I'm not sure I know what you mean by listeners. I don't see any code
that exists that does anything like this. Can you explain where you
think events will be created, and what scope they will have?

(more comments inline)

> -----Original Message-----
> From: [EMAIL PROTECTED]
[mailto:nant-developers-
> [EMAIL PROTECTED]] On Behalf Of Gerry Shaw
> Sent: Saturday, June 01, 2002 7:26 AM
> To: 'NAnt Developers'
> Subject: RE: [nant-dev] New Release Date (June 5th)
> 
> Well I'm back.  Scott and Ian should be able to make releases (I meant
> to have at least Ian have that access).
> 
> >From reading all the nant-dev email this is a summary of what I've
seen
> (I've excluded bug reports and new tasks)
> 
> Issues I've noticed from reading nant-dev since I've come back
> * Check what the default no command line argument action has been
> changed to.  IMO it should execute the global tasks, than the default
> target then exit.  It should also execute any listeners that get
> installed.

This works as you have stated. The program flow is a little strange, and
needs to be re-factored. (It should be easier to detect the
no-global-task, no-targets, and no-default states.)

> * Find out resolution to build file change (not) causing csc to
> recompile.  IMO it should not cause a recompile.

Unless I missed a checkin I don't think anyway has changed the csc, or
compilerbase tasks to use the build file as a dependency.

> * review Project constructor changes
Yep, that would be great.

> * are there configuration files yet?

Nope, I've talked about a few things that I have yet to deliver on.

> * replace onsuccess and onfailure with buildlisteners (they were
present
> but I removed them in favor of a listener approach - this requires
build
> events).

Events sound like a good pattern to support this kind of feature.

> * confirm that basedir on project is gone

It is not yet gone.

> * reflect on the ability to force a called target to execute (are we
> sure this is a good idea?)  I think so but I don't think the Ant team
> would agree.

Hmm.... actually, the AntCall task does indeed create a new instance of
the callee and execute it. This would be the opposite of the existing
behavior we have now. And would lend itself to the suggested changes.

> * define a base set of "core" tasks that will ship with nant (this
> should be the smallest subset of tasks possible to do XP like
> programming).  Everything else should go into the contrib/extras/etc..
> Project that Scott is setting up.

Here, here!. I'd suggest that any task which depends on any
functionality that isn't in the .Net SDK(/Mono SDK in the future) should
go into the NAntContrib Project. The question here is how should these
two project be distributed. I'd like to see a single download what
includes both. In most cases I think that users will want the full set
of available tasks for their use. Making them download and install
(extract) 2+ files could get a little tedius.



_______________________________________________________________

Don't miss the 2002 Sprint PCS Application Developer's Conference
August 25-28 in Las Vegas -- http://devcon.sprintpcs.com/adp/index.cfm

_______________________________________________
Nant-developers mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/nant-developers

Reply via email to