[Bioperl-l] Q for Ewan or ??

Mark Wilkinson mwilkinson@gene.pbi.nrc.ca
Thu, 01 Mar 2001 10:05:32 -0600


> We can provide tarballs at the same time as the 0.7 and a CPAN bundle if
> you like.  I think we weren't sure if you wanted to be included in this
> release.

Our main concern is that, particularly for Windows users, setting up CVS and SSH2 and
all the rest is not particularly easy.  Granted, most Windows users will not have Tk
installed either... but then, neither does Hilmar so what does that tell you  8o)
Still, the installation of Perl and Tk is much less problematic than the Windows
installation of CVS - really, Windows CVS is a right royal pain!  In addition, I
suspect that very few people who are not in the bioperl developers list are even
*aware* that there is a CVS branch called bioperl-gui...

Our desire is to make SeqCanvas as visible/accessible as possible at minimum pain.
The easiest way to achieve this would be for it to be a part of BioPerl itself, but I
understand the reluctance to add such a huge and otherwise unnecessary dependency on
Tk.  I would argue, on the other hand, that it is much "closer" to bio-perl than is,
for example, bio-corba, or bio-java... it was designed to be a bio-perl Sequence
object viewer, and in that regard I do feel that it "belongs" in bioperl.  We are
probably a bit biased v.v. the importance of this module because we use it every hour
of every day to visualize our gene calling pipeline output, but for those of you who
prefer the good old days of yellow on black ascii art I can understand that the
module would seem somewhat superfluous ;-)  ;-)

w.r.t. stability and 'readiness' to be released - we have been running fairly regular
tests on it to ensure that it is compatible with the ever-changing 0.7 landscape and
it seems to be standing tall.  The code changes I have made recently are more or less
cosmetic...  It isn't *fully* 0.7 compliant because it does not yet deal with
fuzziness, but at least it doesn't break...  so I don't think we have any deep
concerns in that regard.  That said, I agree that it has not gone through the
rigorous testing that the rest of BioPerl endures - we only know that it runs on
RedHat 6.2, Win95/98/2000, and even then it is always us testing it so I don't know
how it would fare in someone else's hands....

Anyway, after that long winded monologue I guess the short and sweet of it would be
to say that we would be pleased and appreciative if you could distribute it by any
other method than anonymous CVS, simply due to the barriers that this poses to many
potential users.

cheers all!

Mark

P.S.  N.B. that this entire message is *only* about SeqCanvas.pm, not about the other
GUI modules which are, in some cases, written by other groups/individuals.

--
---
Dr. Mark Wilkinson
Bioinformatics Group
National Research Council of Canada
Plant Biotechnology Institute
110 Gymnasium Place
Saskatoon, SK
Canada