Re: problem with history

[Sanjay Goel]

Chris January wrote:
>> Hi,
>> my cygwin bash does not write anything in .bash_history .  I think
>> it was working fine before and some changes done by me has caused it
>> to stop updating this file.
>> rxvt does not have this problem.
>> any pointers where could be the problem.
> Close bash by typing exit rather than clicking the Close button.
>
> Chris

I always do that.
Sanjay




--
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/

ImageMagick/PerlMagick

[Bill McCormick]

When last we left our valiant hero, he was finally able to get ImageMagic to
config w/ jpeg. Although the support for EXIF seemed to still be missing
something, he compiled ImageMagick and trudged on to the next step in his
quest... PerlMagick.

Issuing the 'perl Makefile.PL' command, his confidence was again bolstered
by a the happy output of a Makefile generation. Then upon issuing the make
command, from out of nowhere like some ghost in the machine, the make
failed.

Our hero took a deep breath, read various docs and searched the lists, but
upon finding no answer decided to start over from the beginning. Upon doing
more thorough navigation at imagemagick.org web site, he finds binaries for
cygwin. With spirits again bolstered, he downloads that along with all the
delegates and other various items ('cause he's a download doggy.) Hoping
just to get PerlMagick working, following strict software installation
guidelines of making small changes, one at a time, he installs only the
cygwin ImageMagick binaries. Disappointment strikes again; PerlMagick is
still unable to make.

Ok  f*ck the story. It's too late. I hope you got a chuckle anyway.

I got Imagemagick to finally config and build w/ jpeg by uninstalling the
jpeg libs from the cygwin setup and then re-installing. The PerlMagick
Makefile.PL makes a Makefile that does not make. Did that make scense?
During make we get to ...

Creating library file: libMagick.dll.a

and then about a zillion messages like ...

Magick.o(.text+0xbd9):Magick.c: undefined reference to `_GetExceptionInfo'
Magick.o(.text+0xc00):Magick.c: undefined reference to `_CloneImage'
Magick.o(.text+0xc14):Magick.c: undefined reference to
`_DestroyExceptionInfo'
Magick.o(.text+0xc2c):Magick.c: undefined reference to `_CatchException'
Magick.o(.text+0xcbc):Magick.c: undefined reference to `_ReacquireMemory'
Magick.o(.text+0xcce):Magick.c: undefined reference to `_AcquireMemory'
Magick.o(.text+0xdec):Magick.c: undefined reference to `_FormatString'
Magick.o(.text+0xe4a):Magick.c: undefined reference to `_AcquireMemory'
Magick.o(.text+0xe58):Magick.c: undefined reference to `_CloneImageInfo'
Magick.o(.text+0xe6a):Magick.c: undefined reference to `_CloneDrawInfo'
Magick.o(.text+0xe79):Magick.c: undefined reference to `_CloneQuantizeInfo'
Magick.o(.text+0xecd):Magick.c: undefined reference to `_MagickError'
Magick.o(.text+0x1378):Magick.c: undefined reference to `_MagickError'

*snip*

Seems to be a missing lib or path to find it.

Takers?


Thanks,

Bill
---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.502 / Virus Database: 300 - Release Date: 7/18/2003


--
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/

documentation on building the cygwin dll

[Carlo Florendo]

Good Day!

I've searched the net regarding building the cygwin dll and found these.

www.cygwin.com/ml/cygwin/2001-04/msg00709.html
www.cygwin.com/ml/cygwin/2001-04/msg00720.html
sources.redhat.com/ml/cygwin/2001-07/msg00816.html
sources.redhat.com/ml/cygwin/2001-07/msg00817.html

However, not one of them refers to info that shows how to build the cygwin dll.

The faq contains an entry on how to build dlls but *not* how to build the cygwin dll.

All the readme files in the distribution  cygwin-1.3.12-4.tar.bz2   don't mention how 
to build the cygwin dll.  The following are
the
readme files.

./README
./README-maintainer-mode
./config/mpw/README
./include/mpw/README
./libiberty/README
./mpw-README
./newlib/README
./winsup/bz2lib/README
./winsup/bz2lib/README.COMPILATION.PROBLEMS
./winsup/cygwin/regexp/README
./winsup/testsuite/README
./i686-pc-cygwin/newlib/README
./i686-pc-cygwin/libiberty/README
./i686-pc-cygwin/winsup/bz2lib/README
./i686-pc-cygwin/winsup/bz2lib/README.COMPILATION.PROBLEMS
./i686-pc-cygwin/winsup/cygwin/regexp/README
./i686-pc-cygwin/winsup/testsuite/README

Could anyone point out where the documentation to build the cygwin dll is?

Thanks!

Best Regards,
Carlo
--
Carlo Florendo
Astra Philippines Inc.
URL: www.hq.astra.ph/resources












-
Carlo Florendo
Astra (Philippines), Inc.
Email: [EMAIL PROTECTED]
Web: http://www.astra.ph



--
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/

spaces in group names

[luke . kendall]

I notice that Windows allows spaces in group names, unlike Unix, so the
output of the "groups" command can't be processed reliably under
Windows.

Any suggestions?

: /home/luke ; id
uid=11021(luke) gid=10513(Domain Users) groups=12919(adaytum),10513(Domain 
Users),13876(MS_VisualStudio),13761(ZoneAlarm)
: /home/luke ; groups
Domain Users adaytum MS_VisualStudio ZoneAlarm

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/

How to tell if ntsec is on or off

[luke . kendall]

On 23 Jul, I wrote:
>  I was going to qualify this with `when ntsec is defined in CYGWIN' 

It's not easy to find out if ntsec is turned on, is it?  When I wrote
the above, I was thinking "ntsec turned on" means $CYGWIN includes the
word "ntsec".

But I think I've just realised that isn't true, is it?

If it's pre Cygwin 1.3.something-like-18, then it's on if and only if
ntsec is in $CYGWIN, but if it's after, it's on unless $CYGWIN includes
nontsec.  So the actual test you'd have to make would be something like
what I've written here (read "~" as "includes"):

version < 1.3.18 then $CYGWIN ~ \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: practicing building the cygwin.dll

[Carlo Florendo]

Oops, sorry!  I forgot to attach cygcheck output.  Here it is...

- Original Message -
From: "Carlo Florendo" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Wednesday, July 23, 2003 12:25 PM
Subject: practicing building the cygwin.dll


> Hello,
>
> I'm trying to build the cygwin dll from scratch.  I downloaded the entire 
> distribution last year including the sources.  I've
never
> upgraded any package since then.
>
> I have cygwin-1.3.12-4 installed. Attached is the output of cygcheck -svr
>
> I assumed that building the cygwin.dll is as simple as going inside the 
> ./cygwin-1.3.12-4/ directory, running ./configure, ./make,
> and ./make install.
>
> Running ./configure worked well.
> However, I get these errors when running make:
>
> 
>
> newlib.stdlib   ..working in
> atexit.c..linked
> atexit.exp  ..linked
> newlib.string   ..working in
> string.exp  ..linked
> tstring.c   ..linked
> loading cache ../config.cache
> configure: error: can not find install-sh or install.sh in .. ./..
> make: *** [configure-target-newlib] Error 1
>
> I wondered why it was looking for install-sh or install.sh in its parent directory.  
> I copied install-sh to the parent directory
and
> still got the same error.
>
>
> The next thing I did was to go to the ./cygwin-1.3.12-4/winsup/ directory to try 
> building it from there.
> Running configure worked well.  When I ran ./make, here's what I got:
>
> make[1]: Entering directory `/usr/include/w32api'
> make[1]: *** No rule to make target `all'.  Stop.
> make[1]: Leaving directory `/usr/include/w32api'
>
> make: *** [w32api] Error 1
>
> My /usr/include/w32api directory exists.
>
> Lastly, I thought that going to the ./cygwin-1.3.12-4/winsup/cygwin/ directory would 
> build the cygwin.dll
> Running configure worked well.  When I ran make, I got:
>
> g++ -c -gstabs+ -O2 -MMD -fbuiltin ... cygheap.cc
> In file included from cygheap.cc:17:
> fhandler.h: In method `select_stuff::select_stuff()':
> fhandler.h:1200: implicit declaration of function `int memset(...)'
> make: *** [cygheap.o] Error 1
>
>
> So, with the classic question found in this M/L: WAIDW (What am I doing wrongly)?
>
> Thanks!
>
> Best Regards,
>
> Carlo
> --
> Carlo Florendo
> Astra Philippines Inc.
> URL: http://www.astra.ph/resources
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> --
> 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 Win95/NT Configuration Diagnostics
Current System Time: Tue Jul 23 10:00:58 2003

Windows 2000 Professional Ver 5.0 Build 2195 

Path:   F:\cygwin\usr\local\bin
F:\cygwin\bin
F:\cygwin\bin
.\
d:\TEXMF\MIKTEX\BIN
c:\WINNT\SYSTEM32
c:\WINNT
c:\WINNT\SYSTEM32\WBEM
d:\ULTRAEDT
d:\BORLAND\CBUILDER3\BIN
c:\Program Files\Common Files\System\ado
d:\Microsoft Visual Studio\Common\Tools\WinNT
d:\Microsoft Visual Studio\Common\MSDev98\Bin
d:\Microsoft Visual Studio\Common\Tools
d:\Microsoft Visual Studio\VC98\bin
e:\temp\img\ImageMagick
d:\borland\bcc55\bin
F:\cygwin\usr\X11R6\bin
F:\cygwin\bin\

