usage.  Tweaking TM to show
page fault activity and handles showed that there was a lot of page
faults from setup, and a steadily growing number of handles (the
..tar.bz2 files which were never released).

The other PC that showed the problem had 512MB of memory, and no
special resource limits.

> >> At the crash point, setup.exe has some 600 open handles and 2 Gb of VM. 
> >> It's
> >> accumulated 2.5 million page faults.
> 
> Wow!
> 
> Well, it's not quite as dramatic here, but I can see a good few leaked 
> handles in Process Explorer.

Yes, we looked in PE too, and it was clear that no .tar.bz2 file handle
was released.  We were trying a fresh install, and selected "All" (i.e.
every package).

> I've committed a fix, but unfortunately, when I begun testing to prepare for 
> a release, I ran into all sorts of nasty problems, including being unable to 
> reinstall Cygwin, having uninstalled it to test setup. I'm going to keep 
> working on it though... after I've completely purged Cygwin from my 
> computer, reinstalled, and set up a VM to do tests in without disrupting my 
> primary machine. Might take a while.

Sorry to be causing you such hassles.  I'm hoping to get a chance to
try the latest Cygwin:

# This file is automatically generated.  If you edit it, your
# edits will be discarded next time the file is generated.
# See http://cygwin.com/setup.html for details.
#
setup-timestamp: 1100546431

luke


--
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: installer crash on post-instal scripts

[Bobby McNulty]

Kris Matthews wrote:
What worked for me:
1. Completely wipe out existing cygwin installation
2. Install all of cygwin *except* X.
3. .. Install X stuff, by itself.
Regards,
Kristopher
On Wed, 17 Nov 2004, Jing Zhou wrote:
 

Hi Kris,
I have the same problem of installing full cygwin 1.5.12.1. It froze
at 93% when installing postinstall in X11 directory. The disk size
continued growing. It was 2G when I eventally decided to delete it.
Then my computer crashed.
It seems to me that you found a solution. Could you please post the solution?
My computer is windows xp home.
thanks,
-Jing
Kris's original message is:
installer crash on post-instal scripts
From: Kris Matthews 
To: cygwin at cygwin dot com
Date: Wed, 17 Nov 2004 11:51:05 -0500 (EST)
Subject: installer crash on post-instal scripts

When I try to install cygwin on windows xp, the installer hangs at around
90%, when it's running the X post install scripts. (From a full install)
   

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

Yes, and KDE works for me now. Don't forget to rebaseall.

--
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: installer crash on post-instal scripts

[Kris Matthews]

What worked for me:

1. Completely wipe out existing cygwin installation
2. Install all of cygwin *except* X.
3. .. Install X stuff, by itself.

Regards,
Kristopher

On Wed, 17 Nov 2004, Jing Zhou wrote:

> Hi Kris,
>
> I have the same problem of installing full cygwin 1.5.12.1. It froze
> at 93% when installing postinstall in X11 directory. The disk size
> continued growing. It was 2G when I eventally decided to delete it.
> Then my computer crashed.
>
> It seems to me that you found a solution. Could you please post the solution?
>
> My computer is windows xp home.
>
> thanks,
>
> -Jing
>
> Kris's original message is:
>
> installer crash on post-instal scripts
>
> From: Kris Matthews 
> To: cygwin at cygwin dot com
> Date: Wed, 17 Nov 2004 11:51:05 -0500 (EST)
> Subject: installer crash on post-instal scripts
>
> 
>
> When I try to install cygwin on windows xp, the installer hangs at around
> 90%, when it's running the X post install scripts. (From a full install)
>

--
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/

installer crash on post-instal scripts

[Jing Zhou]

Hi Kris,

I have the same problem of installing full cygwin 1.5.12.1. It froze
at 93% when installing postinstall in X11 directory. The disk size
continued growing. It was 2G when I eventally decided to delete it.
Then my computer crashed.

It seems to me that you found a solution. Could you please post the solution?

My computer is windows xp home.

thanks,

-Jing

Kris's original message is:

installer crash on post-instal scripts

From: Kris Matthews  
To: cygwin at cygwin dot com 
Date: Wed, 17 Nov 2004 11:51:05 -0500 (EST) 
Subject: installer crash on post-instal scripts 



When I try to install cygwin on windows xp, the installer hangs at around 
90%, when it's running the X post install scripts. (From a full install)

--
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: Can't create Bash here document.

[Hannu E K Nevalainen]

YorkshireSkier BLOB aol.com wrote:
...
> However, if I try to use a command that uses the here document,
> I see the following error:
> 
> $ cat <> 1
>> 2
>> 3
>> fred
> bash: cannot create temp file for here document: No such file or
> directory 
> 
> I can confirm that the variables TEMP and TMP (and TEMPDIR) are
> defined. All point to a valid subdirectory that has write
> permission. I can for instance 'touch $TEMP/fred' without difficulty.
> 
> I found one similar mailing about this in the archive - the
> respondant said this has come up before, but I cannot find the
> original mailing in the archive (or on the web). Please can
> someone point me in the right direction.

What does this display:
$ mount -m

I suspect that you need:
$ mount -fsb "$TEMP" "/tmp"


/Hannu E K Nevalainen, B.Sc. EE Microcomputer systems--72-->

** mailing list preference; please keep replies on 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: scp exits often with -1

[Christopher Faylor]

On Wed, Nov 17, 2004 at 07:20:56PM +0100, Reini Urban wrote:
>or if you like living on the edge there's a new experimental "noparent"
>branch for cygwin1.dll

Sigh.  Do not use this branch.  It is not intended for anyone but me.

--
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/

[OT] RE: scp exits often with -1

[Dave Korn]

> -Original Message-
> From: cygwin-owner On Behalf Of Jörg Schaible
> Sent: 17 November 2004 17:41

> Dave Korn wrote on Wednesday, November 17, 2004 6:24 PM:

> >  *Mine* contained more useful info and was a fraction of the size.
> 
> Well, nonetheless, none of your comments really help with the problem.

  Nor does your strace output.  It does nothing to solve the problem.  It just
sits there unread in several thousand peoples' inboxes.

> So, if you can't or won't help, I prefer no answer.

  You could have gotten _exactly_ that by simply *not* sending 200Kb of strace
to the list.

  Unless anyone specifically asked you for it, you have no reason to assume that
anyone's going to want to plough through it.

