Wendy Smoak wrote:
On 12/14/06, ekio <[EMAIL PROTECTED]> wrote:
Is it intended that system properties don't work with things within
<parent>?
Let say I have the following in the POM, and I call mvn
help:effective-pom
-Dsomething=-dev
It seems like ${something} is not getting filtered...
<parent>
<groupId>group</groupId>
<artifactId>artifact</artifactId>
<version>1.0.0${something}</version>
</parent>
I don't think this can work. One of the first things Maven does is
figure out the pom hierarchy, so if it can't find its parent pom, it
stops. Replacing expressions with values comes later.
I know we had this exchange before, at least now I see I'm not alone...
I don't think it would be a disaster if properties specified via -D
would always be resolved. In fact, it would add consistency and also be
in line with the tradition of honoring the most recently specified
settings over any computed or derived settings. I'm really at a loss to
understand the objections against this. In a similar manner, environment
variables should be resolvable at any time also, but I can live without
that.
On a side note, I think that any setting that has unresolved ${...}
constructs should cause errors, and -not- as it does today create
directories named ${xyz}, which tend to cause interesting problems when
viewed via samba or such :)
Yes, it would be nice if I could do everything within maven, but until
maven offers a usable C/C++ artifact type and the associated plugins,
maven should at least assist in the task of embedding it into a larger
scale build system instead of making it hard to do.
--
cg
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
