Re: [gnugo-devel] HTML regression views updates

[Gunnar Farnebäck]

Dave wrote:
> I can have a look if you want. I've just downloaded the latest
> sources, and was able to run regress.pl on just semeai.tst
> I'm now running it through on --all_batches
> 
> What is the nature of the crash ?  Is it the gnugo executable, or a
> script error, or does the perl interpreter itself crash and burn ?

You probably won't be able to reproduce it...
I found a lead in the output files and it looks like the problem is
that the way I start it (from cron via another script) it finds itself
in an environment without the TERM environment variable, whereby GNU Go
crashes when it tries to draw boards in the semeai traces.

Try this (tcsh terminology):
  unsetenv TERM
  interface/gnugo -l regression/games/semeai/semeai6.sgf --decide-semeai C1/E1 
-t

and see it crash with

  finished examine_position
  Analyzing semeai between C1 and E1, white moves first
  probable_eyes_a: 0000 eyemin: 0 eyemax: 0
  probable_eyes_b: 0000 eyemin: 0 eyemax: 0
  TERM environment variable not set.

I've modified my script to explicitly set TERM to something
reasonable, which will hopefully be enough. Whether it should crash
from a lack of TERM is a different question, which I don't know the
answer to.

> FWIW the rest of the gnugo version output is a bit wierdly formatted,
> with a very long line :
> 
> See http://www.gnu.org/software/gnugo/ or contact
>                         address@hidden for information about GNU Go. GNU Go 
> comes with NO WARRANTY to
> the extent permitted by law. This program is free software; you can
> 
> 
> Was it intentional to have the tabs after \n in the third line
> of the COPYRIGHT in main.c ?  (If it was, \t would be better than a
> literal tab I think)

No, those shouldn't be there. I don't know how they managed to get in
when I updated the copyright notices for 2006, but I've corrected it
in CVS now. Thanks for noticing.

/Gunnar