ant-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Holger Engels <>
Subject Re: [PATCH] ant task with (nested) fileset(s)
Date Sat, 10 Nov 2001 13:57:19 GMT
On Sat, 10 Nov 2001, Jose Alberto Fernandez wrote:

> From: "Peter Donald" <>
> > I am still fairly sure I don't like anton (or javaon) style functionality at 
> > all. Basically it amounts to limited version of foreach where only one var 
> > can change each iteration. I can't see how either can lead to good build 
> > practices - anyone care to enlighten me ?
> > 
> In the case of <javaon> I agree with your argument (although there is still the
issue of
> having allowed <execon> as part of ANT in the first place) those asking for <javaon>,
>  I think are just asking for symmetry on the tasks provided by ANT.
> With respect to <ant> (or <anton>) fileset functionality I disagree with
your assesment. 
> The question is whether the best practice is to have to modify a main-buildfile everytime

> a new subproject is added, or this can be done by simply installing the subproject in
> right location in the source tree.
> For example in a project like tasklibs where you may have a large amount of quasi-independent
> subproject may be a good example. Shall each library be build only independently? shall
there be
> a main-build that needs to be modified every time a new taglib is added? or shall the
main build
> simply pickup the new taglibs as they are added. Different build shops may want to be
> able to use different policies on this regard and I do not see why one should be concidered
> better or worst than any other. It is a question about how much centralized control a
> wants to have of the build-process across an organization.

I totally agree with you .. almost every larger project is divided into 
several more or less independent subprojects.

I think, if the interface (in terms of xml) of my implementation is ok, 
someone should commit it to the repository. So that people can test it (I 
did already).


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

View raw message