SysDir: C:\WINNT\System32
WinDir: C:\WINNT

CYGWIN = `nontsec'
HOME = `e:\fcarlo'
MAKE_MODE = `unix'
PWD = `/cygdrive/c/temp'
USER = `fcarlo'

ALLUSERSPROFILE = `C:\Documents and Settings\All Users'
APPDATA = `C:\Documents and Settings\fcarlo\Application Data'
BASH_ENV = `/etc/bashrc'
CLASSPATH = `D:\Exceed.nt\hcljrcsv.jar;D:\Exceed.nt\hcljrcsv.jar;;'
COLORFGBG = `15;default;0'
COLORTERM = `rxvt-xpm'
COMMONPROGRAMFILES = `C:\Program Files\Common Files'
COMPUTERNAME = `THORIN'
COMSPEC = `C:\WINNT\system32\cmd.exe'
DISPLAY = `:0'
GS_FONTPATH = 
`.:/usr/share/lilypond/1.6.0/fonts/type1:/usr/share/texmf/fonts/type1:/usr/share/texmf/fonts/type1/adobe:/usr/share/texmf/fonts/type1/adobe/utopia:/usr/share/texmf/fonts/type1/bitstrea:/usr/share/texmf/fonts/type1/bitstrea/charter:/usr/share/texmf/fonts/type1/bluesky:/usr/share/texmf/fonts/type1/bluesky/cm:/usr/share/texmf/fonts/type1/bluesky/cmextra:/usr/share/texmf/fonts/type1/bluesky/cyrillic:/usr/share/texmf/fonts/type1/bluesky/euler:/usr/share/texmf/fonts/type1/bluesky/symbols:/usr/share/texmf/fonts/type1/hoekwater:/usr/share/texmf/fonts/type1/hoekwater/context:/usr/share/texmf/fonts/type1/hoekwater/mflogo:/usr/share/texmf/fonts/type1/hoekwater/rsfs:/usr/share/texmf/fonts/type1/hoekwater/stmaryrd:/usr/share/texmf/fonts/type1/hoekwater/wasy:/usr/share/texmf/fonts/type1/public:/usr/share/texmf/fonts/type1/public/antp:/usr/share/texmf/fonts/type1/public/antt:/usr/share/texmf/fonts/type1/public/belleek:/usr/share/texmf/fonts/type1/public/cmcyr:/usr/share/texmf/fonts/type1/public/cs:/usr/share/texmf/fonts/type1/public/marvosym:/usr/share/texmf/fonts/type1/public/mathpazo:

Re: Permission problems with recent Cygwin

[luke . kendall]

(Sorry this took so long; it's taken me days to find the hours required
to investigate.)

I've got some more info, below, and have also read through all the
permissions threads on the cygwin mailing list I could find, and to read
about half the ntsec threads over the last year or so.  I don't think it
explains the problem we're having.

(BTW: one thing that struck me was that the change that turned on ntsec
by default seems to have caused a reasonable number of problems.  And
although I'm also subscribed to the cygwin-announce list, if there was
an announcement about it, it didn't stick in my memory.  But that may
just be my faulty memory, sorry.)

What is especially confusing I think is that you can't trust the mode
output from ls.  Permissions reported via ls may bear little or no
resemblance to the user's actual permissions.  I was going to qualify
this with `when ntsec is defined in CYGWIN', but it seems to be
independent of that.  (See "Odd thing #2.)Anyway ...

Odd thing #1


I compared two systems running Cygwin 1.3.22: a fresh install, vs my old
system (updated via setup.exe).

The odd thing is that in both cases, ntsec is *not* on by default.
It *is* on if I slogin to a machine.  (Using password authentication.)

Perhaps ssh modifies CYGWIN before logging in?

FYI, I've set things up here to define a site-wide setting for CYGWIN by
modifying it in "cisra.sh".  (This is instead of expecting over a
hundred users to manually set CYGWIN in their control panel.)

#
# The env. variable CYGWIN may contain a list of space-separated keywords.
# Correct handling of binary and text files requires the nobinmode keyword.
# In case there is a pre-existing setting (e.g. via ssh), add it to the end.
#
export CYGWIN
case "$CYGWIN " in
*binmode *)
;;
*)
CYGWIN="$CYGWIN nobinmode"
;;
esac

*and* by having the sample .profile created for all users include this
at the end:

#
# If CYGWIN is defined, then its setting has no effect except on subsequent
# shells.  So the last thing we do is exec a shell.  This is needed to get
# correct handling of binary and text files via CYGWIN=nobinmode
# The STARTX is to guard against execing in startx.bat, if people want to
# use the --login argument.
#
if [ -x "$SHELL" -a "$CYGWIN" != "" -a -z "$STARTX" -a -z "$DISPLAY" ]
then
exec $SHELL
fi

Anyway, logging in directly winds up with ntsec *not* set; specifically,
CYGWIN is set to nobinmode.  An slogin results in this bizarre setting:

: /home/luke ; echo $CYGWIN
"binmode ntsec tty"

which I can't explain.  Note the extra double-quotes there, too.

Odd thing #2


The fresh machine belongs to user Joal, who has admin rights on his
machine.  (Here, each machine is set up so the user has admin rights on
their own machine.)

This is what Joal gets by running getfacl on /etc/profile on his machine:

joalt:/etc /etc
$ getfacl /etc/profile
# file: /etc/profile
# owner: desmolej
# group: Domain Users
user::rwx
group::r-x
mask:rwx
other:r-x

And after changing mode to 700:

joalt:/etc /etc
$ chmod 700 /etc/profile

joalt:/etc /etc
$ getfacl /etc/profile
# file: /etc/profile
# owner: desmolej
# group: Domain Users
user::rwx
group::---
mask:rwx
other:---

Compare that to the very different result I get if I run getfacl on
/etc/profile on my machine:

# file: /etc/profile
# owner: robs
# group: Domain Users
user::rwx
user:Administrators:rwx
group::---
group:SYSTEM:rwx
mask:rwx
other:---

For Joal, on his machine, his groups are:
joalt:/home/joalt ~
$ id
uid=14765(joalt) gid=10513(Domain Users) groups=10513(Domain Users)

For me, on my machine and on Joal's:
: /home/luke ; id
uid=11021(luke) gid=10513(Domain Users) groups=12919(adaytum),10513(Domain Users
),13876(MS_VisualC6),13761(ZoneAlarm)

Looking at the security properties via Explorer, it lists

Administrators (DOYLE\Administrators)
Domain Users (CISRA\Domain Users)\
Everyone
Robert Smuhar (CISRA\robs)
SYSTEM

Administrators includes "robs" and "desmolej"; they and SYSTEM have
permissions: Full control, Modify, Read & Execute, Read, and Write.

"robs" has every permission except Full Control.

Domain users and Everyone have only Special Permissions.

After doing a "chmod a+rx /etc/profile" an extra group got added
to the Security properties (?!) like so:

Users (Doyle\Users)

This group persisted even after doing a "chmod 700 /etc/profile"
And even then, this new group has interesting permissions:
Read & Execute, and Read.

This is all on my machine, where I can always read /etc/profile
even when it is this mode:

-rwx--+   1 robs Domain U  640 Sep 17  2002 /etc/profile

On Joal's machine, where this mode prevents users from reading the
file (as if it were a real Unix system), the "Users (Doyle\Users)"
does not appear in the security properties view of the file in
Explorer.

Both are on XP machines, both on NTFS file s

Re: Cygwin version 1.3.20

[Peter A. Castro]

On Mon, 21 Jul 2003, Max Bowsher wrote:

> Henry Da Costa wrote:
> > Hello Mr. Faylor,
>
> You have sent this message to a public mailing list.
>
> > I hope you can help me. I'm looking for a complete Cygwin version 1.3.20
> > set of packages but have so far been unable to find it.
>
> The phrase "complete Cygwin version 1.3.20 set of packages" is meaningless.
> A Cygwin DLL version number does not uniquely identify a set of packages.
>
> > I need that
> > particular version of Cygwin because other versions including the latest
> > one give us an error when used from a development tool for which we
> > don't have the source. Do you know if that Cygwin version is archived
> > anywhere?
>
> It isn't.

That's not entirely true (oh, boy, I'm puttin' my foot in it now...).

If Henry happens to have the setup.ini that he used to install 1.3.20, he
might be able to re-install the versions of those packages, provided he
can find a server which has those package versions still.

FWIW, I have been keeping a running archive of the cygwin website, and
thus all cygwin packages as of 4/15/2002.  I started doing revisioning of
the setup.ini as of 3/11/2003.  The earliest revisions of setup.ini I
have do specify 1.3.20-1 of cygwin, but, of course, that doesn't
guarantee it lists the exact versions of other, related, packages.  It's
pretty much a crap-shoot.

My ftp server is publicly accessable, so, again, provided Henry can find
the correct setup.ini, and provided the latest setup.exe can still
understand the old setup.ini, he quite possibly could recreate the
"complete Cygwin version 1.3.20 set of packages" :)

This is all theory, but if it's that important, then it might be worth
Henry's efforts.  On the other hand, others have already voiced their
concerns on using older verions of packages, and I share those concerns.
However I also understand that sometimes you really do need to re-create
and older environment.

