OK, this is my opinion here.  I want to hear what the other committers have
to say on this topic.

I want to get v1.3 out.  I want us to finish what we have started,
consolidate it, test it, and release it.  I want it to be released before
the end of the year, preferably more in the Sep/Oct timeframe.  We've been
doing some great stuff, we've been doing it for a while, we need to get it
into the hands of our users.  Ceki and I presented the bulk of the features
for v1.3 back in Nov 2003 at ApacheCon.  It will be almost 2 years!

I don't want anything to get in the way of this.  I don't want to start any
new features, I don't want to open any new issues or topics.  That is not to
say that these issues are not important, I agree they are.  I just think
that opening them up is going to delay us from getting the other really cool
1.3 stuff out.

Also, to me, a 2.0 designation could/should be used to indicate some
fundamental changes/shift in the api/architecture.  If we are going to break
some pre-existing longstanding contracts, the 2.0 version is the time to do
it.  Let's go wild.  So, again, I think doing it will require more thought
and discussion than I want to take place to get a release out soon.

Ok, those are my thoughts.

-Mark

> -----Original Message-----
> From: Curt Arnold [mailto:[EMAIL PROTECTED]
> Sent: Tuesday, May 10, 2005 9:56 AM
> To: Log4J Developers List
> Subject: Relabeling log4j 1.3 as 2.0
> 
> There have been suggestions that it may be beneficial to relabel the
> current log4j 1.3 branch as log4j 2.0 and have an interim minor
> release from the 1.2.x branch designated 1.2.11 or 1.4.0 (will be
> raised in a separate thread).
> 
> The current 1.3 branch is a substantial enough change from the 1.2
> code base that there have been (and continue) many inadvertent or
> unexpected compatibility issues with 1.2, however the minor release
> designation has prevented desirable code changes that may cause some
> breakage.  For example, many classes that should not be extended
> (Logger and LoggingEvent) were not declared as final.  It is possible
> that some users have been able to successfully extend those classes
> in limited ways, so designating them as final is not an appropriate
> change for minor release.
> 
> Some potential things I could see that could be tolerated in a 2.0
> release that wouldn't fit in a 1.3 (please let me know any I have
> left out)
> 
> Designating classes not designed for extension as final.
> Lowering visibility of members and methods, particularly making
> protected member variables private and changing package visible
> member methods.
> Dropping JDK 1.1 collection types
> Finer grain synchronization with potentially some changes in locking
> semantics.
> Using JDK 1.5 typed collections on JDK 1.5+ while still supporting
> JDK 1.2 collections.
> 
> Hopefully these could be accomplished without compromising client
> code, but may break some amount of custom appenders, layouts and such.
> We are all anxious to have something to show for the long development
> time of the 1.3 branch, but as much as we've changed, it would be a
> missed opportunity not to break things that should be broken.  I
> would however prefer this not be an excuse for breaking things that
> should not have been broken and pushing back the release targets.
> 
> 
> ---------------------------------------------------------------------
> 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]

Reply via email to