Re: Biology packages in FreeBSD
Hi!
> > fasta, fasta3,
> Here we have probably to do some further investigation: We have the blast2
> package which also seems to contain a version of fasta. Could anybody more
> educated than me please comment on this?
>From my emails changed with upstream author William R. Pearson <wrp@virginia.edu> (this is what he said):
[EMAIL 1]
I am nervous about distributing FASTA with a Linux distribution, because
there is always a chance that the version of FASTA that is distributed
would have a bug, which, by the time someone installed it from the
distribution, would probably be fixed.
I am more comfortable including a script that would automatically
download
the current version from ftp.virginia.edu/pub/fasta/fasta3.shar.Z,
and compile it.
[EMAIL 2]:
It did not occur to me that you would only want to distribute binaries,
since they are not very useful without DNA and/or protein sequence databases
(which are considerably larger). Are you planning on a distribution
that includes sequence databases? I can certainly see an advantage to a
distribution that has pre-compiled versions of the programs (possibly pre-configured
to recognize the databases).
[THE LAST ONE]:
I do not think there is much reason to distribute only my program,
because compiling my program is the smallest of the several steps required to
having a working FASTA system. Downloading the databases and editing
configuration files to find the databases is much more time consuming
and confusing.
So, I would prefer that FASTA not be included in your distribution,
though I encourage you to include a script that would allow people to
download and compile FASTA.
[END]
"your distribution" = Debian
So, I think that fasta is out of Debian, right?
> > gff2ps,
(...)
> Here the reasons are given in the mails of the thread above: It is just
> duplicated functionality with a worse implementation.
As Steffen Moeller said: "BioPerl does not offer the example as a ready-to-use application."
Bioperl's version is different from the one that I cited.
And I found the error on gff2ps. It is something with locales. Using LC_ALL=pt_BR, it fails to build the PS file. LC_ALL=C works OK. I have contacted upstream author and said about that.
Also, from TODO list of gff2ps: "Full re-write in Perl..."
I can release a gff2ps version, using a wrapper with something like
/usr/bin/gff2ps =
#!/bin/sh
LC_ALL=C /usr/share/gff2ps/gff2ps.real
Or something like that. This is a workaround until upstream releases a new version. Or we could wait to the new version, without releasing a version before.
> > gperiodic, grappa, kinemage, lamarc,
> > libgenome, lsysexp, migrate, nab,
> Missing.
Gperiodic is on Debian! :-)
> Matt Hope, who is an official Debian developer maintains a private archive
> of biological software. I have heard that the reason for this is that he
> has not enough time to bring the packages into a shape which is fit for
> official Debian (Matt, please correct me if I'm wrong). My idea would be
> to start group maintainance of these packages. If somebody would be able
> to offer Matt the help he needs we might quickly be able to move all his
> packages to the Debian mirror. I would also spend some time on some of these
> packages (perhaps tagging myself as "Uploader" for the packages).
If there is something that I could help, let me know, please.
Cheers,
Nelson A. de Oliveira
__________________________________________________________________________
UOL Fone: Fale com o Brasil e o Mundo com até 90% de economia.
http://www.uol.com.br/fone
Reply to: