tomcat-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Craig R. McClanahan" <Craig.McClana...@eng.sun.com>
Subject Re: [ANNOUNCEMENT] Tomcat 3.2 Final Release
Date Thu, 30 Nov 2000 18:07:32 GMT
Rogério Meneguelli Gatto wrote:

> Hi, Henri,
>
> I also submitted a patch (yesterday, Nov 29) and before that a request (Nov 20)
> for Windows SMB Paths in Tomcat 3.2 server.xml.  On my first post, I even
> volunteered to make it work again (it did on 3.1 and it does on 4.0).  I had no
> answer to that post, and started to work "on my own".  I came up with the patch
> yesterday and sent it.  No replies either...
>
> I understand that not every contribution is accepted, but I would like to at
> least know why... Wasn't it well implemented?  Too dangerous to commit without
> more testing?  Im my case, are "\\machine\share\webapp" docbases incompatible
> with other goals?
>

This contribution was not rejected -- it just wasn't included in Tomcat 3.2 final.
Let me explain why for a minute.

The way I viewed my responsibility as Release Manager was to actually get a release
created, that worked for the large majority of people and didn't have any critical
bugs that affected most people.  Bug reporting and patch contributions happen
continually -- yours happened to come in the day before a final release was planned
to be created.  The ***last*** thing that a release manager should do is apply a
patch that could destabilize the release, without being able to go through a beta
cycle with it.  In your case, even though the patch didn't break my simple tests, I
judged the risk to be too high, given the timing.

Now that 3.2 final is out there, we can start working on 3.2.1 and integrate
patches that were not included.  We have the luxury of time to test them, without
people breathing down our necks about getting a release done that was promised six
months ago :-).

If someone else is selected as release manager for some future version of Tomcat,
he or she might have a different philosophy.  But this kind of release management
process has worked well for me throughout my career, so I applied it here as well.

Comments on the process (as well as the code) are of course welcome -- Tomcat is
fundamentally the result of a community effort.

>
> Regards,
> Rogério Gatto
>

Craig McClanahan



Mime
View raw message