  So basically what you've done is say to yourself "I'll just send tons and tons
of stuff (which I know will be of no use or interest to 99.9% of the recipients)
to absolutely everyone in the hopes that it might be relevant to even just one
or two people."

  That's the same approach that spammers take.  It's just plain wrong.

  However we're getting way OT, so if you want to discuss this further, I'm
going to have to ask you to step outside.  While flapping your arms and clucking
like a chicken.  TITTTL!

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: scp exits often with -1

[Reini Urban]

Jörg Schaible schrieb:
I know, that there were massive changes handling fork() and the
pipes. So what shall I do, if a the problem now only manifests
regularly if scp is called by Java calling a dos batch calling a
shell script ? Using strace the problem is going to happen much more
unlikely. So what do you mean is it good using a debugger? I really
tried hard to find a combination, that makes the problem more obvious
without such a weird setup and I cried here for any help, that would
enable me to find a better scenario or that enables anyone else to
find the problem.
Hi Jörg :)
So, if you can't or won't help, I prefer no answer.
Downgrading to cygwin-1.3.11 should help.
or if you like living on the edge there's a new experimental "noparent" 
branch for cygwin1.dll
--
Reini Urban
http://xarch.tu-graz.ac.at/home/rurban/

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

Re: long command executed via a variable fails

[Geoffrey KRETZ]

Actually my original unix script use :
if ! $cmd; then
...
fi
AND
if ! eval $cmd; then
...
fi
And it works perfectly
I'm forced to use eval because I'm of the structure of my $cmd variable.
Just a few samples :
- cmd="$LDPRELOAD $FGLRUN ${mainname}.42r -db $dbname -dt $dbtype -ho 
$QAHOST $dvmversionopt -dc $dbcode -dynamic -tn 99",
with LDPRELOAD = "LD_PRELOAD='$FGLDIR/bin/dbmdb28x.so'" and $FGLRUN is a 
runner use to run an executable
- cmd ="$COMPILATOR fglmksdl -d $dbcode $mksdlopt"
with $COMPILATOR = "INFORMIXC='$FGLCC -shared'", $FGLCC is a compilator 
with some option (ie: gcc -mlp64 or cc -Wl,-G,-dy) and fglmksdl a (my 
company) command that generate shared libraries.

I've got to use carefully the quotes and double-quotes and without eval, 
I doesn't work on Unix.

Whatever, I'll try : eval "$cmd" and read the sh man page ;)
Geo
Igor Pechtchanski wrote:
On Wed, 17 Nov 2004, Reini Urban wrote:
 

Geoffrey KRETZ schrieb:
   

I've got a problem with a shell script used with Cygwin 1.5.10-3 on W2000
SP4 and W XP SP 2.
The following part of code works on all the Unix I've tested (HP-UX/AIX/Sun
Solaris/Linux).
With Cygwin,  it doesn't :(
*Code:*
cmd="long shell command with differents parameters"
if [ ! eval $cmd ]; then
echo "Error : $cmd"
exit 1
fi
 

This doesn't do what you thought it would.  Please read the sh manual on
the "if" builtin command.  You're actually invoking the "[" command with
the parameters "! eval $cmd", which is not what you want.
 

instead of eval $cmd, i've tried :
- `eval $cmd`
- eval `$cmd`
- $cmd
- `$cmd`
 

Most of the above (except $cmd) show that you don't understand much about
command substitution.  Please read the sh man page carefully.
BTW, given that sh's "eval" behaves differently from bash's, you should
actually use 'eval "$cmd"' (note the double quotes).
 

I've also try with a function without more success.
*Code:*
execCmd()
{
eval $cmd
return $?
}
cmd="long shell command with differents parameters"
if [ ! execCmd ]; then
echo "Error : $cmd"
exit 1
fi
 

This doesn't work since $cmd is unset in the function.  Either define the
function after setting cmd, or pass $cmd as a parameter, e.g.,
execCmd() {
 cmd="$1"; shift
 eval "$cmd"
 return $?
}
BTW, your "if" test suffers from the same problem as your original one.
Ditch the "[".."]".
 

The only way I've find is :
*Code:*
cmd="long shell command with differents parameters"
eval $cmd
if [ $? -ne 0 ]; then
echo "Error : $cmd"
exit 1
fi
 

Note that this works because you're actually using "[".."]" in the right
way here.
 

Is it possible to make it work like the two first exemple or I'm
obliged to use the third solution ?
 

FAQ: http://cygwin.com/faq/faq_3.html#SEC43
cygwin's /bin/sh is ash, on most other platforms it is /bin/bash.
If you want it to behave it exactly like on other platforms, and you use
bash specific constructs, use the /bin/bash shebang.
   

Well, eval is not a bash-specific construct.  It does behave slightly
differently in ash than in bash, but in this particular case, switching to
bash wouldn't have helped the OP.  Reading the man pages for bash and sh
would have. :-)
HTH,
	Igor
 


--
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: long command executed via a variable fails

[Reini Urban]

Geoffrey KRETZ schrieb:
Reini Urban wrote:
FAQ: http://cygwin.com/faq/faq_3.html#SEC43
cygwin's /bin/sh is ash, on most other platforms it is /bin/bash.
If you want it to behave it exactly like on other platforms, and you use
bash specific constructs, use the /bin/bash shebang.
If I specified  "#! /bin/bash" at the first line of my shell script, is 
it enough to make Cygwin understand that I want to use bash instead of 
ash ??
sure. you'd need to do that on every platform if you use bash specifics. 
it is considered good style.

but as igor already posted, your problem is only marginally better on 
bash. you should really use exact quoting: test 'eval "$cmd"'
or even better use Igor suggestions with testing against the return value.
--
Reini Urban
http://xarch.tu-graz.ac.at/home/rurban/

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

RE: scp exits often with -1

[Jörg Schaible]

Dave Korn wrote on Wednesday, November 17, 2004 6:24 PM:

>   I beg to differ.
> 
>   My post contained the useful information that dumping a
> couple of hundred Kb of unsolicited strace output on the list
> is a pointless waste of time.
> 
>   Yours contained a couple of hundred Kb of unsolicited strace output.
> 
> http://www.google.com/search?q=unsolicited+strace+site:cygwin.com&hl=en&lr=&filter=0
>
>  *Mine* contained more useful info and was a fraction of the size.

Well, nonetheless, none of your comments really help with the problem.

I know, that there were massive changes handling fork() and the pipes. So what 
shall I do, if a the problem now only manifests regularly if scp is called by 
Java calling a dos batch calling a shell script ? Using strace the problem is 
going to happen much more unlikely. So what do you mean is it good using a 
debugger?

