Paulo Gaspar wrote:

> First, you write too much about a name when the question has always been
> having or not a 3.3 in the 3.x branch.
>
> Most of us (for whom having a 3.3 is interesting) are still not concerned
> about having or not a revolution and a Tomcat 5. It is too soon to be
> concerned about when our main priority is to have something better than
> 3.2 for production _real soon_.
>
> 3.3 is the obvious name and the discussion has always been around having
> it or not.
>

Paulo (and others), an important thing to remember is that Apache projects (as
opposed to an arbitrary open source project) operate under a set of rules and
conventions that, in effect, are the "Apache culture".  Many of those rules and
conventions are documented (such as the rules on voting), but some are not.  One
of the things I took away from the PMC meeting yesterday is the need to better
articulate those rules.

However, one of them is that there is no such thing as a "version" of any Apache
project until there is a vote to go that way, and elect a particular code base
to be that version.  See below for more.

>
> Catalina was a revolution, a proposal on following a different path.
>

It was, until it was elected as the code base for 4.0.  Now, it's the
established direction for 4.x.

Note that there was no "jakarta-tomcat-4.1" branch, or any such thing as "Tomcat
4.1", until the vote that took place last week.  Now, there is.  Such a thing
hasn't happened for 3.3.

It is obvious why this hasn't happened -- this is one of those "culture things"
that wasn't clearly spelled out within Jakarta.  Costin agreed to rectify this,
so a "3.3" version proposal is likely to be forthcoming shortly.

>
> At the moment, for me (and possibly others) 3.3 is an evolution.
>

Regardless of whether or not this is true, it's still a new version, and still
needs to follow the same proposal and voting procedures.

NOTE:  When this proposal is made, people who vote on it should remember the
following:
* Electing a code base needs at least three +1 votes and no -1 votes.
* Only votes of committers on Tomcat (*all* versions -- it is all one project
  until someone forks it to a separate name) are binding.
* A +1 vote on electing a code base implies an *obligation* on the part
  of the voter to actively support the code base.  Among other things,
  that includes someone taking on the role of release manager, all +1-ers
  being actively involved in fixing remaining bugs, *and* (after the release
  ultimately happens, if and when it does) supporting users of the release
  -- in our environment, that means answering user questions on TOMCAT-USER.

(FYI:  I am on record -- see the PMC Meeting Minutes that will be published
shortly -- that I will *not* veto a release plan for 3.3 that meets my concerns
about support.)

>
> Maybe (or maybe not) some people already see Costin's work as 5.0 but I
> think that most of us don't go that far. I will not be thinking about
> what 5.0 should be in the near future.
>

So far (to my knowledge), Costin has not proposed it for this purpose.  However,
it is important to note that no vote is necessary to declare a revolution
(starting with the code currently in the HEAD branch of "jakarta-tomcat") and
working towards that goal.  The only restriction is that no one can call it
"Tomcat" in the mean time.

This principle was actually articulated in the "Rules for Revolutionaries"
document, which was triggered when I (incorrectly) tried to use the name
"Tomcat.Next" before there had been any such agreement by the development
community.  The result was the creation of the name Catalina, which did not
become "Tomcat 4.0" until the vote that made it so.

Names are important -- for a variety of reasons, including legal ones (because
the name "Tomcat" belongs to the Apache Software Foundation, not to the
individual committers).  Therefore, we as developers need to respect those
reasons and become more careful a out our use of those names.

Craig McClanahan



---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, email: [EMAIL PROTECTED]

Reply via email to