Vadim Gritsenko wrote:
Stefano Mazzocchi wrote:Vadim Gritsenko wrote:All: Is it ok if I remove old Cocoon releases from the xml.apache.org? To be precise: cocoon-1.7.x cocoon-1.8 cocoon-2.0 cocoon-2.0.1 cocoon-2.0.2 cocoon-2.0.3
-1, Brian asked for space conservatism, but that doesn't mean to destroy the history of this project.
he refers to removing nighly builds, ancient alpha and beta versions. releases should stay right where they are.
And he refers to "every minor release and beta version". What about removing those? We have betas and candidate releases:
Cocoon-2.0b1.zip
Cocoon-2.0b2.zip
Cocoon-2.0rc1a.zip
cocoon-2.0rc2.zip
no problem removing those.
those are not minor releases. Those are "releases", a release is something that you might build your stuff upon. It's a contract. Books, for example, use these contracts for years (before a new edition outdates another one, but that edition can still be used)We have minor releases: cocoon-2.0.1 cocoon-2.0.2 cocoon-2.0.3 cocoon-2.0.4
We can leave latest and previous release - 2.0.3 and 2.0.4 - and archive the rest on cocoondev.org.I would like to keep all of our final releases in one place. I like kernel.org that does that. I consider this historical value and also a sense of respect to all those who may base their software on a version of Cocoon and let it run for 20 years without changing it.
Cocoon distribution is not the lastest version, is the *entire* collection of releases.
Damn, we advocate that open source software removes lock-ins, but how about decades-long ones?
I'm strongly against fragmenting our distribution space on different machines.
I say we do this incrementally:Forever.
Generally speaking, I agree. Especially if they could not be reproduced from CVS. Moreover, all releases should be available and not only the set we have now (see <dream/> in this thread). But given limited hardware, I think it's ok to archive it on another machine.
What do you think?
1) first we remove all the betas, alphas, release candidates from xml and jakarta
2) if the problem persists, we iterate again.
for the future, we'll think about ways to decentralize the distribution system even from the source.
--
Stefano Mazzocchi <[EMAIL PROTECTED]>
--------------------------------------------------------------------
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, email: [EMAIL PROTECTED]