I really tried hard to find a combination, that makes the problem more obvious 
without such a weird setup and I cried here for any help, that would enable me 
to find a better scenario or that enables anyone else to find the problem.

>  I rest my case.

So, if you can't or won't help, I prefer no answer.

- Jörg

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

Re: Problem with setup

[Bobby McNulty]

Max Bowsher wrote:
[EMAIL PROTECTED] wrote:
On 11 Nov, Bobby McNulty wrote:
 Guys, theres a major problem with setup.
 According to the error I just got, setup is not
 closing any of its files after writing them.
 I finally got past xorg, and got up to update_info.
 I'll rerun setup to get a cygcheck.

I can confirm this.  We took a snapshot of October 26th
(setup-timestamp: 1098732614), and in that version of setup,

That number you quoted identifies the version of the setup.ini 
datafile, not the setup.exe program.

none of
the file handles for any of the temp files of the form (where c:\temp
is your temporary package directory) c:\temp\http%.\package.tar.bz2
is ever released.  The system starts page faulting and the number of
handles just grows and grows, and eventually setup dies like this:
Microsoft Visual C++ Runtime Error
Runtime Error!
Program \\samba\install\win32\cygwin\setup.exe
This application has requested the Runtime to terminate it in an
unusual way.  Please contact the application's support team for
more information.

That's nasty. I wonder why we haven't seen more reports of this.
Is there any reason why your machine should be more susceptible to 
this? (Low memory, any kind of resource limits?)

Max.
--
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/

As for for, 256 MB on Celeron II running XP Pro with Service pack 2.
Running SP2 since it came out.
Problem started November 3rd with a full installation.
--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Re: Problem with setup

[Bobby McNulty]

Luke Kendall wrote:
On 16 Nov, Bobby McNulty wrote:
 

Then I'm not the only one noticing this? 
   

No, myself and one other person here have begun noticing it.
luke
--
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/
 

OK.
I have a full cywin running now, with Xorg and KDE.
So I'm ok.
But I had to do it in four stages.
KDE is working again under org.
Like I said before, it was setup.
Bobby
--
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: long command executed via a variable fails

[Geoffrey KRETZ]

Reini Urban wrote:
Geoffrey KRETZ schrieb:
I've got a problem with a shell script used with Cygwin 1.5.10-3 on 
W2000 SP4 and W XP SP 2.

The following part of code works on all the Unix I've tested 
(HP-UX/AIX/Sun Solaris/Linux).

With Cygwin,  it doesn't :(
*Code:*
cmd="long shell command with differents parameters"
if [ ! eval $cmd ]; then
 echo "Error : $cmd"
 exit 1
fi

instead of eval $cmd, i've tried :
- `eval $cmd`
- eval `$cmd`
- $cmd
- `$cmd`
I've also try with a function without more success.
*Code:*
execCmd()
{
 eval $cmd
 return $?
}
cmd="long shell command with differents parameters"
if [ ! execCmd ]; then
 echo "Error : $cmd"
 exit 1
fi

The only way I've find is :
*Code:*
cmd="long shell command with differents parameters"
eval $cmd
if [ $? -ne 0 ]; then
 echo "Error : $cmd"
 exit 1
fi
Is it possible to make it work like the two first exemple or I'm 
obliged to use the third solution ?

FAQ: http://cygwin.com/faq/faq_3.html#SEC43
cygwin's /bin/sh is ash, on most other platforms it is /bin/bash.
If you want it to behave it exactly like on other platforms, and you use
bash specific constructs, use the /bin/bash shebang.
If I specified  "#! /bin/bash" at the first line of my shell script, is 
it enough to make Cygwin understand that I want to use bash instead of 
ash ??

--
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: long command executed via a variable fails

[Igor Pechtchanski]

On Wed, 17 Nov 2004, Reini Urban wrote:

> Geoffrey KRETZ schrieb:
> > I've got a problem with a shell script used with Cygwin 1.5.10-3 on W2000
> > SP4 and W XP SP 2.
> >
> > The following part of code works on all the Unix I've tested (HP-UX/AIX/Sun
> > Solaris/Linux).
> >
> > With Cygwin,  it doesn't :(
> >
> > *Code:*
> >
> > cmd="long shell command with differents parameters"
> > if [ ! eval $cmd ]; then
> >  echo "Error : $cmd"
> >  exit 1
> > fi

This doesn't do what you thought it would.  Please read the sh manual on
the "if" builtin command.  You're actually invoking the "[" command with
the parameters "! eval $cmd", which is not what you want.

> > instead of eval $cmd, i've tried :
> > - `eval $cmd`
> > - eval `$cmd`
> > - $cmd
> > - `$cmd`

Most of the above (except $cmd) show that you don't understand much about
command substitution.  Please read the sh man page carefully.

BTW, given that sh's "eval" behaves differently from bash's, you should
actually use 'eval "$cmd"' (note the double quotes).

> > I've also try with a function without more success.
> >
> > *Code:*
> >
> > execCmd()
> > {
> >  eval $cmd
> >  return $?
> > }
> >
> > cmd="long shell command with differents parameters"
> > if [ ! execCmd ]; then
> >  echo "Error : $cmd"
> >  exit 1
> > fi

This doesn't work since $cmd is unset in the function.  Either define the
function after setting cmd, or pass $cmd as a parameter, e.g.,

execCmd() {
  cmd="$1"; shift
  eval "$cmd"
  return $?
}

BTW, your "if" test suffers from the same problem as your original one.
Ditch the "[".."]".

> > The only way I've find is :
> >
> > *Code:*
> >
> > cmd="long shell command with differents parameters"
> > eval $cmd
> > if [ $? -ne 0 ]; then
> >  echo "Error : $cmd"
> >  exit 1
> > fi

Note that this works because you're actually using "[".."]" in the right
way here.

> > Is it possible to make it work like the two first exemple or I'm
> > obliged to use the third solution ?
>
> FAQ: http://cygwin.com/faq/faq_3.html#SEC43
> cygwin's /bin/sh is ash, on most other platforms it is /bin/bash.
>
> If you want it to behave it exactly like on other platforms, and you use
> bash specific constructs, use the /bin/bash shebang.

