You are viewing a plain text version of this content. The canonical link for it is here.

Posted to alexandria-dev@jakarta.apache.org by di...@multitask.com.au on 2002/05/03 05:01:49 UTC

Namespaces and Transforms

Sam,

I hear you. I understand where you're coming from. I can also see that
there are diverging ideas on project descriptors which have been solved in
two ways:
1) Augment with namespace definitions, and
2) Generate from different document.

The bottom line is that as long as they both work, who cares how the end
result is achieved.

I'd imagine most people who've written code generators have never written a
java compiler.

I'd hope from these exchanges it's clear that we're simply not ignoring the
work. We're working with Gump. If it turns out one descriptor is a simple
transformation of the other, then it'll be easy to adopt Gumps descriptor,
if not, we can propose the changes back to Gump/Alexandria. It may be that
Gump and Maven don't have the same need for the project information, since
they do different things with a 'project', e.g. look @ Ant's 'project
descriptor'.

At the moment, we've not gone the extend with namespaces way, but it really
shouldn't matter.
--
dIon Gillard, Multitask Consulting
Work: http://www.multitask.com.au
Developers: http://adslgateway.multitask.com.au/developers

"Sam Ruby"
<rubys@us.ibm. To: "Jakarta General List" <ge...@jakarta.apache.org>
com> cc:
Subject: Re: You guys are so funny.
05/03/02 12:42
PM
Please respond
to "Jakarta
General List"

dIon Gillard wrote:
>
> I don't see how we can be more accomodating other than downloading,
> installing and running our own Gump. From talking to Vincent on that,
it's
> not a simple process, hence we are relying on the Gump developers to tell
> us where we're going wrong.

Look at it from my perspective for a moment.

I took initiative. I build a working system. I provided some
documentation. I've responded when people have asked for features like the
ability to control their own descriptors. I've asked for feedback. I've
quickly given commit access to anybody who even expresses the slightest
interest and has even some minimal competency.

What feedback do I get?

> We've chosen to generate a descriptor rather than use namespaces, but
other
> than that, I can't tell why you're complaining - throw us a bone....

Imagine somebody writing a code generator, never having installed a
compiler.

Let's be clear - I am not asking anybody sully their hands by actually
running Gump, but there must have been a reason why a different DTD was
chosen than Gump's. I made an effort to document the Gump data definitions
and there certainly is plenty of instance data to look at. Tell me what to
change, tell me what's wrong, or simply tell me they suck. All I ask is
that you don't continue to ignore this work.

Let me be clear: I don't give a rat's behind whether the project
definitions are processed using XSLT, DVSL, or C#. But is it too much to
ask that somebody showing at least some token interest in converging on the
data definitions?

- Sam Ruby

--
To unsubscribe, e-mail: <ma...@jakarta.apache.org>
For additional commands, e-mail: <ma...@jakarta.apache.org>