[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#312265: [Thomas Ruedas] Re: Error messages in texdoctk (was: Bug#312265: tetex-bin: texdoctk: gv is now in /usr/bin/)



--- Begin Message ---
Hi,
just a short reply now as a sign that I haven't ignored your mail.
On Thursday 09 June 2005 10:29, frank wrote:
> If one of the actions (like view) that call external programs fail, the
> error message (spit out on stderr) is either displayed in the terminal
[...]
> many users (/usr/bin/X11/gv: No such file or directory), while "Error
> while calling 'see somefile.ps': /usr/bin/X11/gv: No such file or
Yes, that would probably be helpful. However, I can't think of an easy and 
fast fix for all instances right now, because in particular the viewer is 
started as a child process (at least if texdoctk runs in quiet mode); see 
the following bit in the sub viewslc:
	  } else { # viewer messages written to /dev/null instead of terminal
 	    if ($browser == 0) { # normal viewers or new Netscape/Mozilla
	      system("perl -we 'use strict; \
my \$dump=\"\"; \
open(NULL,\">/dev/null\"); \
open (VIEWQ,\"$viewer $slcdoc |\"); \
while (\$dump=<VIEWQ>) { print NULL \$dump; } \
close(VIEWQ); \
close(NULL);' &");
	    } else { # open doc in existing window (Netscape or Mozilla)

That makes it difficult to capture error messages from the child for use 
in the parent (at least so it seems to me after a first look). In some 
other instances, I think one can deal more easily with this issue.
When I find some time, I will try and concoct a solution, but I can't 
promise that I come up with something soon. If you have an idea, feel 
free to tell me.
> P.S. The e-mail address you give in the about box of texdoctk is
> invalid
If it's the dlc.ku.dk one, yes - my former host instution has ceased to 
exist. In the most recent release (0.6.2) which I sent to Thomas Esser 
some months ago, my current address is given.
Regards,
Thomas
-- 
------------------------------------------------------------------------
Thomas Ruedas
Geological Institute, University of Copenhagen
------------------------------------------------------------------------


--- End Message ---

-- 
Frank Küster
Inst. f. Biochemie der Univ. Zürich
Debian Developer

Reply to: