FAQ Search Today's Posts Mark Forums Read
» Video Reviews

» Linux Archive

Linux-archive is a website aiming to archive linux email lists and to make them easily accessible for linux users/developers.


» Sponsor

» Partners

» Sponsor

Go Back   Linux Archive > Ubuntu > Ubuntu Masters Of The Universe

 
 
LinkBack Thread Tools
 
Old 12-08-2008, 10:25 PM
Angelo Graziosi
 
Default Problems with Cernlib-2006

I do not know if you are still interested to these things but it seems that
there is some problems with Cernlib-2006.


I have a fortran program (GEANT simulation) which outputs figures (one per
page) in a PS file. I can see the PS file with GV or GSView but if I extract a
page from it to another PS file, the page are wrong, only a wrong part of the
figure is visible and GV or GSView do not emit any message.


The same happens with PAW. I have attached a simple test case to reproduce the
thing.


The PAW macro, test.kumac, produces a PS file with four pages (test.ps). If one
extracts, for example the page 3 (with GV, mark page 3 and 'Save Marked'
produces test.ps_pages) then reading it with GV (or GSView) shows that it is
wrong, but without error messages.


I have found these problems both on GNU/Linux Kubuntu 8.04 Hardy with the
CERNLIB of this distribution (are you the maintainer?), both on Cygwin with my
build of Cernlib.


If I use GSView on Windows, the same things happen.

Restoring Cernlib-2005 (Vogt, CERN binaries, my build etc., on linux or
Cygwin...) fixes the problems!


So it looks as a BUG of Cernlib-2006.


Cheers,
Angelo.--
Ubuntu-motu mailing list
Ubuntu-motu@lists.ubuntu.com
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-motu
 
Old 12-10-2008, 02:42 AM
"Kevin B. McCarty"
 
Default Problems with Cernlib-2006

Angelo,

There was a patch regarding printing that went into Cernlib 2006 that
fixed postscript output problems observed by someone else. Probably
this is related. You can take a look at the history of the earlier
bug report at http://bugs.debian.org/329207 .

The patch as suggested by Frank Dohrmann exists in the version of
Cernlib 2005 in Debian "Etch", but upstream changed it slightly before
incorporating it into Cernlib 2006 (now in Debian Lenny and Sid, and
in Ubuntu Gutsy and more recent). I've attached a file "test-etch.ps"
created with your PAW macro on my machine running Etch. If you use
the attached file as input to GV or GSview, does it exhibit the
problem, or not?

- Kevin

On Mon, Dec 8, 2008 at 3:25 PM, Angelo Graziosi
<Angelo.Graziosi@roma1.infn.it> wrote:
>
>
> I do not know if you are still interested to these things but it seems that
> there is some problems with Cernlib-2006.
>
> I have a fortran program (GEANT simulation) which outputs figures (one per
> page) in a PS file. I can see the PS file with GV or GSView but if I extract
> a page from it to another PS file, the page are wrong, only a wrong part of
> the figure is visible and GV or GSView do not emit any message.
>
> The same happens with PAW. I have attached a simple test case to reproduce
> the thing.
>
> The PAW macro, test.kumac, produces a PS file with four pages (test.ps). If
> one extracts, for example the page 3 (with GV, mark page 3 and 'Save Marked'
> produces test.ps_pages) then reading it with GV (or GSView) shows that it is
> wrong, but without error messages.
>
> I have found these problems both on GNU/Linux Kubuntu 8.04 Hardy with the
> CERNLIB of this distribution (are you the maintainer?), both on Cygwin with
> my build of Cernlib.
>
> If I use GSView on Windows, the same things happen.
>
> Restoring Cernlib-2005 (Vogt, CERN binaries, my build etc., on linux or
> Cygwin...) fixes the problems!
>
> So it looks as a BUG of Cernlib-2006.
>
>
> Cheers,
> Angelo.



--
Kevin B. McCarty <kmccarty@gmail.com>
WWW: http://www.starplot.org/
WWW: http://people.debian.org/~kmccarty/
GPG: public key ID 4F83C751
--
Ubuntu-motu mailing list
Ubuntu-motu@lists.ubuntu.com
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-motu
 
Old 12-10-2008, 08:37 AM
Angelo Graziosi
 
Default Problems with Cernlib-2006

Kevin,

your test case works just fine, both on Kubuntu 8.04 Hardy and on Cygwin,
with GV and GSView.

If you see the differences between my and your test case (extracted pages,
for example)

[...]
/Zone {/iy exch def /ix exch def ix 1 sub 2224 mul 1 iy sub 3144
mul t} def
%%EndProlog
+%%BeginSetup
+%%EndSetup
+newpath gsave 20 28 t .25 .25 scale gsave
%%Page: number 1
gsave
1 1 Zone

you can see they seem to be related to the bug 329207.

It looks that the 'slightly' changes 'before incorporating it into Cernlib
2006' have introduced a regression.

So, may I expect an update of Cernlib 2006 on Kubuntu 8.04 Hardy?

Where can I find that patch before the 'slightly' changes, so I can apply
it to my Cygwin builds?


