ant-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Adam Murdoch" <adammurdoch...@yahoo.com>
Subject RE: What is going on in ANT1.x
Date Fri, 08 Feb 2002 09:15:58 GMT


> -----Original Message-----
> From: Stefan Bodewig [mailto:bodewig@apache.org]
> Sent: Wednesday, 6 February 2002 7:08 PM
> To: ant-dev@jakarta.apache.org
> Subject: Re: What is going on in ANT1.x
>
>
> On Wed, 6 Feb 2002, Adam Murdoch <adammurdoch_ml@yahoo.com> wrote:
>
> > This is the magic of UnknownElement, which extends Task, but can
> > turn itself into a data type.
>
> <very good description snipped />
>
> > There's some very brittle stuff in there.
>
> It is, most of it is mine and I'm not proud of it ...
>
> > We would do far worse than to refactor the project building code,
> > and take the compatibility hit.
>
> I'm not sure whether I'm parsing this correctly.
>
> Are you proposing to refactor the project building code (which would
> be fine) without care for backwards compatibility (which would get my
> -1 if you want to do it in Ant 1.x)?
>

No, not without care for backwards compatibility.  But not strictly 100%
backwards compatible either.  I think it would be worthwhile for us to
reclaim some of the internals of Ant, and move them behind a better
specified API (that is, better specified as to which bits are the API and
which bits are the internals).

I'm not talking about an Ant2-style rewrite of the project building code,
just a better separation between API and internals (followed, of course, by
some serious refactoring of the internals).  Problem is, it really isn't
do-able without breaking compatibility somewhere.  The question is, do we
care about this for tasks that are using public code that they really
shouldn't?

This is the 'compatibility hit' I was talking about - breaking some tasks,
to give us more room to move.


Adam


--
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