Well, eval is not a bash-specific construct.  It does behave slightly
differently in ash than in bash, but in this particular case, switching to
bash wouldn't have helped the OP.  Reading the man pages for bash and sh
would have. :-)
HTH,
Igor
-- 
http://cs.nyu.edu/~pechtcha/
  |\  _,,,---,,_[EMAIL PROTECTED]
ZZZzz /,`.-'`'-.  ;-;;,_[EMAIL PROTECTED]
 |,4-  ) )-,_. ,\ (  `'-'   Igor Pechtchanski, Ph.D.
'---''(_/--'  `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-.  Meow!

"The Sun will pass between the Earth and the Moon tonight for a total
Lunar eclipse..." -- WCBS Radio Newsbrief, Oct 27 2004, 12:01 pm EDT

--
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: scp exits often with -1

[Dave Korn]

> -Original Message-
> From: Jörg Schaible  
> Sent: 17 November 2004 17:22

> Dave Korn wrote on Wednesday, November 17, 2004 6:18 PM:
> 
> >> -Original Message-
> >> From: cygwin-owner On Behalf Of Jörg Schaible
> >> Sent: 16 November 2004 17:39
> > 
> >   Sure.  How about 200k of unintelligible gibberish?  I'm
> > sure we'd all _love_ to see it.
> 
> At least it contained more useful information.
> 
> - Jörg

  I beg to differ.

  My post contained the useful information that dumping a couple of hundred Kb
of unsolicited strace output on the list is a pointless waste of time.

  Yours contained a couple of hundred Kb of unsolicited strace output.

http://www.google.com/search?q=unsolicited+strace+site:cygwin.com&hl=en&lr=&filt
er=0

  *Mine* contained more useful info and was a fraction of the size.
 
  I rest my case.


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: scp exits often with -1

[Jörg Schaible]

Dave Korn wrote on Wednesday, November 17, 2004 6:18 PM:

>> -Original Message-
>> From: cygwin-owner On Behalf Of Jörg Schaible
>> Sent: 16 November 2004 17:39
> 
>   Sure.  How about 200k of unintelligible gibberish?  I'm
> sure we'd all _love_ to see it.

At least it contained more useful information.

- Jörg

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

RE: scp exits often with -1

[Dave Korn]

> -Original Message-
> From: cygwin-owner On Behalf Of Jörg Schaible
> Sent: 16 November 2004 17:39

> Hi folks,
> 
> after upgrading to latest cygwin1.dll and openssh I have 
> massive problems with ssh/scp. Using e.g. the same scp 
> command it fails from time to time called from the command 
> line. Unfortunately it fails more often when called from a 
> shell script and quite every time when this shell script is 
> called from a DOS batch called by a Java app.
> 
> See the output of scp -v -v in the good version:

[snip]

> The singel difference is, that in the successful case the 
> client sends an exit-state, that is not received in the 
> unseccussful version. The problem manifests more often with 
> using a shell script and a DOS batch:

[snip]

> Worst case is calling the scpwin.bat from within Java (app 
> also started from the bash shell) using 
> Runtime.exec(Object[]). Then it fails nearly every time.
> 
> Any other information I can provide ?
> 
> - Jörg


  Sure.  How about 200k of unintelligible gibberish?  I'm sure we'd all _love_
to see it.



..Woa!


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: new cygwin has memory problems?

[Christopher Faylor]

On Fri, Nov 12, 2004 at 08:26:05AM -0800, Lester Ingber wrote:
>I was wrong in assuming this was a cygwin base problem.  It must be a
>cygwin-xfree problem.  I have no problems I can see in a cygwin console
>window, only in an xterm (or rxvt) window running under
>XWin -unixkill -emulate3buttons -clipboard -multiwindow -silent-dup-error &
>
>Lester
>
>  +=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=
>I cannot determine why, but with the new Cygwin base I just installed,
>a utility I have used for quite a few years, dired
>[http://www.math.utah.edu/~beebe/software/file-tools.html#dired], now
>freezes my window.  A `top` shows that CPU usage quickly climbs to
>maximum values.
>
>I have to guess that other utilities will have similar problems.

This is a problem caused by this change:

http://cygwin.com/ml/cygwin/2004-10/msg00364.html

And, then, the problem was noted here:

http://cygwin.com/ml/cygwin/2004-10/msg00904.html

The problem is that cygwin is now behaving like linux in that timed
reads from a pty/tty work correctly.  It is not behaving like linux in
that once you get an EOF on a read using stdio, you get EOF forever
after.

I've made a cygwin-specific correction to the newlib library to make
cygwin work more like linux.  I will generate a cygwin snapshot sometime
soon which should have a fix for this problem.  Keep an ey on:

http://cygwin.com/snapshots/

cgf

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

Re: installer crash on post-instal scripts

[Kris Matthews]

Crap crap crap crap.

My apologies, I *started* writing this email, then found a solution to the 
problem, then went back in to _delete_ this message, and sent it instead.

Sheesh :)

Kristopher
[EMAIL PROTECTED]

On Wed, 17 Nov 2004, Kris Matthews wrote:

>
> When I try to install cygwin on windows xp, the installer hangs at around
> 90%, when it's running the X post install scripts. (From a full install)
>
> My logs are at:
>
>
> It seems to be related to:
> http://www.mail-archive.com/cygwin-xfree@cygwin.com/msg12903.html
>
>

--
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/

installer crash on post-instal scripts

[Kris Matthews]

When I try to install cygwin on windows xp, the installer hangs at around 
90%, when it's running the X post install scripts. (From a full install)

My logs are at:


It seems to be related to:
http://www.mail-archive.com/cygwin-xfree@cygwin.com/msg12903.html


--
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/

cygwin1 dll and JAVA-JNI dll loading is buggy

[Gilles Bourgeois]

Hello
Next year (using the current version of cygwin which I can not remember), I 
managed to load from java (JNI) a dll (build against cygwin1.dll) compiled with 
gcc under cygwin.
Now , I have updated my cygwin (1.5.9 for me tooday), and it does not work any 
more.
I saw on the cygwin mailing list that there is problems with DLL and JAVA under 
cygwin.
is it ok know with the latest versions of cygwin1.dll ?
TIA

Gilles
--

Faites un voeu et puis Voila ! www.voila.fr 



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

Re: g++ 3.4.1

[Sam Steingold]

> * Reini Urban <[EMAIL PROTECTED]> [2004-11-17 17:19:33 +0100]:
>
> Sam Steingold schrieb:
>>>* Igor Pechtchanski <[EMAIL PROTECTED]> [2004-11-17 10:23:49 -0500]:
when will 3.4 become the default version?
>>>Sorry, can't answer that.
>> who can?
>
> gerrit (the maintainer) can.