> Max.

-- 
Peter A. Castro <[EMAIL PROTECTED]> or <[EMAIL PROTECTED]>
"Cats are just autistic Dogs" -- Dr. Tony Attwood

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

Re: How to resolve a link?

[Randall R Schulz]

At 20:08 2003-07-22, Christopher Faylor wrote:
...

>>>What would happen on *nix?
>>
>>The same thing as on cygwin.
>
>Really?
>
>Since Cygwin "strace" is not a Cygwin program, it does not get Cygwin
>symbolic links resolved. Whereas on Unix or Linux, a symlink would be
>resolved by the kernel if it was used by strace and it (strace) would
>successfully get its target executed.
AFAIK, we were talking about "which" not strace.  I thought it was obvious
why strace wasn't working.  strace will translate cygwin paths these days
but it won't (yet) follow cygwin symlinks.


Right. My mistake.

RRS

--
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/

practicing building the cygwin.dll

[Carlo Florendo]

Hello,

I'm trying to build the cygwin dll from scratch.  I downloaded the entire distribution 
last year including the sources.  I've never
upgraded any package since then.

I have cygwin-1.3.12-4 installed. Attached is the output of cygcheck -svr

I assumed that building the cygwin.dll is as simple as going inside the 
./cygwin-1.3.12-4/ directory, running ./configure, ./make,
and ./make install.

Running ./configure worked well.
However, I get these errors when running make:



newlib.stdlib   ..working in
atexit.c..linked
atexit.exp  ..linked
newlib.string   ..working in
string.exp  ..linked
tstring.c   ..linked
loading cache ../config.cache
configure: error: can not find install-sh or install.sh in .. ./..
make: *** [configure-target-newlib] Error 1

I wondered why it was looking for install-sh or install.sh in its parent directory.  I 
copied install-sh to the parent directory and
still got the same error.


The next thing I did was to go to the ./cygwin-1.3.12-4/winsup/ directory to try 
building it from there.
Running configure worked well.  When I ran ./make, here's what I got:

make[1]: Entering directory `/usr/include/w32api'
make[1]: *** No rule to make target `all'.  Stop.
make[1]: Leaving directory `/usr/include/w32api'

make: *** [w32api] Error 1

My /usr/include/w32api directory exists.

Lastly, I thought that going to the ./cygwin-1.3.12-4/winsup/cygwin/ directory would 
build the cygwin.dll
Running configure worked well.  When I ran make, I got:

g++ -c -gstabs+ -O2 -MMD -fbuiltin ... cygheap.cc
In file included from cygheap.cc:17:
fhandler.h: In method `select_stuff::select_stuff()':
fhandler.h:1200: implicit declaration of function `int memset(...)'
make: *** [cygheap.o] Error 1


So, with the classic question found in this M/L: WAIDW (What am I doing wrongly)?

Thanks!

Best Regards,

Carlo
--
Carlo Florendo
Astra Philippines Inc.
URL: http://www.astra.ph/resources
















--
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's vanilla sed : capabilities and limitations

[Peter S Tillier]

Brian Dessent wrote:
> [EMAIL PROTECTED] wrote:
>
>> Q1. Querying info sed reveals the expression matcher to be "greedy",
>> matching the longest possible string. Is there a way to make it
>> match the shortest possible, so that echo aaabbbccc | sed 's/^.*b//'
>> (altered but similar) grabs aaab not aaabbb?
>
> If you have perl available (or just a tool that uses perl-compatible
> regexps, i.e. grep -P) you can add the '?' character after any
> qualifier to get the non-greedy version, i.e. '*?' is the non-greedy
> '*', '??' is the non-greedy '?', etc.  But this is a feature of pcre,
> which I don't believe applies to sed in any shape or form.  However,
> most sed scripts are pretty easy to do in perl with little
> modification, so if you
> require this function that's what I'd do.
>
> Brian

The GNU sed maintainer Paolo Bonzini also maintains a version of sed
called ssed (super sed) that does support PCREs in addition to BREs and
EREs (which are both now supported in GNU sed 4.x+).

For more on sed:
http://sed.sourceforge.net/

HTH

Peter S Tillier
"Who needs perl when you can write dc, sokoban,
arkanoid and an unlambda interpreter in sed?"


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

Re: How to resolve a link?

[Christopher Faylor]

On Tue, Jul 22, 2003 at 07:56:02PM -0700, Randall R Schulz wrote:
>Chris,
>
>At 18:29 2003-07-22, Christopher Faylor wrote:
>>On Tue, Jul 22, 2003 at 09:23:13PM -0400, David A. Cobb wrote:
>>>Recently, I was trying to do "strace Xemacs . . ."
>>>First I got a "No such file" error, so I changed to do "strace `which
>>>xemacs` " -- still a failure.
>>>
>>>which xemacs returns "/usr/local/bin/xemacs.exe.lnk"; that is, my normal
>>>handle to launch xemacs is a symlink to the executable whose name or
>>>location varies with the version-number.
>>>
>>>Given that its purpose is to locate what executable file one will use in
>>>a particular environment, should not 'which' resolve the symlink and
>>>return its target?
>>
>>No.
>>
>>>What would happen on *nix?
>>
>>The same thing as on cygwin.
>
>Really?
>
>Since Cygwin "strace" is not a Cygwin program, it does not get Cygwin 
>symbolic links resolved. Whereas on Unix or Linux, a symlink would be 
>resolved by the kernel if it was used by strace and it (strace) would 
>successfully get its target executed.

AFAIK, we were talking about "which" not strace.  I thought it was obvious
why strace wasn't working.  strace will translate cygwin paths these days
but it won't (yet) follow cygwin symlinks.
--
Please use the resources at cygwin.com rather than sending personal email.
Special for spam email harvesters: send email to [EMAIL PROTECTED]
and be permanently blocked from mailing lists at sources.redhat.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: How to resolve a link?

[Randall R Schulz]

Chris,

At 18:29 2003-07-22, Christopher Faylor wrote:
On Tue, Jul 22, 2003 at 09:23:13PM -0400, David A. Cobb wrote:
>Recently, I was trying to do "strace Xemacs . . ."
>First I got a "No such file" error, so I changed to do "strace `which
>xemacs` " -- still a failure.
>
>which xemacs returns "/usr/local/bin/xemacs.exe.lnk"; that is, my normal
>handle to launch xemacs is a symlink to the executable whose name or
>location varies with the version-number.
>
>Given that its purpose is to locate what executable file one will use in
>a particular environment, should not 'which' resolve the symlink and
>return its target?
No.

>What would happen on *nix?

The same thing as on cygwin.
Really?

Since Cygwin "strace" is not a Cygwin program, it does not get Cygwin 
symbolic links resolved. Whereas on Unix or Linux, a symlink would be 
resolved by the kernel if it was used by strace and it (strace) would 
successfully get its target executed.

So wouldn't the behavior on Cygwin differ from a comparable invocation 
on Linux or Unix?

Randall Schulz 

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

Re: How to resolve a link?

[David A. Cobb]

Christopher Faylor wrote:

On Tue, Jul 22, 2003 at 09:23:13PM -0400, David A. Cobb wrote:
 

Recently, I was trying to do "strace Xemacs . . ."
First I got a "No such file" error, so I changed to do "strace `which 
xemacs` " -- still a failure.

which xemacs returns "/usr/local/bin/xemacs.exe.lnk"; that is, my normal 
handle to launch xemacs is a symlink to the executable whose name or 
location varies with the version-number.

Given that its purpose is to locate what executable file one will use in 
a particular environment, should not 'which' resolve the symlink and 
return its target?
   

No.

 

What would happen on *nix?
   

The same thing as on cygwin.

OK, but would strace be equally unable to start the subprocess from a 
link? 
IIRC, one major irritation in Windoz is that one cannot do certain 
operations via a link ("shortcut") whereas in *nix the link is a 
nearly-complete surrogate for its target. 

Of course, if there is a less-obvious but successful solution to the 
problem I'd be happy to hear about it.

--
Please use the resources at cygwin.com rather than sending personal email.
Special for spam email harvesters: send email to [EMAIL PROTECTED]
and be permanently blocked from mailing lists at sources.redhat.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/
 

--
David A. Cobb, Software Engineer, Public Access Advocate
"By God's Grace I am a Christian man, by my actions a great sinner." -- The Way of a 
Pilgrim; R. M. French, tr.
Life is too short to tolerate crappy software.
.


--
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: Skipping the /proc filesystem

[Randall R Schulz]

David,

At 18:15 2003-07-22, David A. Cobb wrote:
Maybe this is something any native *nix speaker knows, but I'm stull 
trudging up the learning curve.
It is entirely non-Cygwin-specific, yes.


If I do a (cygwin) find for some fragment of a filename, I get a whole 
pile of hits in the /proc/registry area - none of which is 
relevant.  I would wish to tell find not to get involved with the 
/proc filesystem at all.

Can that easily be done?
Very easily:

% find / -path '/proc' -prune -o -print

To paraphrase, "find starting in slash pruning away any path names that 
begin with "/proc" and print all others."

