[Bioperl-l] 1.1.1 schedule; SeqIO speed ups

[Ewan Birney]

1.1.1 release will happen probably next week, after the CVS move - Chris -
are we ready to move CVS yet? What is the schedule for that? (the
open-bio servers need to be reorganised somewhat urgently, and chris wants
to swap around the DNS entries for them)



I have been working on SeqIO speed ups, concentrating mainly on fasta
parsing, but always trying to generalise the speedups to effect as many
modules as possible.


It is in fact pretty tricky - somewhere in the object creation we double
the time taken, but it is a mixture of things such as:


   - using _rearrange
   - chaining new functions
   - delegating new off into PrimarySeq meaning for each Seq object we
create, we have two object creation traversals to do
   - passing the sequence on the stack, which, as it can be quite large
itself has overheads - passing a reference is faster


I am trying to sort out a more complete system for all of this, and will
report back I hope by the end of the week.




ewan