Gerrit, could you please speak up?

> but there are yet unresolved issues with the handling of dwarf2 or sjsl
> (longjmp) exception handling on cygwin.
> dwarf2 is much faster but doesn't work with win32 callbacks.
> some folks use locally dwarf2 patched 3.4 versions.
> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14563

Known to work:  3.2
Known to fail:  3.3.3   3.4.2

I don't see how holding 3.4 back is helping this issue.
(I am sure you can enlighten me!)

-- 
Sam Steingold (http://www.podval.org/~sds) running w2k
  
 
nobody's life, liberty or property are safe while the legislature is in session

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

Re: g++ 3.4.1

[Reini Urban]

Sam Steingold schrieb:
* Igor Pechtchanski <[EMAIL PROTECTED]> [2004-11-17 10:23:49 -0500]:
Secondly, gcc-3.4.1 is an experimental package, so you should select
the "Exp" radio button on the "Select Packages" screen to get it.
WFM.

when will 3.4 become the default version?
Sorry, can't answer that.
who can?
gerrit (the maintainer) can.
but there are yet unresolved issues with the handling of dwarf2 or sjsl 
(longjmp) exception handling on cygwin.
dwarf2 is much faster but doesn't work with win32 callbacks.
some folks use locally dwarf2 patched 3.4 versions.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14563
--
Reini Urban
http://xarch.tu-graz.ac.at/home/rurban/
--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Re: long command executed via a variable fails

[Reini Urban]

Geoffrey KRETZ schrieb:
I've got a problem with a shell script used with Cygwin 1.5.10-3 on 
W2000 SP4 and W XP SP 2.

The following part of code works on all the Unix I've tested 
(HP-UX/AIX/Sun Solaris/Linux).

With Cygwin,  it doesn't :(
*Code:*
cmd="long shell command with differents parameters"
if [ ! eval $cmd ]; then
 echo "Error : $cmd"
 exit 1
fi

instead of eval $cmd, i've tried :
- `eval $cmd`
- eval `$cmd`
- $cmd
- `$cmd`
I've also try with a function without more success.
*Code:*
execCmd()
{
 eval $cmd
 return $?
}
cmd="long shell command with differents parameters"
if [ ! execCmd ]; then
 echo "Error : $cmd"
 exit 1
fi

The only way I've find is :
*Code:*
cmd="long shell command with differents parameters"
eval $cmd
if [ $? -ne 0 ]; then
 echo "Error : $cmd"
 exit 1
fi
Is it possible to make it work like the two first exemple or I'm obliged 
to use the third solution ?
FAQ: http://cygwin.com/faq/faq_3.html#SEC43
cygwin's /bin/sh is ash, on most other platforms it is /bin/bash.
If you want it to behave it exactly like on other platforms, and you use
bash specific constructs, use the /bin/bash shebang.
--
Reini Urban
http://xarch.tu-graz.ac.at/home/rurban/
--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Re: Clapack on Cygwin

[Lionel B]

Eleftheria Kavoussanaki writes:

> Hello!
> 
> I tried to install clapack on cygwin and errors occurred. I would very much
> appreciate your help.

/.../

> How can I make the process of timing the clapack routines run in a
> reasonably finite amount of time?

Not sure about clapack, but I have successfully compiled ATLAS
(http://math-atlas.sourceforge.net/) - which includes a subset of the LAPACK
routines and supplies a C interface - under Cygwin. Building the libraries was
pretty much a no-brainer and the routines run like blazes.

Oh, and it took about a day to build (with timing calculations) on a Quad 2.8GHz
Xeon with 1Gb RAM... so patience may be required.

-- 
Lionel B






--
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 with setup

[Max Bowsher]

Luke Kendall wrote:
On 17 Nov, To: [EMAIL PROTECTED] wrote:
 The system starts page faulting and the number of
 handles just grows and grows, and eventually setup dies
I can add this:
At the crash point, setup.exe has some 600 open handles and 2 Gb of VM. 
It's
accumulated 2.5 million page faults.
Wow!
Well, it's not quite as dramatic here, but I can see a good few leaked 
handles in Process Explorer.

I've committed a fix, but unfortunately, when I begun testing to prepare for 
a release, I ran into all sorts of nasty problems, including being unable to 
reinstall Cygwin, having uninstalled it to test setup. I'm going to keep 
working on it though... after I've completely purged Cygwin from my 
computer, reinstalled, and set up a VM to do tests in without disrupting my 
primary machine. Might take a while.

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

Re: g++ 3.4.1

[Sam Steingold]

> * Christopher Faylor <[EMAIL PROTECTED]> [2004-11-17 10:24:21 -0500]:
>
> On Wed, Nov 17, 2004 at 09:58:47AM -0500, Sam Steingold wrote:
>>when I use setup to upgrade, I still get gcc 3.3.3.
>>when I try clicking "New" to get 3.4.1, "gcc-core" keeps jumping to
>>"Keep".
>>
>>when will 3.4 become the default version?
>
> Sometime within the next year.

good, thanks.
who makes the decision?
what is the decision based on?

I am so used to cygwin being more up-to-date than Fedora that I am
surprised that FC3 comes with gcc 3.4 and Cygwin does not.

-- 
Sam Steingold (http://www.podval.org/~sds) running w2k
  
 
Lisp: its not just for geniuses anymore.

--
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/

long command executed via a variable fails

[Geoffrey KRETZ]

Hello,
I've got a problem with a shell script used with Cygwin 1.5.10-3 on 
W2000 SP4 and W XP SP 2.

The following part of code works on all the Unix I've tested 
(HP-UX/AIX/Sun Solaris/Linux).

With Cygwin,  it doesn't :(
*Code:*
cmd="long shell command with differents parameters"
if [ ! eval $cmd ]; then
 echo "Error : $cmd"
 exit 1
fi

instead of eval $cmd, i've tried :
- `eval $cmd`
- eval `$cmd`
- $cmd
- `$cmd`
I've also try with a function without more success.
*Code:*
execCmd()
{
 eval $cmd
 return $?
}
cmd="long shell command with differents parameters"
if [ ! execCmd ]; then
 echo "Error : $cmd"
 exit 1
fi

The only way I've find is :
*Code:*
cmd="long shell command with differents parameters"
eval $cmd
if [ $? -ne 0 ]; then
 echo "Error : $cmd"
 exit 1
fi
Is it possible to make it work like the two first exemple or I'm obliged 
to use the third solution ?

