-----Original Message-----
From: Sam Ruby <[EMAIL PROTECTED]>
To: [EMAIL PROTECTED] <[EMAIL PROTECTED]>
Date: 24 March 2001 21:35
Subject: re: [proposal] fishtank
>Jon Stevens wrote:
>>
>> Is it cool if I import my generic nightly export/binary system into
>> Alexandria? It seems like this project would be the best place for it.
>
>Would it be possible to drive your scripts off of a common set of data
>definitions with any of the current alexandria code bases? The current
>alexandria proposals ("classic", "gump", and "ant") speak different
>dialects of the same basic language, but we are slowly converging. What I
>see in your projects directory is in quite a different direction entirely.
>
+1 It's already hard enough trying to work out what the three different
sytems do is as it is.
>I must confess that I was just about to make a proposal that we split the
>data definitions out into a separate cvs. Something like a
>jakarta-alexandria-data. Now I wish I had made that proposal first ;-) In
>any case, the set of people who should have commit access to the project
>definitions is arguably different (and in general, larger) than the set of
>people who can be trusted to make changes to the implementation itself.
>
Similar thought had been going through my head, It does seem like we're
looking at a set of meta data which should be available outside the realm of
build/docs. The idea of having XML definitions of whole projects does seem
to open up some possiblities, but it's very difficult to maintain sync with
projects so getting the projects themselves to maintain it makes our lives
easier.
The only problem with this is at the moment we don't have the standard
schema. I don't think we can really look ahead without this. I've already
had to put some ideas on hold as my first goal is to get Alexandrai to be
nearer Gump etc. Also there's a lot of duplication that just seems to be
wasting peoples time. I'd much rather be working on the same code as
everyone else.
>Also, I have been slowly making progress in getting projects to accept
>nightly builds using the latest of everything is a useful thing to have. I
>already have jakarta-ant, jakarta-avalon-*, jakarta-slide, jakarta-regexp,
>xml-xalan, xml-soap, and xml-axis signed up - these nightlies are all
>produced by Gump. It is all part of my diabolical plan to get people
>addicted to Gump to the point where the actually care if the builds fail.
>;-)
Doh I really get that test case of to the castor boys so they can sort
things out ;-)
>
>However, I am quite willing to say that the above is simply a "use case",
>and a generic build system should be able to be adapted to multiple
>purposes. But I would still like to see if there is some way that we can
>get all of the various proposals under the jakarta-alexandria umbrella to
>at least be driven off of the same data definitions.
>
>What I envision in a jakarta-avalon-data is a series of "profiles".
>Perhaps one designed to build a system using stable versions of each of the
>components, another which picks up milestone/beta versions, and one
>designed to pick up the absolute latest. In such a future, while little or
>none of the original code in the Gump proposal may live on in its present
>form, what will remain is a profile.
stable/milestone/latest Yup I see it that way to.
>
>So, how about it? It there any hope of adopting a common set of project
>definitions, independent of the "engine" that drives the process?
>
I don't see we have any choice, either that or we go home (oh I am already
;-) )
>- Sam Ruby
>
>
>---------------------------------------------------------------------
>To unsubscribe, e-mail: [EMAIL PROTECTED]
>For additional commands, e-mail: [EMAIL PROTECTED]
>
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
