(dropping private@, adding dev@).

On Mar 19, 2007, at 11:36 AM, Leo Simons wrote:
On Mar 19, 2007, at 3:55 AM, J Aaron Farr wrote:
"Henk P. Penning" <[EMAIL PROTECTED]> writes:

  Some 29 other excalibur artifacts aren't properly signed :

    http://people.apache.org/~henkp/checker/sig.html#unsig-excalibur

  They are all owned by Leo Simons.

Oh dear, I can't believe it. That's been broken for two years or so, then.

  ... I see some (all?) of the objects are symlinks to other things.
  The other things /do/ have proper sigs, so it seems to be a matter
of missing symlinks (to the corresponding .asc's). It would be nice
  if this could be fixed too.

Ug.  We could add the symlinks.  The thing is, the symlinks should be
updated and really, they're not even useful anymore.  Personally, I'd
like to just remove them.

Leo?  Thoughts?

I will find time to fix it tonight.

Should all be done (was indeed just symlinking to do). The main reason to keep the symlinks is google -- if someone searches for "avalon-framework-4.2.0.tar.gz" it is nice if they end up in the right location. Removing the symlinks would hamper that a bit.

I also note that /dist/avalon is missing 35 signatures, all related files owned by Niclas.

Also interesting from Henk's page -- there's a total of 3467 artifacts in /dist, now. Of those, 659 are from excalibur, 640 are from avalon, and 561 are from jakarta. The next runner-up then is tomcat, with a measly 168 artifacts. In other words, you *still* have to consider us rather big by modern-day standard!


Now, should we archive some things? The policy says,

"Older releases that you would no longer recommend to the general public should be placed on archive.apache.org."

that means we should probably get rid of /www/www.apache.org/dist/ avalon/ completely on the active mirrors and just have it on archive in theory, but in practice, files like avalon-framework-4.1.4.jar should simply stay up there in safety gathering dust, since that's still an active dependency for a lot of stuff "out there", and there's not really a big reason to not recommend it for use.

- Leo


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

Reply via email to