(...) Things that should get fixed but aren't, never are (unless there is a major problem such as permission_p).The problem is that you put kludge on kludge on kludge and by version 4.5 it will be a nightmare to maintain. (...) when you want to eliminate acs-messaging you can't because it will break something somewhere. (...)
-- Jon Griffin
I would like to disagree with that. First, though, I have to admit that I do not know the 4.x bboard design other than what I've picked up on this thread and seen in general-comments. From the issues I see in this thread I come to the conclusion that:
- There is a lot of interest in the 'perfect bboard', and I am also counting previous discussions on messaging, IM, email support for bboards, etc.
- the problems are so severe that there will be no way around re-engineering a to get a better messaging/bboard system.
So, yes, I do think the messaging/bboard/CR will be rebuilt, possibly from scratch, *hopefully* as soon as the port is done. That is what point-releases are for, this sounds like an OpenACS 4.1(?) thing. I think the core team is strong enough to pick that up and they will have support. To me that means it is good to have a straight port now.
Although generally annoying, breaking parts of the existing API in the next core release will happen no matter what. If that would not happen, we will likely get serious problems with design-rot - we are not gods, after all 😊 Sure, that means that you will break some or a lot of packages and those will have to be upgraded to the new API. It sucks, but that's life. OTOH, a lot of things will not change so much that parts of the old API can't still exist for backward compatibility. That will give package maintainers some time to catch up.
Don said that a bunch of us are in this for the long haul. That would include me, although I've only recently joined. There appears to be no (opensource) competitor for what OpenACS does. There are probably more people like me who want to set up their semi-private community websites that will hopefully survive a slashdotting.
