[Bioperl-l] 1.0.2

[Hilmar Lapp]

> -----Original Message-----
> From: Jason Stajich [mailto:jason@cgt.mc.duke.edu]
> Sent: Friday, July 12, 2002 4:03 PM
> To: Hilmar Lapp
> Cc: Bioperl
> Subject: RE: [Bioperl-l] 1.0.2
> 
> 
> 
> On Fri, 12 Jul 2002, Hilmar Lapp wrote:
> 
> > I'd not be so optimistic at this point. I'm afraid putting 
> bioperl-db
> > and biosql to more use may instigate API changes at some places. It
> > shouldn't be terrible though. Anyway, I'm just not sure yet 
> we should
> > close 1.0.x after 1.0.2.
> >
> > I'm testing the Location::each_Location fix and commit in a second.
> > Jason, do you have time to merge that to the branch (I'd have to
> > download the entire branch; let me know), or is each_Location not in
> > the branch anyway? If the latter is true, that means that bioperl-db
> > is now incompatible with the 1.0.x branch. Hm. Should we branch
> > bioperl-db then, too? Would be kind of difficult by now and involve
> > some back-porting ...
> >
> Can we just say the bioperl-db HAS to run bioperl 1.1+ or 
> does that just
> get us into trouble?
> 
> I think the back port is the second option rather than 
> branchin bioperl-db
> again and trying to have people remember to check in the fix in both
> places.  I guess adding a method is less API breaking than 
> removing one so
> back porting each_Location would be acceptable.  Are there 
> other things?
> 

Not at this time. But there's the biodatabase object and PrimarySeqI method addition pending, and dumping OMIM into BioSQL will probably cause some changes too until it really works.

I of course have no problem with keeping bioperl-db as Bioperl-1.1-dependent for the time being. I have no idea though where else it is used. (Any show hands here? Anyone's got a problem with bioperl-1.1 dependency for updates?)

Generally adding a method to an API doesn't break it in Perl (as opposed to Java). The problem is on one hand a matter of taste, on the other hand if someone builds an application assuming that added interface method is fine to rely on being implemented by every implementor of the interface, that may cause a real problem. Maybe this is paranoid, maybe it's not, I'm not sure. If you ask me for a vote, I'd be strict and not do it.

	-hilmar

> I can do it this weekend.
> 
> > Opinions?
> >
> > 	-hilmar
> >
> > > -----Original Message-----
> > > From: Jason Stajich [mailto:jason@cgt.mc.duke.edu]
> > > Sent: Friday, July 12, 2002 1:44 PM
> > > To: Bioperl
> > > Subject: [Bioperl-l] 1.0.2
> > >
> > >
> > > Okay folks, I'm going to prepare 1.0.2 release of the core
> > > this weekend
> > > and announce on Monday.
> > >
> > > Unless there is a really good reason, I'd like to close the
> > > branch after
> > > this release and have development only on the main trunk. 
>  1.0 and 1.2
> > > shouldn't be as incompatible as the 0.7 and 1.0 changes so
> > > hopefully it
> > > will be easy enough for systems to upgrade the toolkit
> > > without any changes
> > > to their code.
> > >
> > > If the bioperl pipeline and db teams want to report in, we
> > > probably need
> > > to look at where you are and if a 1.1 release in the next
> > > week or two is
> > > going to still work.  I do NOT think we will need to do 
> separate 1.0.x
> > > branch and 1.2 branch releases for the pipeline/db/run 
> pkgs if the API
> > > stays as stable as it has been.
> > >
> > >
> > > -jason
> > >
> > >
> > >
> > > _______________________________________________
> > > Bioperl-l mailing list
> > > Bioperl-l@bioperl.org
> > > http://bioperl.org/mailman/listinfo/bioperl-l
> > >
> >
> 
> -- 
> Jason Stajich
> Duke University
> jason at cgt.mc.duke.edu
> 
>