On 12/11/06, Mario Ivankovits <[EMAIL PROTECTED]> wrote:
Hi Niall!
> 1) Theres a failing test - does it matter?
> http://tinyurl.com/ymh763
No, I've documented it in the release notes ... though ... hmmm ...
maybe I found a way to make it run too ....
> 2) The dependencies are mis-leading since they include slide and jcifs
> which are in the vfs sandbox and not part of the release:
> http://people.apache.org/~imario/vfs-1.0-RC8/site/dependencies.html
> patch to correct this (and add urls and "optional" to the
> dependencies) here:
> https://issues.apache.org/jira/browse/VFS-100
Ah, yes, thanks for fixing it. (and also the missing headers)
> 3) Is it intentional to not include the vfs sandbox code in the source
> distribution?
How do other projects handle it, I'd prefer *not* to add it to the
source distribution as the "sandbox" is a different module (or project)
e.g. in the context of m2 which will create its own jar and thus its own
-source.jar.
The same goes for the "examples", I'd remove it even from the current
source distribution as its another module.
In MyFaces land we have a separate source jar for each distributed
binary jar which is a one-to-one mapping to the m2 modules.
I guess if you wanted to release separate artifacts that projects
could depend on then you would go the same way. Commons attributes
distributes multiple artifacts - so you could look at that - or maybe
it would be easier to wait until we adopt m2 in Commons
http://jakarta.apache.org/commons/attributes/
In the case of "sandbox" type code then IMO it either gets excluded
and not distributed (as you have it at the moment) or you distribute
just the source code so that for those adventurous souls that are
happy to build and use from source can do so.
Just my 2cents - at the end of the day, its really down to how you
want to play it IMO.
Niall
Ciao,
Mario
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
