On Mon, Jun 26, 2006 at 02:02:16PM -0700, Stephen Hahn wrote: > Having watched the perl/mod_perl/Apache tussels of the last few > releases, I can easily assert that the tiger, though perhaps a bit > lame, is able to deliver nasty and scarring swipes and is not made > from paper. The maintainer should decide on which interpreters they > wish to include, since it is their time that will be impacted by those > deliveries. (WOS integration shares the initial pain of a bug, but > not the work to repair it...)
Okay, I see where you're coming from. I didn't pay close attention to the tussel you mention, but as I recall, there were long-standing problems in the community regarding version skew. I don't think this happens all that often. With respect to the WOS, I'm assuming that we're integrating the interpreters sufficiently long after their introduction that just about anything depending on it will at least have a version available that will support the new version of the interpreter. That is, I don't see us integrating the first alpha version of, say, Perl 7 the day after it comes out and nothing on the planet supports it yet. Whether the maintainer of vim (for instance) wants to take on the burden of having to update to a version that supports the latest perl (say) when the perl integration happens (or even to know when that would be necessary) is something he or she would certainly have to decide in advance. It could be as simple as specifying in the ARC case that the alternate language features' existence was itself Volatile, and might disappear at any time if the maintenance burden due to integration issues became too great. You're certainly right that some care needs to be taken with this, at least. Danek
