ant-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Erik Hatcher" <jakarta-...@ehatchersolutions.com>
Subject Re: [VOTE] Ant2 codebase adoption process
Date Fri, 25 Jan 2002 18:23:06 GMT
I didn't vote as I've been swamped this week and have not had a chance to
catch up yet - and is now a moot point.

I would like to hear counter-points to each implementation - Conor's take on
myrmidon and Peter's take on mutant.  What about each is different and
irreconcilable?  My only initial impression is that myrmidon is quite
different architecturally and 1.x cannot really evolve into it, but that
mutant seems like something we could evolve into by making progress towards
it with successive 1.x releases.  Is that impression accurate?

I have not had time to dig into either, nor do I foresee myself digging into
them much in the near future simply because I, as many others, don't feel a
compelling need switch to a completely new architecture yet.  There are
certainly issues which are making Ant 1.x progress difficult or impossible
(the classloader issues, for instance).  I'm more inclined to favor an
approach that evolves since that gives us something we can all work with on
a daily basis using our existing build files and techniques.  Evolving
doesn't mean we can't eventually make breaks with backwards compatibility
and make dramatic improvements - we just need to note that and move on.  We
can keep older versions of Ant available for builds that depend on those
versions.

Having competing proposals being prototyped doesn't unify ant-dev and only
seems to make it more difficult to accomplish our goals with efforts being
split.  I have felt my own share of frustration over progress in Ant 1.x
being rejected, so I'm not sure how the evolutionary concept will work
though.

I would like to see a larger separation between the core engine and tasks -
I just thought I'd throw that out as a goal that I have.  Currently ant-dev
is caught up maintaining tasks much more than the core, it seems.

    Erik



----- Original Message -----
From: "Conor MacNeill" <conor@cortexebusiness.com.au>
To: "Ant Developers List" <ant-dev@jakarta.apache.org>
Sent: Friday, January 25, 2002 8:59 AM
Subject: Re: [VOTE] Ant2 codebase adoption process


> I would like to try and address people's many comments on this thread
> but first I guess I should declare the vote a failure. Many committers
> did not vote directly and of those that did, while we have some support
> for the cutoff date, (+3.5) votes, there is insufficient support, (+2.5,
> -1) votes for the selection method. Agreeing on a date but not agreeing
> on what to do at that date is not that useful :-)
>
> So, besides the debate about whether we need Ant2 at all, I think we do
> not have a process for getting to Ant2. Sam  suggested the application
> of Duncan's "rules for revolutionaries". I wonder whether those rules
> will work or will the "dampening" effect he suggests just lead to
> indefinite postponement. By suggesting a majority vote on selection of a
> codebase, I wanted to at least ensure a result, a choice, in a bounded
> timeframe. We have been waiting for Ant2 too long already - well at
> least some of us have :-). My suggestion, therefore, is to leave aside
> the question of timeframe and simply try to resolve the process for
> codebase selection. Do you think rules for revolutionaries will work?
>
> Any suggestions would be welcome.
>
> Conor
>
>
> --
> To unsubscribe, e-mail:   <mailto:ant-dev-unsubscribe@jakarta.apache.org>
> For additional commands, e-mail: <mailto:ant-dev-help@jakarta.apache.org>
>
>


--
To unsubscribe, e-mail:   <mailto:ant-dev-unsubscribe@jakarta.apache.org>
For additional commands, e-mail: <mailto:ant-dev-help@jakarta.apache.org>


Mime
View raw message