Danny,
> As I've said before, if someone has a machine on which
> they can make *clean* builds and can upload them to
> their home dir on cvs.apache
I have a system we can do that on, and I am willing to setup a separate copy
of the CVS for doing clean builds. I could probably use some help
automating the process.
Right now, I have cvs and ssh setup to use pki, so interactive login
shouldn't be an issue. What I'd need to do is have a cron job that updates
from the cvs, does the build, and then either [produces the files and uses
scp to move them to a directory under ~noel], or sends e-mail regarding the
build failure. If there are no changes in the CVS, I suppose that it
shouldn't actually do anything.
Do you have something suitable?
I suppose that once we've adopted Maven and/or Forrest, we can enhance the
script(s) to do the web site updates, too.
> 1/ IMO it helps if you change the name of the archive files
in build.xml to include either a date or a version [...]
> 2/ Then TAG the HEAD with this command in the root of
jakarta-james:cvs tag <version>
Let's just put it into the scripts.
But in the meantime, we still need to get that Milestone build posted
manually from Peter's directory. :-)
--- Noel
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
