ant-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Adam Murdoch" <>
Subject RE: [VOTE] Ant2 codebase adoption process
Date Thu, 24 Jan 2002 02:21:32 GMT

> -----Original Message-----
> From: Jose Alberto Fernandez []
> Sent: Wednesday, 23 January 2002 8:23 PM
> To: Ant Developers List
> Subject: RE: [VOTE] Ant2 codebase adoption process
> Guys and gals,
> I think it is too soon to be making decisions about
> one or the other. Pete has had myrmidom out for quite
> a long time, but quite little info outside of code has
> he provided. On the other hand, mutant just showed up
> two days ago, fantastic introduction, but I do not
> think anyone has had time to digest it, or try it, or
> had time to think on what ideas are good or bad or how
> extensible, flexible, or whatever this proposal is.
> I think we need to give the proposals some time to
> mature before we come with an axe and start chopping
> things down and commiting ourselves to one or the
> other.

Good point - there *is* a lot code there to wade through.

I think we can reduce the 'big bang' effect by tackling the proposals a
feature-at-a-time.  Pick a feature, nut it out, and extract out a first cut
into somewhere where people can try it out and build on it (and build *with*
it too).  We would do this gradually, to give people time to ponder the
issues a little.  One thing to keep in mind is that we don't have to get the
feature "right" at this point, nor does it have to be "finished".  It just
needs to look kinda like what we might eventually want.

We can look at the features as they mature in mutant or myrmidon (or
elsewhere), rather than waiting for complete Ant 2 proposals.  If we wait
until then, not only will there be more code to wade through, but we *will*
have to choose one or the other.  This isn't like tomcat 3 and 4, which have
the servlet spec to tie them together;  If you write a build script or task
for myrmidon, it ain't going to work with mutant.  Unless you use the Ant
1.x compat layer, but then why bother?  By tackling features, rather than
entire proposals, we can hopefully get some degree of interoperability
happening.  Then we would have a bunch of interoperable container
implementations to choose from, and that can only be a good thing.


To unsubscribe, e-mail:   <>
For additional commands, e-mail: <>

View raw message