Tim Williams wrote: > David Crossley wrote: > > Tim Williams wrote: > >> David Crossley wrote: > >> > > >> > We intend to "update" the version of Cocoon used by Forrest > >> > to be Cocoon-2.1 branch. > >> > > >> > Please vote. > >> > >> I reckon I've missed discussion on why 2.1.x vs the released 2.2? > >> It's not intuitive why we would "upgrade" to an outdated version. I > >> saw FOR-1016 but I didn't see what the show-stopper was. A search > >> turned up this[1] thread with Thorsten concerned about "pass-through" > >> - but I'm not sure what that means. > > > > That is the ability that lets our sitemaps work. > > If a match is not found then it passes through to the > > next sitemap. That ability was in our old Cocoon-2.2 trunk, > > and we were not sure it was backported to Cocoon-2.1 branch. > > It was and it works for us. > > > > As for the background, please see the recent "Proposal" thread. > > It directs you to > > http://issues.apache.org/jira/browse/FOR-1017 > > and to the old discussion: > > http://marc.info/?t=116969435600001 > > Thanks David, I've read this and still don't see why *not* 2.2.
Because no-one at Forrest is doing the work towards getting that to happen. Cocoon-2.1 is closer to the old version of Cocoon-2.2 that we are using, than we are to the current Cocoon-2.2 head. We must get out of this limbo. It is damaging us in various ways. e.g. Consider what would happen if a major security flaw was found in Cocoon. They could release a fix. However we would not be able to fix Forrest, and something drastic would need to happen. As we said at the beginning of this current proposal, this move does not stop us later going to the new Cocoon-2.2 or even something else. > From all that discussion the only problem I see is the CLI - Thorsten is > currently shopping for a Member to champion his droids lab ATM that > could likely be suitable for that. Maybe. There are a lot of abilities that it will need. As far as i know, upgrading to 2.2 is more than just the CLI. If it was easy then surely one us would at least show that it is operational in the 'forrest run' mode. Also it has moved on 18 months since those threads. Some abilities are now gone. > I assume the other advantages to > upgrading (i.e. XMLFileModule updates, etc.) are applicable to both > versions anyway. Yes, i would expect so. > I'm trusting that it makes sense, but I suggest again that because 2.2 > is the more obvious upgrade path, we'd do well to state explicitly > what's keeping us from it and essentially "reverting" to 2.1. As I > understand it, we've been using a variant of 2.2 for over 3 years now > and the next version of Forrest would potentially have "upgrade to > 2.1" listed in the changes. Yes. Well we have lots of words in these threads to compile into some encouraging statement. > I personally think that's just odd enough > to deserve something more explicit in the voting process than long > references to discussions that weren't exactly fully conclusive. Yes one of us should have tried to better summarise the past discussion and the current state. Thanks for expressing your concerns. That is very important. -David
