Final PDL-2.4.10 Countdown Status:

Here is the current consolidated list of items proposed/planned
for PDL-2.4.10 with the status.  My general feeling is that this
would be a good place to stop working on new functionality and
to focus instead on testing and verification of the various fixes
and features *and* on finishing the new book.

If you have any specific items that you feel should go into our
upcoming PDL-2.4.10 release, please come forward immediately.
Any remaining changes and additions in progress should be done
by Friday, 16-Dec, to give us time for testing, docs, book edits,
and release candidates.

--Chris

Key:

  D  Done
  P  Partially done
  L  Later

  *  A priority for 2.4.10 release
  -  To include if possible
  .  Nice to have (as resources permit)
 ' ' Not it plan (e.g., blank)

D Fixes for segfault bugs.

  All reported bugs with PDL crashing perl have been addressed.


D Update PDL-Graphics-PLplot for PDL core version.

  We need testers to verify that the updated code works.


P pthread computation support for win32 and cygwin

  Pthread support for auto-parallelization of threaded calculations
  works for all platforms.  Still needs a little auto-detection logic
  implemented and verified but all platforms are now supported as
  long as the correct options are given in perldl.conf for those
  cases where the detection fails.


P Complete NiceSlice2 reimplementation using Filter::Simple.

  A framework to select the NiceSlice2 implementation works and
  NiceSlice2 works as far as files go.  Still to do is the rework
  of the support for the interactive shells (e.g., PDL::NiceSlice2
  filtering of strings).  Given the potential risk of breakage for
  this last piece, I'm planning to push it out to PDL-2.4.11.


* Add "other PDL modules" documentation: gnuplot support, CUDA, Stats, Util,...

  This is one of the key things to wrap up in the final few weeks
  towards the PDL-2.4.10 release.


L Finish porting perldl compatibility to pdl2.

  Work continues for this task but finishing the pthread support
  and documentation and book for the PDl-2.4.10 release takes
  priority.


L I would like to see the annoying problem with debugging lvalue PDL
routines fixed.

  The problem appears to be turned into a warning in perl 5.14 or 5.15
  (I'm not sure which) but won't be able to disable the warnings from
  that until a developer with access to that perl version has time to
  investigate the correct placement of 'no warnings' in the code.


D Clear up documentation for PDL::IO::FlexRaw w.r.t.  badvalues,
headers, handles...
P Verify/document/test badval support for PDL::IO::FlexRaw

  The code appears to work and has been further reviewed.  It should
  be good for the PDL-2.4.10 release.


L Add rcsv/wcsv variants to PDL::IO::Misc (compare rasc with rcols)

  The difficult part of this is to ensure that options support
  works by combining any additional user options with the required
  { COLSEP => ',' } one.


D Add wcols support for 2D arrays to match rcols [] stuff.

  This has been verified.


* PDL book updated and on-line for 2.4.10

  Current status TBD (Matt??)  Finishing this update should be the
  top remaining priority for the PDL-2.4.10 release.


L Callback slice operation implementation

  Push to PDL-2.4.11.


L Finishing and releasing DM's new plotting module based on the Prima toolkit

  Push to PDL-2.4.11.


L Sprite-based points for PDL::Graphics::TriD

  TBD.


Sourceforge.net Bugs Closed:

D 3396761  fix rvals arithmetic wrap-around for byte types
D 3396738  PDL::Core::convert default return type
D 3391645  bad printf formats in pdlapi.c and others
D 3391507  pdl2 docs not found by pdldoc
D 3299611  FITS I/O obscure bug
D 3080505  PLplot segfaults on plshades call on OS X
D 3059083  Problems with FITS header handling
P 3018731  NiceSlice affects code where it should not


Sourceforge.net Bugs still active:

. 3363406  PDL::Minuit build can't make libminuit.a
. 3234141  max() fails on nan

  Neither of these bugs is what I consider critical for the
  PDL-2.4.10 release.  The libminuit build problem is being
  investigated and the max() bug for NaN inputs is just the
  existing PDL behavior---which should be cleaned up but is
  not usually a problem except in edge cases.


Other Sourceforge.net Bugs:

. 3316394  'help funname' fails to show multiple function names
. 3314712  make PDL "readme" type files available in POD
. 3314708  PDL install does not install scantree.pl
. 3301031  add multi-image support to PDL::IO::Pnm
. 3299615  PLplot still unusable with X
. 3293345  add mapflex/mapfraw support for win32
. 3283650  add IO by records to PDL::IO::FlexRaw
. 3229286  add user keyword and tags to help/apropos commands
. 3182279  port PDL::Doc, etc. to use Core Pod modules
. 3164794  enhance PGPLOT/pgplot buildability
. 3161459  online docs don't link to functions in other PDL modules
. 3074830  build-time (static) NiceSlice support for PDL installs
. 3064845  add plplot demos to match pgplot ones
. 2995500  perl -d chokes on lvalue functions
. 2989482  CPAN web site bug
. 2955733  better 2D photo/image viewing
. 2947663  CPAN-friendly .pd files

_______________________________________________
Perldl mailing list
[email protected]
http://mailman.jach.hawaii.edu/mailman/listinfo/perldl

Reply via email to