[Bioperl-l] All done with Bio::DB::Query

[Lincoln Stein]

I'm not sure one should ever remove documentation.  Otherwise people reading 
through old code will have trouble when they hit the removed method.  Isn't 
it better to leave the documentation there with a note that the method has 
been superseded by something else?

Lincoln

On Monday 25 November 2002 02:14 am, Hilmar Lapp wrote:
> I was just wondering why we need to deprecate it if it works nicely
> using the new code. Also, if you really want to deprecate it, it
> should become undocumented. My suggestion is to not deprecate it
> (for 1.2), but remove documentation. This way new users won't pick
> it up easily by accident, but old clients continue to run
> undisturbed.
>
> As for module name clash, there is none yet. There might just be one
> in the future, but frankly I don't think it's likely to happen. The
> only thing that lets me favor a name change is that Bio::DB::Query
> in bioperl-db is about object-relational queries, whereas
> Bio::DB::Query in bioperl-live now is about (sequence-)databank
> queries. Might be clearer for users who have both installed to have
> them end up in different namespaces.
>
> 	-hilmar
>
> On Sunday, November 24, 2002, at 05:11 PM, Lincoln Stein wrote:
> > get_stream_by_batch does use the new code; the method still works
> > and is
> > documented, but is marked as deprecated.  get_Stream_by_batch() was
> > originally an optimized form of get_Stream_by_id().  Now that it
> > doesn't
> > contribute any performance improvement, I think we should deprecate it.
> >
> > Sorry about the module name clash.  Perhaps we should rename as
> > Bio::DB::SimpleQuery?
> >
> > Lincoln
> >
> > On Sunday 24 November 2002 04:46 pm, Hilmar Lapp wrote:
> >> Cool Lincoln.
> >>
> >> Can't get_stream_by_batch rewritten to use the new code? Sorry if
> >> the answer is obvious.
> >>
> >> I should have said that Bio::DB::Query is the namespace that I also
> >> use in bioperl-db for object-relational query modules. There's no
> >> clash yet of module names, and the chance for one is probably fairly
> >> low; I just thought I mention this here. If you or anyone else feels
> >> we need to better protect from possible future clashes, we need to
> >> rename either in bioperl-live or bioperl-db.
> >>
> >> 	-hilmar
> >>
> >> On Friday, November 22, 2002, at 08:24 AM, Lincoln Stein wrote:
> >>> Hi,
> >>>
> >>> I'm all done with the following modules:
> >>>
> >>> 	Bio::DB::QueryI
> >>> 	Bio::DB::Query::WebDBQuery  # web utilities
> >>> 	Bio::DB::Query::GenBank   # the only "real" one
> >>>
> >>> I have tested on a variety of Unix systems, but I need a Windows
> >>> tester.  I
> >>> used fork() in a couple of places in order to make SeqIO streaming
> >>> pipelineable (otherwise it has to download the whole set of records
> >>> before it
> >>> can start processing the first one).  The code should detect that
> >>> it is not
> >>> on a platform that supports fork-and-pipe and automatically fall
> >>> back to the
> >>> old way of doing things, but I'm not 100% certain it will work as
> >>> advertised.
> >>>
> >>> I've also done a touch up here and there on the Bio::DB:: modules,
> >>> mostly to
> >>> deprecate Aaron Mackey's get_Stream_by_batch() methods, which are
> >>> superseded
> >>> by the eutils interface.
> >>>
> >>> Let us all be glad that NCBI has committed to maintain eutils as
> >>> its official
> >>> interface.
> >>>
> >>> Lincoln
> >>>
> >>> --
> >>> Lincoln Stein
> >>> lstein@cshl.org
> >>>
> >>> _______________________________________________
> >>> Bioperl-l mailing list
> >>> Bioperl-l@bioperl.org
> >>> http://bioperl.org/mailman/listinfo/bioperl-l
> >
> > --
> > Lincoln Stein
> > lstein@cshl.org

-- 
Lincoln Stein
lstein@cshl.org