Randall Schulz


--
David A. Cobb


--
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: gdb passes different argv/argc to mingw & cygwin

[Shankar Unni]

Sam Steingold wrote:

> args='"4 4" '"'5 5'"

What you're seeing is the side-effect of how mingw-compiled programs
handle wildcards.

mingw-compiled programs automatically link in support for wildcard
expansion of arguments passed in via the Windows GetCommandLine(). So
you'll often see extra quotes being consumed, and other such effects.
It's unavoidable when dealing with Windows..



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

Re: How to resolve a link?

[Randall R Schulz]

David,

Investigate the options to "test" (the binary or the BASH built-in) 
that detect symbolic links: -h or -L (they are synonymous) and the 
command "readlink" (as in "man readlink").

Again, this is stock Unix / Linux stuff.

Randall Schulz

At 18:23 2003-07-22, David A. Cobb wrote:
Recently, I was trying to do "strace Xemacs . . ."
First I got a "No such file" error, so I changed to do "strace `which 
xemacs` " -- still a failure.

which xemacs returns "/usr/local/bin/xemacs.exe.lnk"; that is, my 
normal handle to launch xemacs is a symlink to the executable whose 
name or location varies with the version-number.

Given that its purpose is to locate what executable file one will use 
in a particular environment, should not 'which' resolve the symlink 
and return its target?
What would happen on *nix?

--
David A. Cobb


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

Re: How to resolve a link?

[Christopher Faylor]

On Tue, Jul 22, 2003 at 09:23:13PM -0400, David A. Cobb wrote:
>Recently, I was trying to do "strace Xemacs . . ."
>First I got a "No such file" error, so I changed to do "strace `which 
>xemacs` " -- still a failure.
>
>which xemacs returns "/usr/local/bin/xemacs.exe.lnk"; that is, my normal 
>handle to launch xemacs is a symlink to the executable whose name or 
>location varies with the version-number.
>
>Given that its purpose is to locate what executable file one will use in 
>a particular environment, should not 'which' resolve the symlink and 
>return its target?

No.

>What would happen on *nix?

The same thing as on cygwin.
--
Please use the resources at cygwin.com rather than sending personal email.
Special for spam email harvesters: send email to [EMAIL PROTECTED]
and be permanently blocked from mailing lists at sources.redhat.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: subprocess i/o interaction with shell (bash&cmd): shells compete for input with user program!

[Christopher Faylor]

On Tue, Jul 22, 2003 at 06:12:53PM -0400, Sam Steingold wrote:
>> * In message <[EMAIL PROTECTED]>
>> * On the subject of "Re: subprocess i/o interaction with shell (bash&cmd): shells 
>> compete for input with user program!"
>> * Sent on Tue, 22 Jul 2003 17:27:05 -0400
>> * Honorable Christopher Faylor <[EMAIL PROTECTED]> writes:
>>
>> If the exec paradigm in cygwin was really broken, the mailing list
>> would be filled with people complaining about that fact.
>
>my fault (gcc does not define UNIX, I should have been more careful,
>sorry)
>
>OK, so -- how do I implement execv() ?
>winsup/src/winsup/cygwin/spawn.cc:spawn_guts() is, apparently,
>the place to look at, but it is so hairy ...

It is hairy but, unfortunately, this is not the place to go for help
rolling your own execv.  Cygwin already has a nice, functioning execv.
As always, I'd suggest google.
--
Please use the resources at cygwin.com rather than sending personal email.
Special for spam email harvesters: send email to [EMAIL PROTECTED]
and be permanently blocked from mailing lists at sources.redhat.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/

How to resolve a link?

[David A. Cobb]

Recently, I was trying to do "strace Xemacs . . ."
First I got a "No such file" error, so I changed to do "strace `which 
xemacs` " -- still a failure.

which xemacs returns "/usr/local/bin/xemacs.exe.lnk"; that is, my normal 
handle to launch xemacs is a symlink to the executable whose name or 
location varies with the version-number.

Given that its purpose is to locate what executable file one will use in 
a particular environment, should not 'which' resolve the symlink and 
return its target?
What would happen on *nix?

--
David A. Cobb, Software Engineer, Public Access Advocate
"By God's Grace I am a Christian man, by my actions a great sinner." -- The Way of a 
Pilgrim; R. M. French, tr.
Life is too short to tolerate crappy software.
.


--
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/

Skipping the /proc filesystem

[David A. Cobb]

Maybe this is something any native *nix speaker knows, but I'm stull 
trudging up the learning curve.

If I do a (cygwin) find for some fragment of a filename, I get a whole 
pile of hits in the /proc/registry area -
none of which is relevant.  I would wish to tell find not to get 
involved with the /proc filesystem at all.

Can that easily be done?

--
David A. Cobb, Software Engineer, Public Access Advocate
"By God's Grace I am a Christian man, by my actions a great sinner." -- The Way of a 
Pilgrim; R. M. French, tr.
Life is too short to tolerate crappy software.
.


--
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 Apache hanging problem

[Ajay Nerurkar]

Corinna,

Do you remember what your fix
(http://sources.redhat.com/ml/cygwin/2003-06/msg00055.html) was?

We are using modified older Cygwin sources and we would like to drop
in this fix. 

Thanks.

Ajay

__
Do you Yahoo!?
Yahoo! SiteBuilder - Free, easy-to-use web site design software
http://sitebuilder.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: subprocess i/o interaction with shell (bash&cmd): shellscompete for input with user program!

[Sam Steingold]

> * In message <[EMAIL PROTECTED]>
> * On the subject of "Re: subprocess i/o interaction with shell (bash&cmd): shells 
> compete for input with user program!"
> * Sent on Tue, 22 Jul 2003 17:27:05 -0400
> * Honorable Christopher Faylor <[EMAIL PROTECTED]> writes:
>
> If the exec paradigm in cygwin was really broken, the mailing list
> would be filled with people complaining about that fact.

my fault (gcc does not define UNIX, I should have been more careful,
sorry)

OK, so -- how do I implement execv() ?
winsup/src/winsup/cygwin/spawn.cc:spawn_guts() is, apparently,
the place to look at, but it is so hairy ...

-- 
Sam Steingold (http://www.podval.org/~sds) running w2k
  
 
What garlic is to food, insanity is to art.


--
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: subprocess i/o interaction with shell (bash&cmd): shells competefor input with user program!

[Steve Coleman]

Sam Steingold wrote:

I am not using fork().
just plain exec().
 

If thats the case then the exec() you are linking is not behaving as a 
POSIX exec() call, because if it did you would not still have two 
processes to even compete for the I/O streams. The first process image 
would be "replaced" by the second.

% man execl

8< -
"The  exec family of functions replaces the current process image with a 
new process
image."
-

So, either you are not using Cygwin execl(), or it is VERY broken, which 
I doubt because too many applications have been working correctly for 
too long for the second to be the case.





--
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: subprocess i/o interaction with shell (bash&cmd): shells compete for input with user program!

[Christopher Faylor]

On Tue, Jul 22, 2003 at 04:41:18PM -0400, Sam Steingold wrote:
>> * In message <[EMAIL PROTECTED]>
>> * On the subject of "Re: subprocess i/o interaction with shell (bash&cmd): shells 
>> compete for input with user program!"
>> * Sent on Tue, 22 Jul 2003 13:23:16 -0400
>> * Honorable Christopher Faylor <[EMAIL PROTECTED]> writes:
>>
>> On Tue, Jul 22, 2003 at 12:23:18PM -0400, Sam Steingold wrote:
>> >> Under cygwin you'd accomplish this by having the parent process use
>> >> one of the "exec" calls.
>> >
>> >nope.  doesn't work either.
>> >I get the same "input competition" when I use exec().
>> 
>> If you are using fork/exec, then sure, you'll get that.
>
>I am not using fork().  just plain exec().

Ok.  Then you're not building cygwin applications apparently.   Next
assumption is that you're using the -mno-cygwin option.

If the exec paradigm in cygwin was really broken, the mailing list would
be filled with people complaining about that fact.

cgf
--
Please use the resources at cygwin.com rather than sending personal email.
Special for spam email harvesters: send email to [EMAIL PROTECTED]
and be permanently blocked from mailing lists at sources.redhat.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: subprocess i/o interaction with shell (bash&cmd): shellscompete for input with user program!

[Sam Steingold]

> * In message <[EMAIL PROTECTED]>
> * On the subject of "Re: subprocess i/o interaction with shell (bash&cmd): shells 
> compete for input with user program!"
> * Sent on Tue, 22 Jul 2003 13:23:16 -0400
> * Honorable Christopher Faylor <[EMAIL PROTECTED]> writes:
>
> On Tue, Jul 22, 2003 at 12:23:18PM -0400, Sam Steingold wrote:
> >> Under cygwin you'd accomplish this by having the parent process use
> >> one of the "exec" calls.
> >
> >nope.  doesn't work either.
> >I get the same "input competition" when I use exec().
> 
> If you are using fork/exec, then sure, you'll get that.

I am not using fork().
just plain exec().

> If you use just exec, then cygwin should do the right thing.

it does not.
I use the simple
execl("runtime.exe",NULL);
and I get the "input competition".

