Bioperl Strategy -- Some Thoughts
Tue, 21 Oct 1997 11:32:20 +0000 (GMT)
>From SteveC's mail,
> > This would definitely be worthwile !!! I feel we should stress the
> > importance of Perl for _fast_prototyping_, which (to me) is the niche
> > we best fit in. (Ironically, a bunch of overworked volunteers
> > makes a very _slow_ development team -- we need to change this!
> > Could the development of Bioperl modules in the OMG context
> > be something for which we could obtain funding ?)
> Involvement with the OMG effort would certainly add strength to our
> effort. I wonder, though, in the wake of the Java craze, if Perl work
I remember the C++ craze years ago -- today C++ can offer far less
advantages than Perl if it's compared to Java. If we have the right
strategy, we can prosper no matter whether Java stays, or fades away
in the wake of a new craze.
I feel that our best chance is to join in the Java craze as Java's
fast-prototyping scripting-language companion which is much better
than Tcl. One problem is that Ousterhout promotes Tcl at Sun, so
we better get going soon. If both our CORBA ambitions and Larry Wall's
Perl->Java ``compiler'' materialize, and the Perl compiler is released
soon, we'd be top.
> would be viewed as mis-directed. I personally don't think so but it
> could be countered, "Why not just use Java?". As you state, we need to
> emphasize Perl's niche and huge following.
> We should also be prepared for anti-Perl bias. I recently heard a
> comment along the lines of "With Perl, it is too easy to write code that
> you (or someone else) won't understand in a year". Such a comment could
> actually support Bioperl since it indicates a need for standard modules.
> A key question is, from whom would you ask for Bioperl money? (Whoever
> has alot invested in Perl would be a good start).
We should approach the Perl Institute to get their backing, and limited
funds as an endorsement. Then, we need to decide whether we want to
approach public funders, and/or companies. In this stage, we need
to get the backing of those we're working for -- ask them what to do.
You and SteveC *MUST* get together soon and devise a strategy paper
draft on the future of Bioperl -- if we can't get rid of the
sluggishness that we've had for the last 2 years, no big dreams will
ever come true.
> > If Bioperl adheres to the OMG standards, rewriting selected
> > components to Java and C should be painless, correct ?
> > If so, that would be a heaven for fast prototyping :-)
> Good point.
Even more, compilation from Perl to C and Java is promised by
Larry Wall et al, they will hopefully deliver soon !
> Regarding the RFI:
> > A 1-page response is a must-do, I think.
> Yes. They indicate that some submissions may provoke a request for
> further information, which we should be prepared for.