On 2008-12-24 11:57-0500 Hazen Babcock wrote:
>
> Hello,
>
> Due a number of significant changes it has been requested that the 5.9.2
> release of PLplot should happen in a few weeks time. Tentatively I'm
> going suggest the weekend of January 17-18th. Is this acceptable?
That would be great!
There is an exclamation point there because I am the one that originally
requested a quick release. There are a rather large number of bugs that
have been fixed since 5.9.1. Furthermore, we now have much more
comprehensive testing (almost all our public API is now tested) and much
more convenient testing procedures. If the developers here run ctest in the
build tree and "make test" and "make test_interactive" in the install tree
on their various platforms as a matter of routine, that should generate a
lot of confidence in the cross-platform quality of this next release.
Here is the current compare test status report (compiled from "make -j3 test"
in the install tree). I have excluded c++, f77, f95, and ocaml because they
are perfect (which is great to see):
java
Missing examples : 19
Differing postscript output :
Missing stdout :
Differing stdout :
octave
Missing examples : 19
Differing postscript output :
Missing stdout :
Differing stdout :
python
Missing examples : 19 21
Differing postscript output : 20 27
Missing stdout :
Differing stdout :
tcl
Missing examples : 19
Differing postscript output : 11 13 15 16 20
Missing stdout :
Differing stdout :
perl
Missing examples :
Differing postscript output :
Missing stdout :
Differing stdout : 31
ada
Missing examples : 31
Differing postscript output : 15 29
Missing stdout :
Differing stdout : 23
adathick
Missing examples : 31
Differing postscript output : 15 29
Missing stdout :
Differing stdout : 23
It would be nice but not essential to get some of these remaining
problems fixed before the release.
I will deal with the python issues outside example 19. The python example 21
missing issue is because I am testing with Numeric (HAVE_NUMPY=OFF) rather
than numpy. I don't think the Numeric interface to PLplot change that would
be required to fix example 21 is worth worrying about because that fix would
be time consuming, and we are not going to be supporting Numeric
indefinitely. I don't know what the example 20 and 27 issues are with
Numeric. (IIRC those examples are fine under numpy.) If the issues are easy
to fix I will do so, otherwise I will let is slide like example 21.
I would also be willing to take responsibility on the PLplot side of things
for the perl fix, but Doug Hunt needs to update the external
PDL::Graphics::PLplot module first as I have discussed with him off list.
Of the many recently discovered bugs, the only ones left that I am aware of
are the Tk example segfaults and fixing the visibility issues for
plplotcanvas.h in a way that does not interfere with pygcw. Bugfixes for
these issues would be _really_ nice, but in my view the release should just
go ahead if nobody gets to these issues in time.
Alan
__________________________
Alan W. Irwin
Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).
Programming affiliations with the FreeEOS equation-of-state implementation
for stellar interiors (freeeos.sf.net); PLplot scientific plotting software
package (plplot.org); the libLASi project (unifont.org/lasi); the Loads of
Linux Links project (loll.sf.net); and the Linux Brochure Project
(lbproject.sf.net).
__________________________
Linux-powered Science
__________________________
|