The following issues have been reported with this version of PDL:


- The default TriD controls are based on having at least
  a two button mouse.  For Macs with one button, you can use
  Ctrl+Click to generate the Right-click.  It is also possible
  to configure the Macbook tracpad to generate a secondary
  click---see the control panel.


- VRML output driver does not work for PDL::Graphics::TriD


- FFTW 3.2.2 is the current release version of the FFTW library
  but is not yet supported by PDL::FFTW.  You may install version
  2.1.5 of FFTW which is supported, fall back to using the PDL's
  built-in fft routines (PDL::FFT), or submit a patch to update
  PDL::FFTW to the version 3 API. 


- Various make problems with build:

  - Using parallel make causes the build to fail so don't use -j # options.

  - Make fails with message like

      "*** No rule to make target `pptemplate', needed by `pptemplate.pod'.  Stop."

      If you have this error, please contact the PDL developers via the perldl
      mailing list and help us track this one down.  Thanks!


- Extremely verbose but not particularly helpful text output
  from configure, built, test process.


- Lvalue subroutine errors have been reported for perl 5.6.2
  installation on a linux 2.4 kernel system.  Perl versions 5.8
  and above are *really* recommended.


- PDL::IO::Dumper has test failures on some systems that appear
  to result from an incorrectly detected/use flavor of uuencode
  or uudecode.  Installing Convert::UU should resolve the problem.


- PDL configuration failures have come about because needed temp
  files can not be created.  This appears to be a result of platform
  misconfiguration of TEMPDIR and not PDL per se.


- 16bit image file IO fails in t/pic_16bit.t apparently due to
  an older version of NetPBM which does not support the raw 16bit
  formats.  If this test fails, try upgrading your NetPBM library.


- Documentation for the various convolution routines are unclear.
  The implementation assumes that the kernel argument is flipped
  and then calculates the correlation of that with the image.  This
  makes no difference for symmetric kernels but can give unexpected
  results with non-symmetric ones or if you are trying to do your
  own correlation.


- fftconvolve does not give the same results as the brute force
  summation calculation, see discussion for SF bug # 2630369.


- The following SourceForge bugs are outstanding at time of the
  PDL-2.4.5 release:

  2630369    PDL::FFT::fftconvolve - bad scaling?  
  2524068    Build breaks for PDL-2.4.4/Lib/Minuit (code not ISO C)  
  2339818    need lib64 searches in Graphics::TriD  
  2153898    default PDL build output too long
  1994610    TriD build problem w pthreads
  1994584    PDL Has no signed 8-bit integer datatype
  1994442    imagrgb problems
  1582350    badflag called in PDL::Core and PDL::Complex
  1505171    failure in MatrixOps::eigens in CVS
  1435189    Installation with cpan2rpm
  1205359    PGPLOT Window does not store full state info
  1059020    Problem with FFTW version 3.0.1


For more information on these and other PDL issues, and for
submissions of patches (bug patches are always welcome!), 
see the PDL mailing lists. Links to archive list discussions
and how to register for the mailing lists can be found at
http://pdl.sourceforge.net/maillists/ .
