To Andrew, Phil, and Jim:

These starting three paragraph are new, but the rest of this post you
have seen before under a slightly different subject line.  But I am
including "the rest" of this post now because it appears plplot-devel
is working again, and I want everyone there to be aware of the changed
release date that is alluded to in the last paragraph.

To be more specific about that, the new tentative release date is
Wednesday, March 4th.  The reason for this delay is I haven't had a
chance to even start the preliminary steps in the
release process yet (including most importantly comprehensive testing) because 
of mail troubles
slowing down our communications and all the specific time-intensive
testing I have been doing to help you guys figure out some of the
remaining nastier wxwidgets/plbuf issues.

In other words specific testing that bugs have been fixed for a
particular build configuration is essential, but that has to be 
followed up with comprehensive testing to explore all the corner cases
such as our three major configurations, and our three different build
systems.  Thus, I plan to start the release preliminaries now
including comprehensive testing and if I have some early success with
complete comprehensive checks for the Debian stable case, then I will
start comprehensive testing on MinGW/MSYS/Wine and encourage others
here to also start comprehensive testing on their accessible platforms
following the directions in
http://sourceforge.net/p/plplot/wiki/Testing_PLplot/ ==> Comprehensive
testing. Of course, if our joint comprehensive testing reveals
additional issues that need to be fixed before the release, then the
release date might slide still further into the future. But I have
been quite happy with the bug fixing that has been going on for the
nastier wxwidgets/plbuf bugs we have encountered so far.  Therefore,
however long this release process takes is definitely going to be
worth it!

@Andrew:

I didn't receive your recent e-mail concerning rt configuration, but
spotted that message just now in the Plplot-devel archive, and I have
also discovered all sorts of other mail disruption I have at this time
with sending anything to SourceForge or with SourceForge sending
anything to me.

So it is very likely the release is going to be
delayed beyond February 28th because of my mail troubles with
SourceForge.  Note my ir...@beluga.phys.uvic.ca e-mail address appears
to work fine so if anybody here want to urgently reach me with regard
to some release issue, use that address rather than any
SourceForge-related one at this time.

Also, to be fair, I am pretty far behind on my testing so that may delay
the release for a few days even if SourceForge restores mail
connectivity for me.

I think I now understand why rt has to be explicitly mentioned for
the wxPLViewer build (as you discovered on Ubuntu).  As already
noted by Jim, nm --undefined shows that wxPLViewer explicitly
depends on

U shm_open@@GLIBC_2.2.5
U shm_unlink@@GLIBC_2.2.5

and <http://linux.die.net/man/3/shm_open> indicates that
on Linux they must be linked with -lrt.  However, this
explicit linking not needed with wxwidgets because it
does not call either of the above functions (again according
to nm).

However, ldd shows that wxwidgets and wxPLViewer are linked to
a library that itself is linked to rt on Debian stable which
is why the linking of wxPLViewer scraped through for me
on Debian stable.  But that indirect link "save" did not
apply for the Ubuntu libraries.  The correct solution is, as
always, to mention all libraries that are explicitly needed
so I have now made that fix (commit id a01a677).

@Phil:

Sorry for conflating two e-mails, but "needs must" when
I cannot communicate via SourceForge.

There are some recent regressions in wxwidgets.

1. I noticed while testing the above fix that there are now affine
transformation issues again for 3D plot labels for example 8 with -dev
wxwidgets.  The xwin, xcairo, and qtwidgets devices do not show this
issue even when they are resized.  (So I think that xwin experiment
rules out the problem is in plbuf.)

> From my and Andrew's recent superb experiences with git bisect, we
highly recommend that method of finding the offending commit
that caused the above issue.

2. There is now also a text rotation transformation issue for example
4 page 2 and example 26 page 2 for wxwidgets which was not there
recently.  You might not need git bisect for this one since I suspect
it is the fairly recent example 3 rotation fix (by the way example 3 still
looks fine) that caused the other issues.

The above are strong "would be nice's" for the release if you back
them up by looking at every example to make sure a fix for one example
page does not cause issues for other example pages (like happened to
create these latest two regressions).  Of course, if a change like
Andrew's recent fix to plbuf is necessary to fix all devices, then he
obviously that has to be done even if it causes problems for wxwidgets
(until the two wrongs making a right approach that worked before is
fixed).
But "git bisect" should tell the tale about the exact cause of
the two regressions above.

Another quite strong "would be nice" is to work out what is wrong with
the exotic glyphs for example 24.  Note there are only two missing
glyphs (on page 4) for example 23 and none for examples 18 (pages 6
and 8) and 26 (page 2).

To review, example 23 specifies the require glyphs with ascii escape
sequences unique to PLplot which the core PLplot library translates
into UCS4 (32-bit unsigned integer representation of unicode).  The
rest of the mentioned examples specify the glyphs with utf8 characters
which the core PLplot library also translates into UCS4. The drivers
then translate that UCS4 data into instructions to the various Qt4,
pango/cairo, and wxwidgets libraries using generic font families such
as "sans", "serif", and "mono", which those libraries translate into a
search of all sans, serif, or mono system fonts to find the "best"
system glyph to display that desired unicode character.  So since
wxwidgets appears to handle most unicode without issues, I am
virtually positive that that system glyph search process is breaking
down for the wxwidgets device for the exotic glyphs for example 24 and
the 2 missing glyphs for example 23.  Note, this did not happen for
the old wxwidgets device for the default (wxGC?) backend that did not
use the deprecated plfreetype method. (By the way, I am very happy to
see no mention of freetype in the rewritten wxwidgets driver, and no
mention of it in plbuf.)

Finally, another strong "would be nice" is to rewrite drivers/README.wxwidgets
which your work has completely outdated now (so it is positively
misleading).  And also please put a paragraph into README.Release
describing your recent wxwidgets work.

My own feeling is it will likely be at least mid-week next week
(March 4th) before I will be ready to release for the reasons
mentioned above, so this allows you extra time to do some of the above
before release.

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); the Time
Ephemerides project (timeephem.sf.net); PLplot scientific plotting
software package (plplot.sf.net); 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
__________________________

------------------------------------------------------------------------------
Dive into the World of Parallel Programming The Go Parallel Website, sponsored
by Intel and developed in partnership with Slashdot Media, is your hub for all
things parallel software development, from weekly thought leadership blogs to
news, videos, case studies, tutorials and more. Take a look and join the 
conversation now. http://goparallel.sourceforge.net/
_______________________________________________
Plplot-devel mailing list
Plplot-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/plplot-devel

Reply via email to