On 7/6/05, Simon Kitching <[EMAIL PROTECTED]> wrote: > On Wed, 2005-07-06 at 20:36 -0700, Phil Steitz wrote: > > I am not sure that I have understood the thread(s) above fully, but it > > would seem to me that we need to immediately get the correct, signed, > > voted, *released* 1.0 jar into *both* java-repository and ibibilo. My > > guess would be that via a combination of pom and .properties config > > errors, someone published a snapshot (still named 1.0) to > > java/repository and this got replicated to ibiblio. > > Replacing the bad 1.0 on ibiblio with the correct one will make things > even more confusing I think. People who have already downloaded the 1.0 > will not get the update (including maven users where the bad jar is > cached in ~/.maven/...). >
But then at least what we have advertised as cli 1.0 would be the version that we released officially and new integrations with dependencies on that *released* jar will be built correctly. We need to acknowledge this problem on the web site, but getting the 1.0 version on ibiblio the same as the version on the mirrors seems better to me than leaving things in the state where they are. Again, could be I am not looking at this the right way. > It should be possible to get a 1.0.1 out within 24 hours I think, and as > this problem has been around for over a year that should be soon enough. > > And once 1.0.1 is out I am very keen to see 1.0 deleted completely, so > people get errors that should cause them to check the cli website, see > the problem info, and upgrade to 1.0.1. Why force users of the original release to upgrade? > > Pushing out the "good" 1.0 now only to delete it after 1.0.1 is released > within the next 24 hours doesn't seem terribly useful. The value would be to get the unreleased jar out of circulation and to replace it with the correct, released version. > > > > It goes without saying that we also need to find a way to prevent this > > from recurring, and also find out immediately if there are other > > situations like this. Anyone interested in helping to find a solution > > to this should subscribe to repository@ and discuss the general > > problem there. > > I agree. And the problem could well be a result of someone running a > maven task which pushed this out without them realising. Maven is very > powerful but can be confusing (at least I find it so). So looking into > whether there is something that can be done to avoid future mistakes > would be good. Agreed. I posted to repository@ and we'll see what ideas people have. Another thing we can do is add some docs to the commons web site explaining how the deployment properties work. I will start another thread to discuss this under [docs] Phil > > Regards, > > Simon > > > --------------------------------------------------------------------- > 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]
