ant-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jose Alberto Fernandez" <>
Subject RE: Peter's ANT2 design proposals
Date Wed, 09 May 2001 19:19:30 GMT
> From: Peter Donald []
> At 12:09  9/5/01 +0100, Jose Alberto Fernandez wrote:
> >This would be the completly wrong thing to do. If we are
> going to create yet
> >another *Config kind of crap, this is not going to work.
> >What I would like, in terms of Ant1 terminology:
> >
> .....
> dynamic templating is not really part of build process it is part of
> autoconfig stage. As Ant2 probably won't solve this just yet
> we will most
> likely keep the scripty tasks ala AntOn around - hopefully
> ant3 will kick
> this bad idea out for good ;)

autoconfig is garbage, sorry. You finish having two build files. The
"source" which in is some pseudo language (not ANT's) and the "compiled"
file which is ANT. Every time you modify anithing you have to remember doing
it on "source" and regenerate. Every time you modify or add more directories
or watever you have to regenerate your build file.

One of the beauties of ANT is that it eliminates all that crap we had to do
in make everytime you added a new package to your build. I want a one step
solution that uses a one conprenhensible file that I can understand and
maintain. No magic 3 liners that no one has an idea what is doing.

I worked in build environments all based on make templates and at the end
they needed an entire support group to maintain the makefiles. And the worst
was that very few people understood and use what was in there.
I will argue that that level of magic is very very bad.

BTW, I do not think <script> is the right tool either. I want a
"declarative" as much as possible approach. That is simple to read and
understand and it does not put constraints where they should not be one.

Jose Alberto


Jose Alberto

> Cheers,
> Pete
> *-----------------------------------------------------*
> | "Faced with the choice between changing one's mind, |
> | and proving that there is no need to do so - almost |
> | everyone gets busy on the proof."                   |
> |              - John Kenneth Galbraith               |
> *-----------------------------------------------------*

View raw message