Sam Ruby wrote:
>
> Pier Fumagalli wrote:
> >
> > So, here I stand, my vote is a big -1 on a 3.3 as a newly
> > architected servlet container
>
> Pier, I beg of you to reconsider.
Read my email in detail... Read that phrase. I don't vote -1 on 3.3. I
vote -1 if 3.3 is based on a new architecture would introduce more bugs
and differences as 4.0 is doing.
> I was not present in the ApacheCon in 1998. Nor was I in the room when the
> Jakarta decision was made. Nor was I on the state at JavaONE, when
> Patricia Sueltz made the announced with you, Jon, Stefano, and Brian.
I love myself :) Sometimes I can just be sooooooo poethic :) :) :)
> I can tell you that from my perspective, after all that transpired - and a
> merge twelve months ago - Jakarta Tomcat was hardly an open source project.
> It took a lot of hard work to make it open and viable. And a lot more that
> those four mentioned above were involved.
No, I disagree. I might have not contributed to the Tomcat 3.x tree
(hey, I'm stupid, I can't understand that code, I admit it), but the Sun
folks - including Costin - were always very open and tried hardly to
build this community. And they succedeed, we wouldn't be here arguing
about this whole mess if they didn't.
> Pier, take a step back. Look at where servlets were as a technology two
> years ago. One year ago. Today. Project where you think they will be one
> year from now, two years from now. Do you really believe that irreparable
> harm is going to come from today's explorations and fleshing out possible
> alternatives - particularly since this is all being done in the open and
> under a license that would very much encourage code sharing and reuse?
No, again, read my posts on this topic.... Sam, I'm not asking anyone to
fork or go away. Shit, all I'm saying is that we decided that Catalina
was the container for 4.0, and that Tomcat 3.2 and its container would
have been a bug-fix and performance increase only.
I don't want a fork. All I'm saying is that, if 3.3 is -as I see it- a
major rearchitecture of the container that would introduce more bugs,
and support issues, that should be done as a proposal, to be evaluated
for Tomcat.NEXT.
> I very much believe the contrary - that the only potential irreparable harm
> that can come from this is the stifling of innovation.
And I'm not here to stop innovation. Shit, I didn't say no when we
dropped JServ in favour of Tomcat, even if it was MY baby. And I'm not
saying NO to other proposals for the next generation of Tomcat.
> Pier, Is now the right time, and is this the right way to fight this
> particular battle?
No, it's not... It's too late, and I already written "closed" on this
chapter at least 2 times. But it seems that part of this community is
not considering the "closed" word with the same strength another part
does....
> I don't personally care whether any particular change made by any committer
> is described as a bug fix, a feature enhancement, or an architectural
> change - these descriptions, after all, are subjective.
Bug fix and feature enhancements are one thing, and architectural
changes ARE another... You IBM folks taught me that when I was working
there :) :)
> My suggestion is that discussion returns to the technical merits of the
> various approaches, and exploration of ways in which the various branches
> can exploit and cannibalize the best ideas from each other. Independent of
> the labels currently associated with each proposal.
Agreed. And AGAIN, all I say, if anyone wants to make a proposal for a
new Servlet Container, please, our CVS is open, the mailing list is
here, and we're all going to examine it for Tomcat.NEXT. But now 3.x is
bugfix only and 4.0 is development, wasn't this what we agreed upon?
Pier
--
Pier Fumagalli <[EMAIL PROTECTED]> <http://www.betaversion.org/~pier>