Thanks,
Angelo.



On Tue, 9 Dec 2008, Kevin B. McCarty wrote:

> Angelo,
>
> There was a patch regarding printing that went into Cernlib 2006 that
> fixed postscript output problems observed by someone else. Probably
> this is related. You can take a look at the history of the earlier
> bug report at http://bugs.debian.org/329207 .
>
> The patch as suggested by Frank Dohrmann exists in the version of
> Cernlib 2005 in Debian "Etch", but upstream changed it slightly before
> incorporating it into Cernlib 2006 (now in Debian Lenny and Sid, and
> in Ubuntu Gutsy and more recent). I've attached a file "test-etch.ps"
> created with your PAW macro on my machine running Etch. If you use
> the attached file as input to GV or GSview, does it exhibit the
> problem, or not?
>
> - Kevin
>
> On Mon, Dec 8, 2008 at 3:25 PM, Angelo Graziosi
> <Angelo.Graziosi@roma1.infn.it> wrote:
> >
> >
> > I do not know if you are still interested to these things but it seems that
> > there is some problems with Cernlib-2006.
> >
> > I have a fortran program (GEANT simulation) which outputs figures (one per
> > page) in a PS file. I can see the PS file with GV or GSView but if I extract
> > a page from it to another PS file, the page are wrong, only a wrong part of
> > the figure is visible and GV or GSView do not emit any message.
> >
> > The same happens with PAW. I have attached a simple test case to reproduce
> > the thing.
> >
> > The PAW macro, test.kumac, produces a PS file with four pages (test.ps). If
> > one extracts, for example the page 3 (with GV, mark page 3 and 'Save Marked'
> > produces test.ps_pages) then reading it with GV (or GSView) shows that it is
> > wrong, but without error messages.
> >
> > I have found these problems both on GNU/Linux Kubuntu 8.04 Hardy with the
> > CERNLIB of this distribution (are you the maintainer?), both on Cygwin with
> > my build of Cernlib.
> >
> > If I use GSView on Windows, the same things happen.
> >
> > Restoring Cernlib-2005 (Vogt, CERN binaries, my build etc., on linux or
> > Cygwin...) fixes the problems!
> >
> > So it looks as a BUG of Cernlib-2006.
> >
> >
> > Cheers,
> > Angelo.
>
>
>
> --
> Kevin B. McCarty <kmccarty@gmail.com>
> WWW: http://www.starplot.org/
> WWW: http://people.debian.org/~kmccarty/
> GPG: public key ID 4F83C751
>

--
Ubuntu-motu mailing list
Ubuntu-motu@lists.ubuntu.com
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-motu
 
Old 12-10-2008, 08:05 PM
"Kevin B. McCarty"
 
Default Problems with Cernlib-2006

On Wed, Dec 10, 2008 at 1:37 AM, Angelo Graziosi
<Angelo.Graziosi@roma1.infn.it> wrote:
>
> Kevin,
>
> your test case works just fine, both on Kubuntu 8.04 Hardy and on Cygwin,
> with GV and GSView.
[...]
> It looks that the 'slightly' changes 'before incorporating it into Cernlib
> 2006' have introduced a regression.

OK. Thanks for testing. I will try to revert the state of the code
to what it was with (Cernlib 2005 + Frank Dohrmann's patch). I'll
include this in the next upload of Cernlib to Debian Sid. I can't
guarantee that said upload will make it into the next Debian stable
release (Lenny).

> So, may I expect an update of Cernlib 2006 on Kubuntu 8.04 Hardy?

For this you would need to ask the Ubuntu folks. I do not give it
very good odds, since the policy of updating stable releases (in
Debian and, as far as I know in Ubuntu) is that a bug has to be fairly
serious (e.g. security issue or program crashes) to outweigh the risks
of introducing new code into a stable release.

You can certainly download the Ubuntu source package for Hardy, apply
the patch, and rebuild to obtain your own packages ... the comments at
http://people.debian.org/~kmccarty/cernlib/faq.html#22 apply. In
particular you'll want to issue the "fakeroot debian/rules patch"
command BEFORE editing the source code yourself and then running
"fakeroot debian/rules binary". (Unless you want to play with
dpatch.)

Note 2: if you are running PAW and Geant321 on an x86_64 (amd64)
system, you'll also need to rebuild the PAW package against the new
Cernlib packages since PAW is statically linked on 64-bit systems.

> Where can I find that patch before the 'slightly' changes, so I can apply
> it to my Cygwin builds?

I'll send you a patch this evening that reverts Cernlib 2006 back to
the state of (Cernlib 2005 + patch).

--
Kevin B. McCarty <kmccarty@gmail.com>
WWW: http://www.starplot.org/
WWW: http://people.debian.org/~kmccarty/
GPG: public key ID 4F83C751

--
Ubuntu-motu mailing list
Ubuntu-motu@lists.ubuntu.com
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-motu
 

Thread Tools




All times are GMT. The time now is 12:51 PM.

VBulletin, Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Content Relevant URLs by vBSEO ©2007, Crawlability, Inc.
Copyright 2007 - 2008, www.linux-archive.org