Sorry, I did walk away there for a little bit.

On 8/27/06, Brett Porter <[EMAIL PROTECTED] > wrote:


On 28/08/2006, at 7:14 AM, Wendell Beckwith wrote:

> Take toady's latest example, say you want to remove an ant build
> file and do
> things the maven way, so you decide to use the dependency plugin.
> The web
> site examples have the group and artifactId being
>
> <groupId>org.apache.maven.plugin</groupId>
> <artifactId>dependency-maven-plugin</artifactId>

The dependency plugin was accepted to this project, but hasn't yet
been released here. IMO, we should remove it from the plugin list or
put it in a separate section as it shouldn't be considered ready for
use here yet.

Still, please do file bugs against it where there are documentation
issues.


Done. http://jira.codehaus.org/browse/MDEP-34


> 1.) Publish a project plan and commit to periodic milestones.

Yes, we need a roadmap. Development on the Maven core has been on the
backkburner as we fix peoples pressing issues and work on the plugins
and, funnily enough, the documentation. As you'll have seen on this
list recently, John has been putting a lot of topics together for
discussion and they come up from time to time and get recorded. At
some point in the near future we'll have a roadmap for 2.1 out.

> A lot of plugins still are attached to beta APIs even when there
> are 2 or
> more released versions of the artifact available.

Specific examples? I don't see this in any plugins that aren't
themselves beta plugins.

> For each milestone
> release all code should be compiled with the latest as the rule
> rather than
> the exception.

I'm not really sure what this achieves for the end user, and whether
you are talking about just maven, or all its plugins too. I assume
you are referring to us learning from Callisto here, which I've
already ranted about on my blog, but I'd be interested to hear from
someone who is closer to that community that knows the tangible
benefits it brough.


I suppose this depends on where you draw the line between maven core
(whatever that is) and its plugins.  In all honesty I don't know where that
line lies. So some of what I see may be related to the plugins and not maven
itsself.  At present I can't think of where I saw a maven plugin using a
beta api but I do remember that while tracking something done in the repo I
noticed that there were 2 full versions ( x.0 and x.0.1) already available
and that the code was calling for an 2.0-beta version.  I recently wiped and
rebuild my local repo, and ran a search across all poms in the
org.apache.maven group and didn't turn up anything, so perhaps it's all
better now.  Until I can produce some reliable evidence I would mark this as
unreproducible.

The plan will let the community know what's coming and when
> we can expect every milestone build between now and the release.
> The plan
> is not static as you can updated whenever you want.

Yes, that's a good idea.


For my team, I have been using, with minor adaptations, the eclipse dev
process and in general I think it has the right amount of "agility".  We
post our plan early with our commited, proposed, deferred and rejected items
for the next release and we revise it through out the release process.  We
use confluence for posting so that people interested in it can subscribe
just to that page to cut down on unwanted emails.  Therefore, when we make
updated everyone who wants to be notified is notified and they can either
comment on the issues we have attached to each plan item or start a forum
discussion.  I've create a template that maybe of use to you all if you
wanted to go this way.


> 2.) Produce nightly and weekly integration builds.

We already do. We could do it better. I've brought this topic up a
couple of times on the Continuum list.


I'm not on that list but I guess I will have to be to get a better picture
of what's going on.


Maybe this is
> happening, but how would I know?  Both the Maven 2 and Continuum
> websites
> have a dead link to the Continuous Integration server,
> http://maven.zones.apache.org:8080/continuum.<http://
> maven.zones.apache.org:8080/continuum >

This seems to be the problem. Our nightly builds are produced from an
old system that we were intending to move to Continuum so hadn't
published links to. On the Continuum side, we had to move the server
"temporarily" due to resource constraints and the links haven't been
updated yet. Please file an issue for these.


Done. http://jira.codehaus.org/browse/MNG-2535


> 3. Update the website regularly.
> Just split the thing down the middle into released info (doc,
> tutorials,
> examples, etc) and development current info which at a minimum
> would be the
> last stable milestone.

There's been significant discussion on this on the list already which
I can give you pointers to if you need them, but I'm not rehashing
them again. I'm happy with the plan we have.

Unfortunately, when people have put forward proposals recently
they've been met with silence. We need more participation to get this
moving.

> However, could it
> not be more efficient to release them in mass, especially if they
> have been
> continuously updated to current API's/fixes.

Our experience is the opposite. We did this in Maven 1 and plugin
releases were always stuck behind a core release. It takes a lot more
effort to release every plugin at once.


I'm not asking that you abandon releases when you think there is merit or
for critical fixes.  Instead I'm suggesting that you publish ahead of time
that you are going to roll a milestone/release at a certain periodic time in
the future.  For the most part, no one on any eclipse list requests a
build/release because we (community and devs) know from reviewing the plan
when they will be.  Because the community knows, the community waits.
Because the devs know the community is going to pounce on the milestone like
the slowest Zebra ,they work to get it in shape slightyl ahead of time.  In
no way should this prevent you from release maven-foo-plugin between
milestones/releases.

The APIs are not a moving target. I like that we can push out a
single plugin when it has fixes, and that old plugins continue to
work with new releases of Maven.

On the other hand, it means we need a greater eye over everything to
make sure plugins do get released regularly. This needs work.

> Now I don't get to bitch, hit send and walk away, so what areas of the
> website/documentation are available for a person who has some free
> time.
> I'm hesitant to signup for something big due to day job and night time
> commitments, but I do have some time and I'd like guidance on what
> areas I
> can invest it with respect to making maven better.  I just want the
> pain to
> stop.  Maven's a great tool and we receive a lot of benefits from
> its use.
> However, we likely could do more, faster if some serious sharp
> edges were
> child proofed.

I'm happy to guide you into any area where you are interested to help
out. So, is it documentation that you want to help with? We have a
list of outstanding tasks which I can put in one place.

Or would you like to help pull together the roadmap for external
consumption?


I'm open to working with either or both.  I do believe that production of
the roadmap can help guide how to prioritize what documentation will be
needed and when though.

- Brett


Wb

Reply via email to