Thanks in advance,
Geoffrey

--
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/

Can't create Bash here document.

[YorkshireSkier]

Hi,

I've installed Bash on several Win2k workstations. Most are
working fine, but one particular machine is giving trouble. 
I am attempting to install only the base package - no compilers,
editors, X11 or anything else. The setup program appears to run 
correctly. After the installation 'cygcheck -c' shows all OK.
However, if I try to use a command that uses the here document,
I see the following error:

$ cat < 1
> 2
> 3
> fred
bash: cannot create temp file for here document: No such file or directory

I can confirm that the variables TEMP and TMP (and TEMPDIR) are
defined. All point to a valid subdirectory that has write permission. I can for 
instance 'touch $TEMP/fred' without difficulty.

I found one similar mailing about this in the archive - the
respondant said this has come up before, but I cannot find the original mailing 
in the archive (or on the web). Please can someone point me in the right 
direction.

Thanks,

Simon

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

Re: g++ 3.4.1

[Sam Steingold]

> * Igor Pechtchanski <[EMAIL PROTECTED]> [2004-11-17 10:23:49 -0500]:
>
> On Wed, 17 Nov 2004, Sam Steingold wrote:
>
>> when I use setup to upgrade, I still get gcc 3.3.3.
>> when I try clicking "New" to get 3.4.1, "gcc-core" keeps jumping to
>> "Keep".
>
> Umm, first off, where did you find "New" in setup?

select packages window, column "New".


> Secondly, gcc-3.4.1 is an experimental package, so you should select
> the "Exp" radio button on the "Select Packages" screen to get it.
> WFM.

thanks!

>> when will 3.4 become the default version?
> Sorry, can't answer that.

who can?

-- 
Sam Steingold (http://www.podval.org/~sds) running w2k
  
 
The only intuitive interface is the nipple.  The rest has to be learned.

--
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: chown

[Igor Pechtchanski]

On Wed, 17 Nov 2004, George wrote:

> From the "Recovering from Windows penchant for spaces in user and file
> names" Department:
>
> Maybe I'm being really dense, but the --from switch isn't quite
> working as advertised:
>
> [EMAIL PROTECTED] ~/test
> $ ls -ln
> -rwxrwxrwx1 544  51348 Nov 17 03:59 testfile
>
> [EMAIL PROTECTED] ~/test
> $ chown --from=514:513 1008:513 testfile
>
> [EMAIL PROTECTED] ~/test
> $ ls -ln
> -rwxrwxrwx1 544  51348 Nov 17 03:59 testfile
>
> Thanks.

What CGF said...  However, as a datapoint, chown's --from switch works
just fine for me, as long as I specify the current owner and group
correctly... :-)  I.e., in your case, you should have said

$ chown --from=544:513 1008:513 testfile
   ^^^
Hope this helps,
Igor
-- 
http://cs.nyu.edu/~pechtcha/
  |\  _,,,---,,_[EMAIL PROTECTED]
ZZZzz /,`.-'`'-.  ;-;;,_[EMAIL PROTECTED]
 |,4-  ) )-,_. ,\ (  `'-'   Igor Pechtchanski, Ph.D.
'---''(_/--'  `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-.  Meow!

"The Sun will pass between the Earth and the Moon tonight for a total
Lunar eclipse..." -- WCBS Radio Newsbrief, Oct 27 2004, 12:01 pm EDT

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

Re: g++ 3.4.1

[Igor Pechtchanski]

On Wed, 17 Nov 2004, Sam Steingold wrote:

> when I use setup to upgrade, I still get gcc 3.3.3.
> when I try clicking "New" to get 3.4.1, "gcc-core" keeps jumping to
> "Keep".

Umm, first off, where did you find "New" in setup?  Secondly, gcc-3.4.1 is
an experimental package, so you should select the "Exp" radio button on
the "Select Packages" screen to get it.  WFM.

> when will 3.4 become the default version?

Sorry, can't answer that.
HTH,
Igor
-- 
http://cs.nyu.edu/~pechtcha/
  |\  _,,,---,,_[EMAIL PROTECTED]
ZZZzz /,`.-'`'-.  ;-;;,_[EMAIL PROTECTED]
 |,4-  ) )-,_. ,\ (  `'-'   Igor Pechtchanski, Ph.D.
'---''(_/--'  `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-.  Meow!

"The Sun will pass between the Earth and the Moon tonight for a total
Lunar eclipse..." -- WCBS Radio Newsbrief, Oct 27 2004, 12:01 pm EDT

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

Re: g++ 3.4.1

[Christopher Faylor]

On Wed, Nov 17, 2004 at 09:58:47AM -0500, Sam Steingold wrote:
>when I use setup to upgrade, I still get gcc 3.3.3.
>when I try clicking "New" to get 3.4.1, "gcc-core" keeps jumping to
>"Keep".
>
>when will 3.4 become the default version?

Sometime within the next year.

--
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: chown

[Christopher Faylor]

On Wed, Nov 17, 2004 at 04:25:30AM -0800, George wrote:
>From the "Recovering from Windows penchant for spaces in user and file
>names" Department:
>
>Maybe I'm being really dense, but the --from switch isn't quite 
>working as advertised:
>
>[EMAIL PROTECTED] ~/test
>$ ls -ln
>-rwxrwxrwx1 544  51348 Nov 17 03:59 testfile
>
>[EMAIL PROTECTED] ~/test
>$ chown --from=514:513 1008:513 testfile
>
>[EMAIL PROTECTED] ~/test
>$ ls -ln
>-rwxrwxrwx1 544  51348 Nov 17 03:59 testfile
>
>Thanks.

http://cygwin.com/problems.html would probably be instructive.

--
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: a newbie's question about directory

[Igor Pechtchanski]

On Wed, 17 Nov 2004, Marcel Telka wrote:

> On Wed, Nov 17, 2004 at 04:27:47PM +0800, ZF Tang wrote:
> > Hi,all
> > I'm a newbie to cygwin/linux/unix. Can somebody tell me what's the
> > difference between the directories /bin, /usr/bin, and /usr/local/bin ?
>
> http://www.pathname.com/fhs/pub/fhs-2.3.html

With one teeny Cygwin-specific correction:
.
Igor
-- 
http://cs.nyu.edu/~pechtcha/
  |\  _,,,---,,_[EMAIL PROTECTED]
ZZZzz /,`.-'`'-.  ;-;;,_[EMAIL PROTECTED]
 |,4-  ) )-,_. ,\ (  `'-'   Igor Pechtchanski, Ph.D.
'---''(_/--'  `-'\_) fL a.k.a JaguaR-R-R-r-r-r-.-.-.  Meow!

