On the vein of RadicalSuggestions, this one is really radical. (Dick, let us know when it's enough) It occurred to me that neither SIGPLAN nor SIGSOFT truly capture the elusive CoreMeaning of this conference, and consequently its community. Why not create a new SIG?, say SIG on Programming. This SIG could run the mature version of (OO)PSLA every year. Its academic-technical backbone would be a journal (say, Transactions on Programming (TOP)) published every month. Given the existence of that journal for attracting academics, the conference could afford to be much broader and more inclusive.
If this were to happen, besides the obvious benefits in the stable state, this would also be a way of redesigning (OO)PSLA with a positive spin (new SIG based on an existing conference, hence the right time to refocus the conference), rather than a negative one (OOPSLA is declining, let's revive it).
This is in itself a reasonable idea. But it reminds me that Dijkstra on many occasions pointed out that there were conferences and journals about programming languages, but none about programming. A consequence is that people who focus on algorithms, as Dijkstra did, may feel they inherit this space.
The design patterns idea illustrated the problem: we had long studied algorithms and data stuctures, but we didn't even have a name for the ways in which we structured programs. --RobertBiddleI'm strongly favor of PLSA! My only concern is if there is really a mandate and if it is sustainable? It isn't clear that either SIGPLANSIGSOFTACM nor OOPSLA attendees really support SIGPSLA as a concept. I confess that I'm not well versed (thankfully) in ACM/SIG politics but I think it would be at least useful to look at what freedom there is beyond 2005 (clearly 2005 can break from tradition but it would be good to understand if that is sustainable).