Alex,
Before I give you my "priority list", let me tell you that I am a
retired biochemist/molecular biologist, enjoying my retirement years as
an adjunct professor in Chemistry at Georgia State University -- a sort
of hobby scientist, trying to remediate a bioinformatics approach I was
too old to participate in from scratch.
That said, I'll keep my list of packages I would like to see updated to
10.8.x to the "sci" category.
1) "clustawl" and/or "tcoffee" are very widely used and indispensible
sequence alignment programs. Hard to judge which is best; clustawl is
more widely used.
2) The "ncbitools", particularly the "blash" modules are workhorses of
molecular biology database searching. There may be alternative
approaches to getting NCBI service on Mac 10.8 -- I'm looking at the
possibiities.
3) In my active, working years, "emboss" (we ran in on unix boxes) was
our every0day, comprehensive sequence analyis package. I personally
don't use it these days -- but it remains prominant.
4) "hmmer" is a mature and widely used, hidden Markov approach
characterizing protein (that's how I use it) sequence characteristics,
with which one can then search the databases for proteins with similar
characteristics.
5) "phylip" is perhaps a little out-moded in the field, but the only
representive of phylogenetic inference software in the fink package
collection.
6) I personally am a fan of the "snna" neural net software. It is
flexible and comprehensive; perhaps not much updating since it's
introduction in the '90's, but (I find) more accessible than, say, the
NN modules of Octave (or MATLAB)
Even if some or all of them were ported as "unstable", I'd be happy to
give them a spin and report my experiences.
Bob Wohlhueter
On 5/23/13 9:48 PM, Alexander Hansen wrote:
Glad you're up and running again.
We didn't move packages over for 10.7 without testing them, since we
thought it would better to know that everything works. (Though now
there may be packages that work on 10.7 but not 10.8)
Let us know if there any packages that you're missing that you'd like
to have, and there's a reasonable shot that we might be able to do
something fairly quickly. A fair number of packages haven't made it
over yet simply because their maintainers aren't active and nobody has
requested them.
--Alex
On 5/23/13 6:15 PM, Robert Wohlhueter wrote:
Dr. Hansen,
Thanks for the good pointers. The cook-book you pointed to serveed well,
and I'm up-to-date and running.
The reinstallation was a little bumpy, in that the several previously
installed packages which no longer exist required a manual intervention
to get beyond the offending package. But such a house-cleaning was in
order.
The whole process reminds me of what a great resource fink is.
Bob Wohlhueter
On 5/22/13 5:55 PM, Alexander Hansen wrote:
On 5/22/13 2:44 PM, Robert Wohlhueter wrote:
I've used fink for years under OS 10.6 (but still consider myself a
"beginner") When I upgraded recently to 10.8.3, my fink installation
was broken. Consulting various upgrade-recipes on the web, I
downloaded
fink-0.34.7.tar.gz, and bootstrapped it(without removing any of my
previous installation in /sw...) The process worked, but insisted on
putting the new, bootstrapped installation in /sw2
I knuckled under to that demand, thinking I would simply be able to
merge my pre-existing database (of applications installed with fink)
with the new program installation. Well, that was naive! I can't
describe exactly what I did, but obviously the attempted merger had
its
shortcomings.
The new fink software itself works, and commands like `fink list -i`
indeed recognize my pre-existing packages. But several commands, for
example `fink selfupdate` culminate in opague messages like:
sent 93023 bytes received 4670742 bytes 142201.94 bytes/sec
total size is 11592648 speedup is 2.43
Scanning package description files
Failed: not a reference at /sw/lib/perl5/Fink/Services.pm line
1880
The syntax of Se vices.pm, and why perl5 in particular is "not a
reference" is not at all clear to me. (command line perl works, and
shows version 5.12.)
That message comes untrapped right out of Perl. The error looks
vaguely familiar, and I'm guessing it's due to the fact that Storable
databases are not compatible between Perl versions.
I'm resolving to starting from scratch, but still see no way around
the
problem of having to bootstrap a new fink into a unique directory
tree,thus losing my previous database info. Any pointers to the
proper
way of doing this?
Thanks,
Bob Wohlhueter
http://www.finkproject.org/download/10.8-upgrade.php?phpLang=en
I just noticed that this needs some fixes and updates, but I believe
the meat of it is accurate.
------------------------------------------------------------------------------
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_may
_______________________________________________
Fink-beginners mailing list
Fink-beginners@lists.sourceforge.net
List archive:
http://news.gmane.org/gmane.os.apple.fink.beginners
Subscription management:
https://lists.sourceforge.net/lists/listinfo/fink-beginners