"The Sun will pass between the Earth and the Moon tonight for a total
Lunar eclipse..." -- WCBS Radio Newsbrief, Oct 27 2004, 12:01 pm EDT

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

Re: g++ 3.4.1

[Sam Steingold]

when I use setup to upgrade, I still get gcc 3.3.3.
when I try clicking "New" to get 3.4.1, "gcc-core" keeps jumping to
"Keep".

when will 3.4 become the default version?

-- 
Sam Steingold (http://www.podval.org/~sds) running w2k
  
 
Microsoft: announce yesterday, code today, think tomorrow.

--
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 with setup

[Max Bowsher]

[EMAIL PROTECTED] wrote:
On 11 Nov, Bobby McNulty wrote:
 Guys, theres a major problem with setup.
 According to the error I just got, setup is not
 closing any of its files after writing them.
 I finally got past xorg, and got up to update_info.
 I'll rerun setup to get a cygcheck.
I can confirm this.  We took a snapshot of October 26th
(setup-timestamp: 1098732614), and in that version of setup,
That number you quoted identifies the version of the setup.ini datafile, not 
the setup.exe program.

none of
the file handles for any of the temp files of the form (where c:\temp
is your temporary package directory) c:\temp\http%.\package.tar.bz2
is ever released.  The system starts page faulting and the number of
handles just grows and grows, and eventually setup dies like this:
Microsoft Visual C++ Runtime Error
Runtime Error!
Program \\samba\install\win32\cygwin\setup.exe
This application has requested the Runtime to terminate it in an
unusual way.  Please contact the application's support team for
more information.

That's nasty. I wonder why we haven't seen more reports of this.
Is there any reason why your machine should be more susceptible to this? 
(Low memory, any kind of resource limits?)

Max.
--
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 with setup

[Max Bowsher]

Luke Kendall wrote:
I should also add that after a setup.exe crash at 9:38am after being
left running overnight, later attempts to install that day (with the VM
size steadily increasing, and setup.exe appearing to "stick" at the
stage where it is installing the zip package), that setup.log was not
updated by the later installation attempts.
Shouldn't setup.log be getting opened in append mode?  We examined it
at various times during a later installation attempt, and the contents
didn't change and the file modification time stayed at 9:38am.
At the moment, the log files are written to disc in one bulk write when 
setup.exe completes. That's probably not a great design feature.

Max.
--
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/

Clapack on Cygwin

[Eleftheria Kavoussanaki]

Hello!

I tried to install clapack on cygwin and errors occurred. I would very much
appreciate your help.

In particular:

I downloaded clapack.tgz from http://www.netlib.org/clapack/, unpacked it
with WinRAR and followed steps 1-5 of the clapack readme.install
successfully. I used the make.inc.LINUX instead of the default make.inc
(after renaming it to make.inc). I did not
change the platform (PLAT in make.inc) specification to _Cygwin. I left it
as it was: _Linux. On step 6 of the readme.install instructions I did:

cd clapack/testing/matgen and then make, which worked fine. Then I did
cd clapack/testing and then make. Things seemed to proceed without errors
until I got the following message:

NEP: Testing Non Symmetric Eigenvalue Problem Routines
./xeigtsts < nep.in > snep.out 2 > & 1
signal 11
make: *** [snep.out]
Error 139

The process underway stopped. I inspected all the *.out files that I could
find. The stest.out was without errors (I suppose this was referring to the
first step). The snep.out was empty which could be because of this
interruption due to that error that occurred.

I tried to reproduce that problem and everytime I tried I was getting a new
error message until I got no errors at all...it seemed as if the error was
going down from snep.out to dnep.out to cnep.out until it could no longer
find any *nep.out files to affect...

So, I thought step 6 is past me and continued with step 7. I went into the
clapack/timing directory and typed make. Things seemed to run forever. Thus,
./xlintimc < ctime.in > ctime.out 2> &1 (what I am attempting now) is up on
the screen for more than 20 min...Initially, I had a problem with ./xlintims
< stime.in > stime.out 2>&1 but somehow re-running the make command solved
(?) the problem (as before). So, now I got stuck with the complex routines
and if I overcome this I suspect I will get an error with the double
precision routines...

As I was searching for clues over the internet, I found a suggestion which
was saying that if things run forever with the timing of the clapack
routines, this could be due to a faulty sprintf command. I added the
recommended line- cure into the appropriate file but things did not seem to
speed up.

How can I make the process of timing the clapack routines run in a
reasonably finite amount of time?

(I am running all the make commands after I do startx on bash)

I hope that my way of posing the problem is descriptive enough..:)

I would be grateful of your help.

Sincerely,

kele


--
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/

chown

[George]

>From the "Recovering from Windows penchant for spaces in user and file
names" Department:

Maybe I'm being really dense, but the --from switch isn't quite 
working as advertised:

[EMAIL PROTECTED] ~/test
$ ls -ln
-rwxrwxrwx1 544  51348 Nov 17 03:59 testfile

[EMAIL PROTECTED] ~/test
$ chown --from=514:513 1008:513 testfile

[EMAIL PROTECTED] ~/test
$ ls -ln
-rwxrwxrwx1 544  51348 Nov 17 03:59 testfile

Thanks.

-- 
George


--
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: Oracle, ProC and cygwin - troubles

[Bødker . Martin Hendersen MHB]

Thanks for your reply.

We have now modified the registry with the keys, as described by Gerrit, to 
increase the heap size. The result is the same.

I can give some more details that we have encountered. It is possible to run 
the program in gdb, but only if we compile and link with oraSQL10.lib:
gcc -g -DNOTIMEOUT -DORACLE -DNO_ACS -o sample1 sample1.c 
cygdrive/c/oracle/product/10.1.0/Db_1/precomp/LIB/oraSQL10.lib

But when we run the program from the command line it gives this output, and we 
are sure there is only 1 cygwin1.dll:

c:\cygwin\home\sample1.exe (1096): *** cygheap version mismatch detected .
You have multiple copies of cygwin1.dll on your system.
Search for cygwin1.dll using the Windows Start->Find/Search facility
and delete all but the most recent version.  The most recent version *should*
reside in x:\cygwin\bin, where 'x' is the drive on which you have
installed the cygwin distribution.