-- 
Sam Steingold (http://www.podval.org/~sds) running w2k
  
 
Lisp is a language for doing what you've been told is impossible. - Kent Pitman


--
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/

PTY read/write problems

[Hirsch, Matthew]

Hi list,

I'm having trouble sending data back and forth between two processes over a
pseudo terminal.  I'm able to send data from the slave tty to the master pty
without any problem.  However, no data can be read from the slave side of
the pty (tty).  I've attached an example program that demonstrates my
problem.  I compiled it with

gcc ptyrw.c -g -o ptyrw -Wall

I have cygwin version 1.3.22-1 (Curr in setup as of yesterday).

I've attempted to follow the suggestions made in earlier discussions of
pseudo terminals (http://sources.redhat.com/ml/cygwin/2002-11/msg00064.html
for example).  Has anyone written an example of a working pty/tty connection
under Cygwin?  I would appreciate it very much if someone could set me
straight on this.

The output I'm getting with my test program is as follows:

$ ./ptyrw
MASTER: Master Alive: My PID: 232
<< Master writing to Slave >>
*Max Read Size: 2147483647
MASTER: Created master pty: 3
MASTER: Associated Slave: /dev/tty0
MASTER: Forking Slave...
MASTER: Preparing to write: Sleeping for 2 to let slave get set up
SLAVE: Slave Alive: My PID: 128
SLAVE: Setting Session ID
SLAVE: Opening slave end of pty
SLAVE: Preparing to read, sleeping for 5 while master writes
MASTER: Writing "Master Send"
MASTER: Wrote 15 bytes: "Master Send"
MASTER: Sleeping for 10 while slave finishes
SLAVE: Attempting to read from fd 4
READERROR: Data did not become available in 3 seconds
SLAVE: Read Error
SLAVE: Quitting...
MASTER: Quitting...

Whereas in this direction things work as expected:

$ ./ptyrw 1
MASTER: Master Alive: My PID: 1124
<< Slave writing to Master >>
*Max Read Size: 2147483647
MASTER: Created master pty: 3
MASTER: Associated Slave: /dev/tty0
MASTER: Forking Slave...
MASTER: Preparing to read, sleeping for 5 while slave writes
SLAVE: Slave Alive: My PID: 1896
SLAVE: Setting Session ID
SLAVE: Opening slave end of pty
SLAVE: Wrote 15 bytes: "Slave Send"
SLAVE: Sleeping for 10 while master finishes
MASTER: Attempting to read from fd 3
MASTER: Read 7 bytes: "Slave S"
MASTER: Sleeping for 10 while slave finishes
SLAVE: Quitting...
MASTER: Quitting...

The fact that read returns data in blocks of 7 bytes or less seems a bit
cumbersome but not out of spec.

Thank you very much for any suggestions.
Matt Hirsch



ptyrw.c
Description: Binary data
--
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/

More 1.5 success

[Gary R Van Sickle]

Actually using the 7-10 snapshot, plus everything in [test] as of a few
hours ago.  Compiling and installing distcc on two machines (one Why2K, one
XP) worked with nary a hitch.  Using distcc to compile gcc cvs HEAD as a
cross-compiler on the two boxes worked fantastically.  Distcc was in tcp
mode (distcc --daemon), didn't try it over ssh.

I love it when a plan comes together.

Still getting these crazy double-plusses and minuses though.

--
Gary R. Van Sickle
Braemar Inc.
11481 Rupp Dr.
Burnsville, MN 55337


--
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/

/etc/profile as installed. Minor correction.

[Mr Tharic Jroskj]

The file /etc/profile ships with these lines:

# uncomment to use the terminal colours set in
DIR_COLOR
eval `dircolors -b /etc/DIR_COLOR`

This should be with an S appended to reflect the
shipped file name:

# uncomment to use the terminal colours set in
DIR_COLORS
eval `dircolors -b /etc/DIR_COLORS`

Enjoy,

Charles Merriam
cygwin atsign truegift.com



__
Do you Yahoo!?
Yahoo! SiteBuilder - Free, easy-to-use web site design software
http://sitebuilder.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: Postmaster failing 'not enough core'

[Jason Tishler]

Kevin,

[Sorry for the sluggish response time, but I have been distracted by the
impending Python 2.3 release...]

On Sat, Jul 19, 2003 at 02:33:35PM +0100, Kevin Jones wrote:
> I'm running Cygwin 1.3.22.1 on Windows 2003.

Note that I do not have access to Windows 2003, so you will probably
have to debug this yourself...

> I've installed ipc-daemon as a service and I'm trying to get
> postmaster started.
> 
> [snip]
> 
> "IpcMemoryCreate: shmget(key=5432001, size=1499136, 03600) failed: Not
> enough core"
> 
> followed by the rest of the message. I've tried reducing buffer sizes to
> 16 (the minimum) and the max connections to 8 (as can be seen above) and
> I'm still getting the problem.

This could be a cygipc under 2003 problem -- adjusting PostgreSQL
parameters most likely will not have any affect.

> I suspect it's a permissions problem (most of these seem to be:) ) Do I
> need to do anything special to the postgres user, or allow it access to
> any particular directory?
> 
> Any help would be much appreciated.

Does ipctest work?

$ ipctest h 
Test v0.03
Shared memory segment 132 created
shmat return code= 0
shmat ok adr= 0x68
ok

Does removing the cygipc temp files help?

$ rm -fr /tmp/MultiFile* /tmp/cygipc_?

Does making /tmp world writable help?

$ chmod 777 /tmp

If none of the above helps, then you will have to build a debug version
of cygipc, rebuild PostgreSQL against it, figure out which cygipc
operation is failing with ENOMEM, and *why* it is failing.

Jason

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

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

Re: make 3.80 and VPATH

[Christopher Seawood]

Christopher Faylor wrote:

> On Sat, Jul 12, 2003 at 11:19:17AM -0400, Christopher Faylor wrote:
> >>That example works fine if I do 'make --win32' but fails when just 
using
>>>'make'.  That is a change in behavior from the previous version of 
make.
> >> The testcase works with either of --unix & --win32 when using make
> >>3.79.1-7 .
> >
> >As it should.  Colon is a path separator.  So, to answer your 
question, yes
> >it is by design that VPATH is not translating MS-DOS paths.  If it 
was doing
> >so before, it was clearly a bug.

Ok.  That's what I was attempting to determine.  We have been relying 
upon that behavior for the past year that we've been using cygwin to 
build Mozilla.  That will clearly need to change.

> For the record, I've just resurrected a make 3.79 version and tried 
this and
> I don't detect any difference in behavior.

That's really odd since others are reporting this exact same problem 
that goes away when they downgrade to 3.79.1-7.  You tried it with 
--unix as well?

(Sorry about the delayed response.  I was offline last week for Comic-Con.)

- cls



--
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 history

[Chris January]

> Hi,
> my cygwin bash does not write anything in .bash_history .  I think it was
> working fine before and some changes done by me has caused it to stop
> updating this file.
> rxvt does not have this problem.
> any pointers where could be the problem.
Close bash by typing exit rather than clicking the Close button.

Chris


--
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: subprocess i/o interaction with shell (bash&cmd): shells compete for input with user program!

[Christopher Faylor]

On Tue, Jul 22, 2003 at 12:23:18PM -0400, Sam Steingold wrote:
>> Under cygwin you'd accomplish this by having the parent process use
>> one of the "exec" calls.
>
>nope.  doesn't work either.
>I get the same "input competition" when I use exec().

If you are using fork/exec, then sure, you'll get that.  If you use
just exec, then cygwin should do the right thing.

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: [SEMI-OFFTOPIC] Request for Ralf Habacker

[Ralf Habacker]

> Ralf,
> I appreciate the acknowledgement on your kde-cygwin web site but I
> really would rather not have my email address available on your
> acknowledgement page.
> 
> I know that Corinna probably feels the same way and I suspect that
> Charles Wilson, Egor Duda, and Robert Collins probably all would rather
> not have their addresses available for spammers to harvest.  There are
> a few other people displayed there as well.  I can't speak for them but
> I would assume that most would feel similarly.
> 
> I don't think there is any reason why users of kde-cygwin would need
> to contact any of us personally, so I would suggest just referring people
> to the cygwin web site if they have general questions or concerns.
> 
> I would appreciate your prompt attention to this matter.  I see several
> attempts by spammers to send email to the address that you display on
> your web site, so I know that your site is currently being harvested.
> --
No problem. The email adresses are removed. 

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/

Re: subprocess i/o interaction with shell (bash&cmd): shellscompete for input with user program!

[Sam Steingold]

> * In message <[EMAIL PROTECTED]>
> * On the subject of "Re: subprocess i/o interaction with shell (bash&cmd): shells 
> compete for input with user program!"
> * Sent on Tue, 22 Jul 2003 10:57:02 -0400
> * Honorable Christopher Faylor <[EMAIL PROTECTED]> writes:
>
> On Tue, Jul 22, 2003 at 10:30:52AM -0400, Sam Steingold wrote:
> >So, how can I exit driver and have the shell notice that its child
> >(driver) left a heir (runtime) and that the shell (bash & cmd) should
> >wait for runtime to finish.
> 
> You have two processes reading standard input at the same time?

yes, bash/cmd & runtime do read stdin simultaneously,
but I do NOT want that!

> The expected behavior in that situation is that it is unpredictable
> which process gets which keystroke.

yep.

> You want the parent of the parent of a process to know that there is a
> grandchild process sitting around waiting for input and have the
> grandparent (bash or cmd) wait for the grandchild to finish?

precisely!

> Under cygwin you'd accomplish this by having the parent process use
> one of the "exec" calls.

nope.  doesn't work either.
I get the same "input competition" when I use exec().

> There is no way to do what you want in a normal Windows program.

are you sure? :-(

-- 
Sam Steingold (http://www.podval.org/~sds) running w2k
  
 
Daddy, why doesn't this magnet pick up this floppy disk?


--
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/

[SEMI-OFFTOPIC] Request for Ralf Habacker

[Christopher Faylor]

Ralf,
I appreciate the acknowledgement on your kde-cygwin web site but I
really would rather not have my email address available on your
acknowledgement page.

I know that Corinna probably feels the same way and I suspect that
Charles Wilson, Egor Duda, and Robert Collins probably all would rather
not have their addresses available for spammers to harvest.  There are
a few other people displayed there as well.  I can't speak for them but
I would assume that most would feel similarly.

I don't think there is any reason why users of kde-cygwin would need
to contact any of us personally, so I would suggest just referring people
to the cygwin web site if they have general questions or concerns.

I would appreciate your prompt attention to this matter.  I see several
attempts by spammers to send email to the address that you display on
your web site, so I know that your site is currently being harvested.
--
Please use the resources at cygwin.com rather than sending personal email.
Special for spam email harvesters: send email to [EMAIL PROTECTED]
and be permanently blocked from mailing lists at sources.redhat.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: X server crashed when Netscape 4 & 7 visiting sites

[Christopher Faylor]

On Tue, Jul 22, 2003 at 03:40:16PM -, Joel Handler wrote:
>I just downloaded the entire Cygwin package and am experimenting around with 
>it.  I am running the cygwin X server and am ssh (tunneling X) to a linux 
>machine and running either Netscape 4 or 7.  Both seem to crash the X server 
>(sometimes netscape crashes with a Bus Error, other times it just closes and 
>other times the entire X server crashes) when visiting certain pages (one 
>example is www.citrix.com).  If I use a different X server (Hummingbird 
>Exceed) everything works fine.  
>
>Any ideas or help would be greatly appreciated,

Redirecting to correct mailing list.

--
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/

X server crashed when Netscape 4 & 7 visiting sites

[Joel Handler]

I just downloaded the entire Cygwin package and am experimenting around with 
it.  I am running the cygwin X server and am ssh (tunneling X) to a linux 
machine and running either Netscape 4 or 7.  Both seem to crash the X server 
(sometimes netscape crashes with a Bus Error, other times it just closes and 
other times the entire X server crashes) when visiting certain pages (one 
example is www.citrix.com).  If I use a different X server (Hummingbird 
Exceed) everything works fine.  

Any ideas or help would be greatly appreciated,
Thanks,
---
Joel T. Handler
Institute for Advanced Study
School of Natural Sciences - NT System Administrator




--
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: subprocess i/o interaction with shell (bash&cmd): shells compete for input with user program!

[Christopher Faylor]

On Tue, Jul 22, 2003 at 10:30:52AM -0400, Sam Steingold wrote:
>So, how can I exit driver and have the shell notice that its child
>(driver) left a heir (runtime) and that the shell (bash & cmd) should
>wait for runtime to finish.

Am I reading this correctly?  You have two processes reading standard
input at the same time?  The expected behavior in that situation is that
it is unpredictable which process gets which keystroke.  You want the
parent of the parent of a process to know that there is a grandchild
process sitting around waiting for input and have the grandparent (bash
or cmd) wait for the grandchild to finish?

Under cygwin you'd accomplish this by having the parent process use one
of the "exec" calls.  There is no way to do what you want in a normal
Windows program.

cgf
--
Please use the resources at cygwin.com rather than sending personal email.
Special for spam email harvesters: send email to [EMAIL PROTECTED]
and be permanently blocked from mailing lists at sources.redhat.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/

subprocess i/o interaction with shell (bash&cmd): shells competefor input with user program!

[Sam Steingold]

I have a driver.exe program which calls runtime.exe.
runtime.exe is an interactive console application, i.e.,
it will be doing i/o via the stdio.
If I make driver.exe exit before runtime.exe, the console interaction
is severely broken (both under bash and cmd, both when compiled with and
without cygwin)

E.g.:
=== start driver.c ==
#include 
#include 
#include 
#include 

int main (int argc, char *argv[]) {
  PROCESS_INFORMATION pi;
  STARTUPINFO si;
  si.cb = sizeof(STARTUPINFO);
  si.lpReserved = NULL;
  si.lpDesktop = NULL;
  si.lpTitle = NULL;
  si.cbReserved2 = 0;
  si.lpReserved2 = NULL;
  si.dwFlags = STARTF_USESTDHANDLES;
  si.hStdInput  = GetStdHandle(STD_INPUT_HANDLE);
  if (si.hStdInput == INVALID_HANDLE_VALUE)  goto w32err;
  si.hStdOutput = GetStdHandle(STD_OUTPUT_HANDLE);
  if (si.hStdOutput == INVALID_HANDLE_VALUE) goto w32err;
  si.hStdError  = GetStdHandle(STD_ERROR_HANDLE);
  if (si.hStdError == INVALID_HANDLE_VALUE)  goto w32err;
  printf(" * start (%d) %s\n",argc,argv[0]);
  if (!CreateProcess("runtime.exe","runtime.exe",NULL,NULL,1,0,
 NULL,NULL,&si,&pi))
goto w32err;
  if (argc>1)
WaitForSingleObject(pi.hProcess,INFINITE);
  printf(" * finish %s\n",argv[0]);
  return 0;
 w32err:
  fprintf(stderr," * error %s: %s\n",argv[0],strerror(GetLastError()));
  return 1;
}
=== end driver.c ==
=== start runtime.c ==
#include 
#include 
#include 

int main (int argc, char *argv[]) {
  char buf[BUFSIZ];
  printf(" * start (%d) %s\n",argc,argv[0]);
  gets(buf);
  printf(" * you entered: [%s]\n",buf);
  printf(" * finish %s\n",argv[0]);
  return 0;
}
=== end runtime.c ==

I compile them with
gcc -mno-cygwin -g -W -Wall -Wstrict-prototypes -Wshadow driver.c -o driver
gcc -mno-cygwin -g -W -Wall -Wstrict-prototypes -Wshadow runtime.c -o runtime

When I give the driver an argument so that it will wait for runtime to
finish, I get what one would expect, both under bash and cmd:

 start bash interaction ===
bash$ ./driver.exe wait
 * start (2) d:\sds\c\driver.exe
 * start (1) runtime.exe
qwert
 * you entered: [qwert]
 * finish runtime.exe
 * finish d:\sds\c\driver.exe
bash$
 end bash interaction ===

 start cmd interaction ===
D:\sds\c>driver wait
 * start (2) driver
 * start (1) runtime.exe
qwert
 * you entered: [qwert]
 * finish runtime.exe
 * finish driver

D:\sds\c>
 end cmd interaction ===

now, if I do not give the `wait' argument to the driver, I get
something quite unexpected under cmd:

 start cmd interaction ===
D:\sds\c>driver
 * start (1) driver
 * finish driver

D:\sds\c> * start (1) runtime.exe
qwer
'qwer' is not recognized as an internal or external command,
operable program or batch file.

D:\sds\c>asdf
 * you entered: [asdf]
 * finish runtime.exe
 end cmd interaction ===

i.e., I get the cmd prompt right after driver exits and the first thing
I input is caught by cmd (and the second line goes to runtime).

 start bash interaction ===
bash$ ./driver.exe
 * start (1) d:\sds\c\driver.exe
 * finish d:\sds\c\driver.exe
bash$  * start (1) runtime.exe
defxctutirfefe
bash: defxctutirfefe: command not found
bash$ 3sd
bash: 3sd: command not found
bash$
bash$ p4khdsf
bash: p4khdsf: command not found
bash$ ^C
bash$
 end bash interaction ===

i.e., I get the bash prompt right after driver exits, BUT the input
becomes weird: only about a quarter of the characters I type are echoed
to the screen (the `3sd' string was generated by me repeatedly
typing "asdasdfasdfadfasdfadsf") and the only way out is Ctrl-C (then
everything is restored to normal, but I never see the finishing messages
from runtime.exe).

Ironically, if I compile them with cygwin:
gcc -g -W -Wall -Wstrict-prototypes -Wshadow driver.c -o driver
gcc -g -W -Wall -Wstrict-prototypes -Wshadow runtime.c -o runtime

the results are identical on bash and cmd and somewhat in between the
mingw behavior (i.e., bash is better and cmd is worse):

 start bash interaction ===
bash$ ./driver.exe
 * start (1) ./driver
 * finish ./driver
bash$  * start (1) runtime
sfgggdfrg
bash: sfggdrg: command not found
bash$ rwr
bash: rwr: command not found
bash$
bash$
bash$
bash$
bash$
bash$
bash$
bash$
bash$
bash$
bash$ 54
bash: 5: command not found
bash$ 5676
 * you entered: [gf4676]
 * finish runtime

bash: 5: command not found
bash$
bash$
 end bash interaction ===

 start cmd interaction ===
D:\sds\c>driver
 * start (1)

typo in description of the "cmake" package

[Joe Buehler]

There is a typo in the description of the "cmake" package as displayed
by the "setup" program:
cmake: A cross platform build manger
--
Joe Buehler


--
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] Test/cygwin-1.5.0 version of Midnight Commander

[Pavel Tsekov]

The Midnight Commander visual shell has been updated to version
4.6.0a-20030721-1 .

