Mailing-List: contact tomcat-dev-help@jakarta.apache.org; run by ezmlm
Date: Mon, 3 Jan 2000 15:57:42 -0800 (PST)
Message-Id: <200001032357.PAA24948@calterra.eng.sun.com>
From: Eduardo Pelegri-Llopart <pelegri@eng.sun.com>
To: tomcat-dev@jakarta.apache.org
Subject: re: open processes...
Reply-to: Eduardo Pelegri-Llopart <pelegri@eng.sun.com>

First a clarification: I have been following jakarta, tomcat, xerces,
xalan, cocoon, (JSP-interest & Servlet-interest), et al. via digests,
since I get quite a bit of other mail every day.  Unfortunately,
digests make it difficult to properly participate in email
conversations; and, to make things worse, digests from all the apache
sites were not sent for a couple of weeks.

Assuming software works properly, I will switch to undigest on Tomcat
starting tomorrow.  I hope that will make me more responsive on this
mailing list.  For now I'll keep participating on the other lists via
digests.


I'll now try to (start) answer(ing) a few threads that started around the
23rd....


First on spec development processes.  I have been directly involved in
the JavaBeans specification, the JavaHelp spec, and the JSP specs.
I've been indirectly involved in several other standards including
ANSI, OMG, and have been around people who have been involved with
several W3C processes [[I have not been involved with IETF]].  I do
not claim that the JCP (java community process) is the ideal process,
but it is the best one I've seen, and the same for everybody that I've
talked to.  The process is relatively quick, leads to pretty good
implementable and stable specs, and it is quite balanced and fair to
everybody.

The process can be improved (by design the JCP is under-specified),
and I hope we can do so in the next go-around, where we will have to
find a way that will work well for the open source community.  But
developing a spec is not easy, and a good process does not mean one
where everybody feels their needs are matched perfectly. A spec is an
excercise in compromise; anybody who participates in the expert group
should understand that, or will risk being frustrated.  There are
several portions in JSP 1.0 and JSP 1.1 that I would have done
differently had I been the only designer and had I been involved with
it from the beginning.

I should not be the one to praise our current spec process, you may
want to talk directly to people who were in the expert group; in the
JSP side, the representatives either have been pretty happy or they
have been not been straight-forward in conversations with me.  I've
done my best to satisfy the requirements described above.


On specifics:

| Stefano says:

| Ask
| Eduardo: first time I saw that I had to sign a paper to get into the JCP
| process, I jumped off the chair.

Yes, Stefano complained about the standard NDA that everybody had to
sign to participate in JCP.  The intention of the NDA is to protect
the development of the Java platform, and the interests of the
companies and individuals that participate in the process.  Without it
we would not have been able to get some companies to participate.

Note that many many companies and individuals, including many active
members in the open source community and in tomcat, have signed that
NDA.

But, as said above, we have to find a version of the JCP that will
work for the open source community and I will work on that.


| ...W3C...

I think you should talk with somebody who has been actively working in
the W3C.  Some groups work well, some do not.  And, organizationally,
the whole process is less open and accountable than JCP.


Stefano says...

| I'm already invited to partecipate to
| the JCP process of developing Servlet 3.0 and JSP-NG where I'll propose
| Cocoon ideas and solutions for integration with standards

Yep, I invited Stefano and several other people to endorse the "JSR"
(Java Specification Request) for the next version of Servlet and JSP.
Endorsing the JSR just indicates support for the creation of a new
spec, the "call for expert group members" will follow later.  My
intention was to craft a first draft of the JSR and circulate it
around several places, including this mailing list, before Christmas
but that has not yet happened.  I hope to do that in the next few days.


Ruby says...

| Personally, I would like to see a more open process for the designing and
| implementing of APIs.

I think that most people who participated in the process felt it was
quite fair.  In general the hardest partner to deal with is Sun
internal: it is hard to explain to them that indeed we are running a
fair process.


The bottom line to me is that the JCP has worked pretty well, and that
it is flexible enough that it can be made to work better.  I'd
encourage you to work through it; I personally pledge to continue to
do my best to use the process to improve the Java Platform.


Hope this helps.

	- eduard/o