maven-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jason van Zyl <>
Subject RE: How set in project.xml ?
Date Fri, 12 Dec 2003 21:57:17 GMT
On Fri, 2003-12-12 at 15:38, Lester Ward wrote:
> > Your analysis is simply erroneous. We don't make changes 
> > arbitrarily for the sake of making changes or to cause users 
> > long-term grief. So far I think I've done all right in OSS 
> > using similiar practices that I employ for Maven.
> I agree. Maven is a wonderful piece of technology.
> > Velocity, Apache XmlRpc, OJB, BCEL are all and 
> > haven't fallen prey to disuse yet.
> Nor has Maven. Sorry if I gave the impression that I thought it had. 

No, no. I think I read you correctly and I understand the sentiment but
it's been ~4 years now I've been doing OSS full-time so I've seen what
you talk about. 

> My
> point was only that I've seen projects disintegrate when they began to
> insist that the rest of the world conform to them _unnecessarily_.

I think the opposite has happened with Maven though. Compared to Ant you
are greatly more restricted in what you can do.

> I think the basic issue I (and, I think, some of the other posters) have is
> that they don't see why fixing the target directory in place is _necessary_.
> What benefit does it provide to fix it in place? Why is that benefit worth
> more than the flexibility of the current system?

In the long run in may not be, but I always lean toward the notion of
restriction for the sake of maintainability. For example, if the build
directory wasn't configurable ever would it have crossed people's minds
to make it configurable or would they have worked with it. I'm not sure.

> > Again, I believe you are wrong and that given the benefits 
> > users derive from Maven they will eventually start asking 
> > makers of tools to accommodate Maven's methods of development.
> Some will. Some won't. That will cause pain (if Maven becomes less flexible)
> for those who want to use the systems that won't conform. My experience is
> that open source developers tend not to have to deal with such pain, so are
> overly unsympathetic towards it. I can agree to disagree here, though.

There's really no restrictions on integration given pre/post goal
manipulation. Anything is possible and that mechanism will always
remain. Your example of moving the build directory could be done using a
post goal. Not as effeciently but could be done nonetheless.

> > I don't feel compelled to defend my philosophy because it manifests
> > itself in Maven and you're obviously using it so you must 
> > already agree to some extent. And I can see that you care because you're 
> > arguing with me which I take as a compliment. 
> I do care. The reason I am posting is that you appear to be on the verge of
> changing the philosophy used in Maven (i.e. replacing a currently flexible
> property with a rigid standard).

Again I find that interesting because my view of what Maven is hasn't
changed from the day I started it. I'm also not going to do things like
arbitrarily remove the configurable build directory just because I feel
like it. I want to start doing little surveys of usage and if it was
found it was a feature used 20% or more users I wouldn't consider
changing it. 

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

Jason van Zyl

In short, man creates for himself a new religion of a rational
and technical order to justify his work and to be justified in it.
  -- Jacques Ellul, The Technological Society

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

View raw message