[Biocorba-l] idl 0.2 discussion

[Matthew Pocock]

Hi.

I have to confess that I haven't started to write any of this code yet - I was
hoping that somebody else would feel inspired, but mabey nobody cares. Sniff.

Ewan Birney wrote:

> On Fri, 17 Nov 2000, Jason Stajich wrote:
>
> > Well, I haven't seen a tremendous amount of discussion on the language
> > lists or bioxml.  I suspect that much of the CORBA stuff doesn't apply to
> > enough people at this point to get heated debates, or we all like it so
> > much there is nothing to say.  I'm not sure when Alan gets back, but I'd
> > like to put a date on finishing discussion on the proposal and moving
> > towards freezing the idl as version 0.2.
>
> Yeah. My only dispute now is that I do think that maxlenght should be on
> the AnonymousSeq interface. Although alan is right that it is a property
> of the server/database, this does not help a client which has a subroutine
> that is declared to take an anonymousseq. This subroutine has to know how
> much it can yank out of the server at any one time.
>

I feel uneasy about this whole maximums issue. But, if some servers are going to
be limitited to string lengths that are below the lengths representable by corba
longs then I guess we do need a maxlength field, and it should be on
AnonymousSeq. How big are corba longs? Can we think of any servers that would be
unable to retrieve chunks this size (i.e. where maxlength < MAX_LONG)?

Matthew