Re: [Factor-talk] Graphviz unit tests

2011-09-18 Thread John Benediktsson 

 - The current tests pass on my machine.  But then, my machine = a Debian
 box
  with Graphviz 2.26.3 installed (since I guess 2.28 is too bleeding-edge
 for
  Debian) and erg's plugin-list fix reverted (which I'm certainly thankful
 for,
  but 2.26 doesn't have the gvPluginList function; ugh).  So, grain of salt
 I
  guess.


I get a kernel-error trying to run the first smoke-test (5 K_n), with
Graphviz 2.28.0 on Mac OS X 10.7.1 x86_64.  And running graphviz test
causes Factor to quit unexpectedly... which might be a different bug?
--
BlackBerryreg; DevCon Americas, Oct. 18-20, San Francisco, CA
Learn about the latest advances in developing for the 
BlackBerryreg; mobile platform with sessions, labs  more.
See new tools and technologies. Register for BlackBerryreg; DevCon today!
http://p.sf.net/sfu/rim-devcon-copy1 ___
Factor-talk mailing list
Factor-talk@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/factor-talk

Re: [Factor-talk] Graphviz unit tests

2011-09-18 Thread Alexander James Vondrak 
 I get a kernel-error trying to run the first smoke-test (5 K_n), with
 Graphviz 2.28.0 on Mac OS X 10.7.1 x86_64.  And running graphviz test
 causes Factor to quit unexpectedly... which might be a different bug?

I hate Graphviz.  I really, really do.

I'm not sure about the kernel-error, but if the unexpected quitting is anything
like previous issues I've come across, their source code habitually calls
exit(), which kills the UI listener since they share the same process (I
experimented once with forking the FFI calls, but that's a no-go on Windows,
so).  If you -run=listener to test, you might be able to see if a message gets
printed by Graphviz to stderr.  Maybe that'll give us more info?

(If it's further similar to other problems I've had with Graphviz, the error
message will be a single line buried in some undocumented corner of some messy
source code that hasn't been discussed anywhere that's indexed by Google.  Joy.
I hope for Graphviz's sake that I'm just really incompetent at reading C code.)

Thanks for testing,
--Alex Vondrak
--
BlackBerryreg; DevCon Americas, Oct. 18-20, San Francisco, CA
Learn about the latest advances in developing for the 
BlackBerryreg; mobile platform with sessions, labs  more.
See new tools and technologies. Register for BlackBerryreg; DevCon today!
http://p.sf.net/sfu/rim-devcon-copy1 
___
Factor-talk mailing list
Factor-talk@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/factor-talk

Re: [Factor-talk] Graphviz unit tests

2011-09-18 Thread John Benediktsson 
Unfortunately, this is all I get from the command-line.  I'll need to do
more debugging at some point:

( scratchpad ) USE: graphviz
( scratchpad ) graphviz test
Unit Test: { { t } [ 5 K_n smoke-test ] }
factor(53761,0x7fff7512c960) malloc: *** error for object
0xc000: pointer being freed was not allocated
*** set a breakpoint in malloc_error_break to debug
Abort trap: 6



On Sun, Sep 18, 2011 at 7:43 PM, Alexander James Vondrak 
ajvond...@csupomona.edu wrote:

  I get a kernel-error trying to run the first smoke-test (5 K_n), with
  Graphviz 2.28.0 on Mac OS X 10.7.1 x86_64.  And running graphviz test
  causes Factor to quit unexpectedly... which might be a different bug?

 I hate Graphviz.  I really, really do.

 I'm not sure about the kernel-error, but if the unexpected quitting is
 anything
 like previous issues I've come across, their source code habitually calls
 exit(), which kills the UI listener since they share the same process (I
 experimented once with forking the FFI calls, but that's a no-go on
 Windows,
 so).  If you -run=listener to test, you might be able to see if a message
 gets
 printed by Graphviz to stderr.  Maybe that'll give us more info?

 (If it's further similar to other problems I've had with Graphviz, the
 error
 message will be a single line buried in some undocumented corner of some
 messy
 source code that hasn't been discussed anywhere that's indexed by Google.
  Joy.
 I hope for Graphviz's sake that I'm just really incompetent at reading C
 code.)

 Thanks for testing,
 --Alex Vondrak

 --
 BlackBerryreg; DevCon Americas, Oct. 18-20, San Francisco, CA
 Learn about the latest advances in developing for the
 BlackBerryreg; mobile platform with sessions, labs  more.
 See new tools and technologies. Register for BlackBerryreg; DevCon today!
 http://p.sf.net/sfu/rim-devcon-copy1
 ___
 Factor-talk mailing list
 Factor-talk@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/factor-talk

--
BlackBerryreg; DevCon Americas, Oct. 18-20, San Francisco, CA
Learn about the latest advances in developing for the 
BlackBerryreg; mobile platform with sessions, labs  more.
See new tools and technologies. Register for BlackBerryreg; DevCon today!
http://p.sf.net/sfu/rim-devcon-copy1 ___
Factor-talk mailing list
Factor-talk@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/factor-talk