If we link against the two oracle dll's as described earlier, we got this error 
as described earlier: 
6 [main] ? 2884 cygheap_fixup_in_child: Couldn't reserve 9891528 bytes of
space for cygwin's heap (0x6178 <0xB4>) in child, Win32 error 487
c:\cygwin\home\z6mhb\sample1.exe (2884): *** m.AllocationBase 0x0, m.BaseAddress
 0x6178, m.RegionSize 0x4, m.State 0x1

Kind regards,
Martin




___
www.kmd.dk   www.kundenet.kmd.dk   www.eboks.dk   www.civitas.dk   
www.netborger.dk

Hvis du har modtaget denne mail ved en fejl vil jeg gerne, at du informerer mig 
og sletter den.
KMD skaber it-services, der fremmer effektivitet hos det offentlige, 
erhvervslivet og borgerne.

If you received this e-mail by mistake, please notify me and delete it. Thank 
you.
Our mission is to enhance the efficiency of the public sector and improve its 
service of the general public. 


--
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: Sound generator starting package for Linux/Cygwin

[Reini Urban]

Theo Verelst schrieb:
Reini Urban wrote:
Theo Verelst schrieb:
I've started a sound generator package on Linux and Cygwin consisting
...
   http://82.168.209.239/Soundtest
Why not use the standard puredata, which is also based on tcl/tk and 
portaudio, but is stable and mature?
Well, as quick response, to begin with I didn't know about that package.
Second, having quickly looked at it, it looks like the package I've seen
at IRCAM, which frankly sounded not so good at all.
??
this must have been an old version. pd is imho simply the best, and the 
facto standard.

The approach of readable messages to control a sound core is hardly new, 
I've used it decades ago in other context such as graphics (e.g. AVS), 
and I liked (and still like) the approach in combination with a good 
interpreted language, for which this example is meant.
we'll see.
pd uses some kind of "graphical" interpreted language, which connects 
dlopen'ed compiled objects. (highly optimized)

I forward this to pd-dev @ iem.at, which is right around my corner.
Pd probably stems back to NeXT, which as first desktop had a builtin DSP
for sound processing. I'm interested in DSP, too, (see 
http://82.168.209.239/Dsp , http://82.168.209.239/Xilinx ), I'm sure
it's quite an issue to do sound algorithms right 
(http://mini.net/tcl/11991)...
yep, that's true.
pd is the open source version for ircam's MAX, which ran on NeXT and 
irix and Mac.

I'll check Pd, but am quite sure I don't want to go that way myself.

Oh, the idea was also to try out an approach which could make scripting 
and C core work on Linux and cygwin, preferably with as little porting
effort as possible, and also in a distributed way: more than one sound 
cores on different machines.

And I've made a (very complicated) sound generator core ( 
http://82.168.209.239/Articles/pms.html ) for string simulation
(presented at FOSDEM last year http://82.168.209.239/Fosdem ) which I 
want to be able to control not just over midi in real time, and though
I'd experiment a bit with a practical frame for that.
--
Reini Urban
http://xarch.tu-graz.ac.at/home/rurban/
--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

Re: Sound generator starting package for Linux/Cygwin

[Theo Verelst]

Reini Urban wrote:
Theo Verelst schrieb:
I've started a sound generator package on Linux and Cygwin consisting
...
   http://82.168.209.239/Soundtest
Why not use the standard puredata, which is also based on tcl/tk and 
portaudio, but is stable and mature?
Well, as quick response, to begin with I didn't know about that package.
Second, having quickly looked at it, it looks like the package I've seen
at IRCAM, which frankly sounded not so good at all.
The approach of readable messages to control a sound core is hardly new, 
I've used it decades ago in other context such as graphics (e.g. AVS), 
and I liked (and still like) the approach in combination with a good 
interpreted language, for which this example is meant.

Pd probably stems back to NeXT, which as first desktop had a builtin DSP
for sound processing. I'm interested in DSP, too, (see 
http://82.168.209.239/Dsp , http://82.168.209.239/Xilinx ), I'm sure
it's quite an issue to do sound algorithms right 
(http://mini.net/tcl/11991)...

I'll check Pd, but am quite sure I don't want to go that way myself.
Oh, the idea was also to try out an approach which could make scripting 
and C core work on Linux and cygwin, preferably with as little porting
effort as possible, and also in a distributed way: more than one sound 
cores on different machines.

And I've made a (very complicated) sound generator core ( 
http://82.168.209.239/Articles/pms.html ) for string simulation
(presented at FOSDEM last year http://82.168.209.239/Fosdem ) which I 
want to be able to control not just over midi in real time, and though
I'd experiment a bit with a practical frame for that.

Regards,
   Theo Verelst
--
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: a newbie's question about directory

[Marcel Telka]

On Wed, Nov 17, 2004 at 04:27:47PM +0800, ZF Tang wrote:
> Hi,all
> I'm a newbie to cygwin/linux/unix. Can somebody tell me what's the
> difference between the directories /bin, /usr/bin, and /usr/local/bin ?

http://www.pathname.com/fhs/pub/fhs-2.3.html

-- 
+---+
| Marcel Telka   e-mail:   [EMAIL PROTECTED]  |
|homepage: http://telka.sk/ |
|jabber:   [EMAIL PROTECTED] |
+---+

--
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: Sound generator starting package for Linux/Cygwin

[Reini Urban]

Theo Verelst schrieb:
I've started a sound generator package on Linux and Cygwin consisting
of a sound server program based on Portaudio library connected with
a Tcl/Tk script with user interface, in this case starting with
a 8 harmonics sliders like in an organ.
See:
   http://82.168.209.239/Soundtest
Why not use the standard puredata, which is also based on tcl/tk and 
portaudio, but is stable and mature?

http://pure-data.sourceforge.net/
http://puredata.info/
--
Reini Urban
http://xarch.tu-graz.ac.at/home/rurban/
--
Unsubscribe info:  http://cygwin.com/ml/#unsubscribe-simple
Problem reports:   http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ:   http://cygwin.com/faq/

a newbie's question about directory

[ZF Tang]

Hi,all
I'm a newbie to cygwin/linux/unix. Can somebody tell me what's the
difference between the directories /bin, /usr/bin, and /usr/local/bin ?




--
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/