This release will _NOT_ work with versions of Cygwin prior to 1.5.0 .

Along with the ability to work with large file this new version will bring
you a lot of
bug fixes and some exciting new features. This release is based on the CVS
codebase
as of 21 July 2003.

A list of some of the visible changes follows:

* The ministatus bar is redrawn if a search invoked by M-s or C-s was
canceled.

* The 'Free VFS now' menu item was removed. The functionality can be 
   accessed now from 'Active VFS list' dialog.

* Treat Shift-Backspace as Backspace.

* Add backward search support to the internal viewer.

* Some fixes to the ftp VFS.

* Emacs style file locking support for the internal editor.

Please, give this version a test and report any bugs you might
encounter to the Cygwin mailing list.

NOTE: It appears that MC triggers a crash in the Cygwin dll if one
  tries to use the internal viewer on multigigabyte files. I am
  aware of this issue so please do not send reports bug reports
  about it.

Pavel Tsekov
Midnight Commander maintainer for Cygwin

INSTALLATION:

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 'mc' from the 'Utils' category. Do not forget that this is a test 
version so you need to click on the 'Exp' radio buttong before
selecting the package.

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.

In the US, ftp://mirrors.rcn.net/mirrors/sources.redhat.com/cygwin/
is a reliable high bandwidth connection.

In Japan, ftp://ftp.u-aizu.ac.jp/pub/gnu/gnu-win32/ is already
updated.

In DK, http://mirrors.sunsite.dk/cygwin/ is usually up-to-date.

If one of the above doesn't have the latest version of this package
you can either wait for the site to be updated or find another
mirror.

Please send questions or comments to the Cygwin mailing list at:
cygwin at sources dot redhat dot com . If you want to subscribe go to:
http://cygwin.com/lists.html I would appreciate if you would use
this mailing list rather than emailing me directly. This includes
ideas and comments about the setup utility or Cygwin in general.

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

*** 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:

cygwin-announce-unsubscribe-YOU=YOURDOMAIN dot COM at cygwin dot com

-- 
+++ GMX - Mail, Messaging & more  http://www.gmx.net +++

Jetzt ein- oder umsteigen und USB-Speicheruhr als Prämie sichern!


--
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: expat-1.95.6-1

[Gerrit P. Haase]

expat-1.95.6-1 has been uploaded to the Cygwin net distribution.

Expat is an XML parser library written in C.

It is a stream-oriented parser in which an application registers 
handlers for things the parser might find in the XML document 
(like start tags).  Homepage: http://www.libexpat.org/

CHANGES (since expat-1.95.5-1):

- Added XML_FreeContentModel().
- Added XML_MemMalloc(), XML_MemRealloc(), XML_MemFree().
- Fixed a variety of bugs: see SF issues 615606, 616863, 618199, 653180, 673791.
- Enhanced the regression test suite.
- Man page improvements: includes SF issue 632146.


INSTALLATION:

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 `expat' from the `Libs' category.

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.

In the US, ftp://mirrors.rcn.net/mirrors/sources.redhat.com/cygwin/
is a reliable high bandwidth connection.

In Japan, ftp://ftp.u-aizu.ac.jp/pub/gnu/gnu-win32/ is updated fast.

In Denmark, http://mirrors.sunsite.dk/cygwin/ is usually up-to-date.

If one of the above doesn't have the latest version of this package
you can either wait for the site to be updated or find another
mirror.

For a brief descripton of what this package is and what it contains,
see http://cygwin.com/packages/ .


Please send questions or comments to the Cygwin mailing list at:
[EMAIL PROTECTED] .  If you want to subscribe go to:
http://cygwin.com/lists.html .  I would appreciate if you would use
this mailing list rather than emailing me directly.  This includes
ideas and comments about the setup utility or Cygwin in general.

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


  *** 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]
  

Gerrit P. Haase
-- 
=^..^=   http://cygwin.com/cygwin/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 starting cron as a service

[Bowden, Todd]

Bill,

The docs should be /usr/doc/Cygwin/cron.README.  The docs should be there.

Todd C. Bowden 


-Original Message-
From: Bill McCormick [mailto:[EMAIL PROTECTED] 
Sent: Monday, July 21, 2003 4:22 PM
To: [EMAIL PROTECTED]
Subject: RE: problem starting cron as a service


Ok, I missed the -D param. Just going from the man pages, cron takes no
params. I also didn't find any info regarding running cron as a service in
the /usr/doc's. So, as always, it's the littlest things that bite.

Any idea regarding my previous post about ImageMagick?

thanks mucho,

Bill

> -Original Message-
> From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] 
> Behalf Of Bowden, Todd
> Sent: Monday, July 21, 2003 3:21 PM
> To: '[EMAIL PROTECTED]'
> Subject: RE: problem starting cron as a service
>
>
> You have to install it as a service in order to see in from a Windows 
> perspective.
>
> cygrunsrv -I cron -p /usr/sbin/cron -a -D
> Cygrunsrv -S cron
>
> net start | grep -i cron
>
> I see it in my Services.  Why wouldn't you see it in yours
>
> The doc /usr/doc/Cygwin/cron.README tells you how to install it.
>
> Todd C. Bowden
> HP Certified
> AtosOrigin
> 5000 S. Bowen
> Arlington, TX 76017
> Office: 817-264-8211
> E-mail: [EMAIL PROTECTED]
>
>
> -Original Message-
> From: Bill McCormick [mailto:[EMAIL PROTECTED]
> Sent: Monday, July 21, 2003 3:12 PM
> To: [EMAIL PROTECTED]
> Subject: RE: problem starting cron as a service
>
>
> Oops ... my bad.
>
> It's running; it's just not displayed as running as a service to 
> windows. ps -x -f shows the grand truth. That being the case, doesn't 
> cron startup really belong in rc.local? Umm ... does rc.local even get 
> called on system boot?
>
> Thanks,
>
> Bill
>
> > -Original Message-
> > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
> > Behalf Of Vince Hoffman
> > Sent: Monday, July 21, 2003 11:10 AM
> > To: [EMAIL PROTECTED]
> > Subject: RE: problem starting cron as a service
> >
> >
> > Opps thought from your previous message you already knew or i would 
> > have said. the easiest way is the "at" trick, from a command 
> > line,type "at \\yourcomputername 15:49 /interactive cmd.exe" this 
> > will give you a cmd shell running as SYSTEM. from there cd to your 
> > cygwin\bin directory and type "bash --login -i" which will give
> you a bash
> > shell running as SYSTEM.
> >
> > > -Original Message-
> > > From: Bill McCormick [mailto:[EMAIL PROTECTED]
> > > Sent: 21 July 2003 16:41
> > > To: [EMAIL PROTECTED]
> > > Subject: RE: problem starting cron as a service
> > >
> > >
> > > How do I run a prog as localsystem?
> > >
> > > > -Original Message-
> > > > From: [EMAIL PROTECTED]
> > > [mailto:[EMAIL PROTECTED] Behalf
> > > > Of Vince Hoffman
> > > > Sent: Monday, July 21, 2003 9:38 AM
> > > > To: '[EMAIL PROTECTED]'; '[EMAIL PROTECTED]'
> > > > Subject: RE: problem starting cron as a service
> > > >
> > > >
> > > > Hi Bill,
> > > > Off hand there are a few things to try, but Mark Harig
> > > on the cygwin
> > > > list seems pretty up on cron, and has a script which will
> > > diagnose common
> > > > problems, heres a link to a message which has it attached, 
> > > > 
> > > >
> > > > You'll do better keeping messages on list as theres a
> > > whole lot more
> > > > experience out there than I have. One thing you could try
> > > is running cron
> > > > manualy from the command line as localsystem and seeing if you 
> > > > get any clues in the output. (appologys if you have and i missed 
> > > > that
> > > > message.)
> > > >
> > > > Vince
> > > > ps. the output of cygcheck -s -v -r as an uncompressed 
> > > > attachment might help too.
> > > > > -Original Message-
> > > > > From: Bill McCormick [mailto:[EMAIL PROTECTED]
> > > > > Sent: 21 July 2003 15:18
> > > > > To: Vince Hoffman
> > > > > Subject: RE: problem starting cron as a service
> > > > >
> > > > >
> > > > > Thanks Vince,
> > > > >
> > > > > Nothing in cron.log (file created but empty) and mount is 
> > > > > system. Any other ideas?
> > > > >
> > > > >
> > > > > Bill
> > > > > > -Original Message-
> > > > > > From: [EMAIL PROTECTED]
> > > > > [mailto:[EMAIL PROTECTED] Behalf
> > > > > > Of Vince Hoffman
> > > > > > Sent: Monday, July 21, 2003 4:08 AM
> > > > > > To: '[EMAIL PROTECTED]'; Cygwin
> > > > > > Subject: RE: problem starting cron as a service
> > > > > >
> > > > > >
> > > > > > anything in /var/log/cron.log ?
> > > > > >
> > > > > > and are you mounts system or user ?
> > > > > > ie.
> > > > > > $ mount |grep "/ type system"
> > > > > >
> > > > > > should give something like
> > > > > >
> > > > > > h:\cygwin on / type system (binmode)
> > > > > >
> > > > > > > -Original Message-
> > > > > > > From: Bill McCormick [mailto:[EMAIL PROTECTED]
> > > > > > > Sent: 21 July 2003 08:25
> > > > > > > To: Cygwin
> > > > > > > Subject: problem starting cron as a service
> > > > > > 

Re: ImageMagick

[Gerrit P. Haase]

Hallo David,

>  o libwmf works too, but isn't distributed with cygwin

Wouldn't it be a good idea to find a maintainer who cares about
libwmf, littlecms, libmng and some other graphic libraries like
GD, libexif, libmnote, jasper, ...?


