Back (for now)

Georg Fuellen fuellen@dali.Mathematik.Uni-Bielefeld.DE
Fri, 27 Jun 1997 10:53:23 +0000 (GMT)

Steve wrote,
> Hi Folks.
>   Georg: Thanks for the quick, useful reply!
> On Thu, 26 Jun 1997, Georg Fuellen wrote:
> >   [ Note, I'm on vacation July 5 -> approx July 25. ]
> Ok -- that corresponds with most of the time I'm away.
> > >   Congratulations on the great job with the OiB electronic poster.  I
> > > think you guys did a great job of putting that together.
> > 
> > thanks -- SteveC and Chris deserve most of the credit !! 
> > How about basing a Perl Journal article on it ??
> Great idea! How should we arrange this?  Maybe SteveC and I could
> coordinate this once I get to Stanford, with more input from you guys?  I

That would be good.

> contacted Jon Orwant some time ago, and he indicated a willingness to have
> a bioperl article at pretty much any time.  I think this should be
> published after the CPAN release. 
> > My interest will probably remain limited to sequences, at most I expect to deal
> > w/ RNA structures in my PhD thesis work, and I'm not sure generic code 
> > concerned w/ both RNA _and_ protein structures is a useful thing to have, or? 
> > I'm mentioning this b/c as soon as you guys start working on modules for
> > structures, you may shortly consider RNA structures, too, and maybe find
> > out that I'm too pessimistic ?!
> I think that the protein 3D structure module might reasonably include RNA
> (and DNA)  tertiary structures.  However, I think RNA secondary structure
> is very different from protein 3D or secondary structure.  Moreover, I
> don't see any real advantage of conflating RNA secondary structure with
> proteins.

ok - agreed.

> > > 3)  Releasing these darn things finally!!
> > > 
> > > Background info: the bioperl list at mole is now gone; the sysadmin
> > > removed it because of junk mail it was attracting. 
> > 
> > Not so nice. Given this, what's the best way to avoid manual subscription 
> > of the bioperl@mole subscribers to our vsns-bcd-perl-announce list ??
> Um, in your announcement message, you said that everyone had already been
> put on vsns-bcd-perl-accounce and just gave instructions for removal.  So,
> I don't see any change here.  I'm confused by the problems you suggest.

You're correct, I didn't remember that in Feb/Mar, I agreed that I'll 
be handling removals manually. I'll have everything set up after my 
vacation, ok? (I gather that the announcement will have to wait until August 
anyway. I have not yet put ppl on vsns-bcd-perl-announce b/c majordomo then
sends them welcome messages which are just stupid and noisy if there's no 
real announcement following up on them. We've already got a kinda bad 
reputation for talking much about releasing ``really soon now'' and not much
has been delivered. Maybe we've got our standards too high ?! ;-)

> > I won't start reworking the announcement texts until the code is submitted
> > to CPAN -- that was a (small) waste of time doing it in Mar/Apr, I should 
> > have known better ;-) 
> :)
> > I suggest that as soon as you're at Stanford, you bundle things together,
> > incl. makefiles. Maybe SteveC will help, etc. Just today I realized that
> > I'll probably soon need to replace some functions by C code, so I'd be happy
> > if the make process would be designed w/ this in mind - if that's possible 
> > at all. 
> I am tentatively happy with this, but am concerned that it could be a
> major delay. I am just at Stanford for 1 month before I disappear again

Either bundling/getting ``make'' to work is as simple as you said (and then,
there's no major delay), or it's more complicated, and then I'm not the
right person anyway (I'm really feeling lost & lack time)...

> (for a few weeks), and during that time I'm attending two conferences and
> need to make up my presentations for both of them!  Plus, I need to get a
> new apartment, move in.  
> I would not recommend using any C code at this point, as it is likely to
> delay things further (though making the module designed to permit it at
> the future is good idea). 

Sorry about my bad wording... I don't wanna ``replace some functions by C code'',
instead I'd like to provide access to a C-implemention as an alternative, 
mostly for the closures. (The closure setup returns a pointer to an ANON 
subroutine, and I hope that it's easy to alternatively return a pointer to
a C function. To get an idea on what I'm talking about, take a look at the 
current UnivAln module (search for the word ``Using closures;'' in
Hints/help would be warmly welcome).

> It would be best if, for example, UnivAln and PreSeq could be set up so as
> to build out-of-the-box.  If this means removing some features, I think

I think it doesn't -- but take a look.

> that is better than requiring manual editing.  (You could require manual
> editing to reinstate teh features.)  Then, if you can get those to me by
> August 1, I promise to make a major effort to get the items on CPAN as
> quickly as possible once I get to Stanford.

I can't, no time :-/

> > > 4)  A bioperl meeting
> > First of all, let's hear Chris on the issue -- any news on G.I.'s side ?
> > Perl for G.I.'s "2nd generation" code ?? - would be great !!
> > 
> > I cannot travel to the US without the additional justification of an
> > International bioinformatics conference where I can present my PhD research
> > stuff -- I've heard that there's gonna be another OiB meeting next year 
> > in the US, coupled w/ ISMB 1998, it may be good to meet berore/during/after.
> When/Where is ISMB 1998?  I have been scouring the web for this, without
> any luck whatsoever. 

ISMB 1998 is in Montreal, approx in June. There's also RECOMB '98,
which may be scheduled on a rather short notice for early 1998 --
I'm currently trying to find out.

> > Other outstanding issues on my list:
> > Link to WWW-talk/poster from homepage (SteveC: is that ok, or did you do 
> > last-minute changes that are not yet on the WWW ?)
> Ok with me.
> > My big prob with Bio::PreSeq/Bio::Parse is correctness: how do you verify 
> > it's doing every conversion ok? test suite? regression w/ ReadSeq?
> I assume that it at least works most of the time!  Unless we're confident
> that this works perfectly, why don't we list it as a 'use at your own
> risk'.  We should make up a test suite, but I think that is unecessary for
> the release of 'PreSeq'
> > Should we have clone() or copy() for the function
> > that returns a clone/copy of the object ?
> > Maybe one is non-recursive (shallow), but the other is ``deep''?!
> I liked your (Georg's) idea:  clone() is deep, copy() is shallow.  

copy() is deep right now (in Bio::UnivAln), and I think that's more 
intuitive... ?!

> > (Which lab? I know quite some ppl there..)
> I'll be at the Bioinformatics Center, National University of Singapore...
> Best regards,
>   Steve

best wishes,