On Mon, Jun 26, 2006 at 01:01:28PM -0500, Eric Boutilier wrote:

> Danek Duvall wrote:
> >Stephen Hahn wrote:
> >>  My objection is a little bit about size, but mostly that the Vim
> >>  maintainer(s) will have to redeliver (or evaluate the need to
> >>  redeliver) when any of the included interpreters are updated.  More
> >>  interpreters means more work for the maintainer(s); all I suggest is
> >>  that the maintainer calculate that cost...
> >
> >It might also be worth checking to see whether factoring out the
> >languages into shared objects would be a savings.  
> 
> A savings in executable size, but not maintenance cost though, right?

A bit in executable size (though the total size of the vim deliverable
would be about the same), but I was thinking that the maintenance cost
might decrease if the language modules were sufficiently decoupled from
both the vim distribution and the core language distribution.

But I think I've convinced myself that that's only the case if the modules
were built entirely independently, and not with vim at all -- since their
link to vim is already negligible, but their binding to the languages is
likely to be (I haven't checked) much tighter.

Of course, since the proposal is for vim to become part of the WOS, then it
would make sense (IMHO) to support only languages that were part of the
WOS.  Vim would be redelivered with every build, and built with the latest
bits of the interpreters (at least potentially), and Stephen's redelivery
argument, I think, is a paper tiger.

Danek

Reply via email to