Gerrit
-- 
=^..^=


--
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: ImageMagick

[Gerrit P. Haase]

Hallo Bill,

You wrote:

> I thought I had the jpeg lib from cygwin installed and that I needed
> something more. Your note prompted me to check again and sure enough, I
> didn't have it installed - DUH!! So after installing that - Imagemagick is
> now ready to fly with jpeg.

Fine.

> I DL'd your lib's from your link (note that the files want to save as
> *.tar.tar as opposed to *.tar.bz2,) renamed them and untarzip'd them into
> place.

I always use wget to download over HTTP.


> I run Imagemagick configure again and it looks like things are still
> broken there:


> *snip*
> EXIF  --with-exif=yes   no (failed tests)
> *snip*

> So now I'm trying to rebuild from your sources ...

> [EMAIL PROTECTED]/usr/local/exif-0.6

> *snip*

> checking for libexif >= 0.5.9... gnome-config: not found
> gnome-config: not found

This is the problem, IIRC you're missing the gnome-common package.


> Package libexif was not found in the pkg-config search path.
> Perhaps you should add the directory containing `libexif.pc'
> to the PKG_CONFIG_PATH environment variable
> No package 'libexif' found

> configure: error: Library requirements (libexif >= 0.5.9) not met; consider
> adjusting the PKG_CONFIG_PATH environment variable if your libraries are in
> a nonstandard prefix so pkg-config can find them.

> 

> The same for gexif-0.5.

> So off I go pouring over doc's and such (actually, now I need to work on
> something else for a while and I don't need to chew camera tags at the
> moment.) I'll get back to this before too long though.

> I'll check the PKG_CONFIG_PATH env var - could that be the whole problem?.
> I'm pretty sure I need to use the --enable-shared switch during configure
> and I think that means I neither need nor want --enable-static; True? Is
> there a specific build order I need to follow or maybe I've yet to install
> some other prerequisite?

See comment above.  I don't know exactley what all is needed to get IM
compiled.  I usually do `autoreconf --install --verbose --force`
before configuring to ensure that ltmain.sh and the other autotool
scripts are updated to the most recent versions, then libtool takes
care of building shared libs automatically if the package supports
shared libs (that is ...LIBLDFLAGS = -no-undefined ...), but I
believe, IM has already Cygwin support and it should work out of the
tarball.


Gerrit

P.S.: Please fix your quoting (that means, don't include email
  adresses in the body of emails when posting to public mailing
  lists, 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: GCC: Split it ?

[Gareth Pearce]

> IMHO, this looks like a PTC situation..

prehaps more likely MOTC.

;)


--
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: GCC: Split it ?

[Ronald Landheer-Cieslak]

IMHO, this looks like a PTC situation..

packaging issues belong on cygwin-apps, by the way..

rlc

On Tue, 22 Jul 2003, Frédéric L. W. Meunier wrote:

> The GCC 3.2-3 package is almost 3 times the size of the
> 2.95.4-10 mainly due to the Java part. Wouldn't it help split
> it in various packages, like one for C and C++, and others for
> the remaining (and less used) languages ?
> 
> Just a suggestion.
> 
> --
> 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: GCC Errors - two versions released

[Ronald Landheer-Cieslak]

On Mon, 21 Jul 2003, William J. Moulton wrote:
> Has anyone noticed that two versions of GCC were released with the
> latest versions of cygwin?
As always, of course..

> I experienced many compile and makefile errors only to discover it was
> the GCC compiler version.  And have watched problem after problem be
> reported to the newsgroup about the compiler and makefile.  Version 2.95
> is called by adding -2 to the call for the compiler (eg. gcc-2 or g++-2,
> etc.)  Using just GCC (or g++) calls version 3.x.  After making this
> change all my compile and makefile problems disappeared.
.. which has been the case for quite a while now..

gcc V3 is significantly different from gcc V2, and some projects (may) 
depend on V2. Hence, our gracious gcc maintainer decided in his 
benevolence to provide us (the grateful and ingrate users alike) with 
both. gcc2 contains version 2.95.3; gcc contains version 3.2

> Also wouldn't it make sense to release the most stable version as gcc
> and the newer version as gcc2 of gcc-dev?  The everyone could avoid some
  ^ -> older?
> headaches.
Yes, that would make sense, which is exactly why this is the way it is. 
(reading s/newer/older/, of course)

rlc



--
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 with history

[Sanjay Goel]

Hi,
my cygwin bash does not write anything in .bash_history .  I think it was
working fine before and some changes done by me has caused it to stop
updating this file.
rxvt does not have this problem.
any pointers where could be the problem.
TIA
Sanjay




--
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: when is it useful to attach strace outputs to the M/L

[Ronald Landheer-Cieslak]

Though it is contrary to my common practice to chime in on a thread cgf 
has already replied to, a message I sent was cited here, so I might as 
well clarify.

On Tue, 22 Jul 2003, Carlo Florendo wrote:
> Sometimes, when I post problems here, I attach the output of cygcheck
> and strace.  Could anyone point out when it's relevant to attach strace
> outputs here?  In the post
> http://cygwin.com/ml/cygwin/2003-07/msg01072.html the output of strace
> was not deemed to be useful.
The reason the strace output was not useful for debugging the problem I 
had (have) is that the problem itself does not show up on the strace. 

Strace shows the system (Cygwin DLL) calls made by a Cygwin program. It 
can be useful for debugging problems withing the Cygwin DLL. Most problems 
are not directly related to the Cygwin DLL though (most are cockpit 
errors) which means in most cases, strace output is not useful.

The Lynx problem is (or rather: seems to be) a Cygwin DLL problem which is 
kinda hard to debug (see the thread about it for more info). I sent the 
strace output (or rather: made it available so as not to pollute mailboxes 
with useless attachments) in the hope the Cygwin core developers would see 
something I didn't - that hope was in vain.

That being clarified, cgf, the Ruler, Czar and Monarch of Cygwin is, of 
course, right (but IMHO, adding a link to a site containing strace output 
never hurts).

rlc



--
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: tiff-3.6.0-2 NEW: libtiff-devel-3.6.0-2, libtiff3-3.6.0-2

[Charles Wilson]

The tiff package has been updated to version 3.6.0-2.  This is NOT a
64bit, cygwin-1.5.0 test release.  It is a cygwin-1.3.22 release;
upgrade now.  However, the upcoming 64bit release will have a DLL with a
new name, because libtiff depends on libjpeg, and because libjpeg's ABI
changed, so did libtiff's.  But that's all in the future.  For now, this
release is purely preparatory for the 64bit release.

I repeat, THIS release is NOT a cygwin-1.5.0 test release.  It is
cygwin-1.3.22-compatible; upgrade now.

The only difference in this release and the previous 3.6.0-1 release, is
that here we have split into three:

libtiff3-- contains cygtiff3.dll
libtiff-devel   -- contains headers, import/static libs, man3 pages
tiff-- everything else (incl. utilities)

This is the ONLY change between 3.6.0-1 and 3.6.0-2

Charles Wilson
tiff volunteer maintainer for cygwin

INSTALLATION:

To update your installation, click on the "Install Cygwin now" link on
the http://sources.redhat.com/cygwin web page.  This downloads setup.exe
to your system.

Run setup and answer all of the questions.  You must choose a
mirror, since direct downloads from 'cygwin.com' are no longer
allowed.

ftp://mirrors.rcn.net/pub/sourceware/cygwin/ (US)
ftp://linux.sarang.net/mirror/development/compiler/cygwin/ (Korea)
ftp://ftp.mirror.ac.uk/sites/sourceware.cygnus.com/pub/cygwin (UK)

The setup.exe program will figure out what needs to be updated on your
system and should install the tiff package automatically.

If you have questions or comments, please send them to the Cygwin
mailing list at:  [EMAIL PROTECTED] .  I would appreciate if you
would use this mailing list rather than emailing me directly. This
includes ideas and comments about the setup utility or Cygwin general.

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

*** CYGWIN-ANNOUNCE UNSUBSCRIBE INFO ***

If you want 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]

NOTES:

o The DLL exports a few variables that are "bad" -- in the
sense that they cannot be auto-imported (arrays of structs, etc).
However, these are library private and should NOT be accessed by
external programs.  If you find a rogue program out there that
messes with these internal data objects, let me know...

On the other hand, --enable-pseudo-relocs should take care of the 
problem, as an interim solution for those rogues.


--
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/