Results of Socket Standard Workshop at SS2001

Pete give me 4 lines here to summarize what happened:

Pete We decided that there were several levels to what needs to be done. at the lowest level is the primitive interface (Smalltalk<-->C). Since all socket communications are (usually) done through the Berkeley API, and since the Berkeley API is a "known" on each platform (lots of documentation, works "almost" the same across platforms, etc) we decided to expose the Berkeley API through a very thin Smalltalk wrapper.

Pete Once that is done, the next thing to do is to abstract all of the various platform and protocol indiosyncracies and create a "usable" set of Smalltalk classes for sockets and name resolution, addresses, etc. that can hide some of the differences in the various socket implementations based on the BSD API.

Pete On top of that will go the "nice" high-level things like Streams etc.

Pete My job, of course, is the primitive API. And I got buy-in from all of the vendors for them to implement the required primitives. The idea is for each vendor to implement the primitives but not support or expose them. If we have the primitives then we can start working on all the rest, independent of vendor support or activity.

Pete That's the short summary :-)

Pete when I say "My job..." I of course mean "Me and anyone willing to lend a hand" ;)

reinz Sounds good, have any deeper details been finalized, eg method names, required Globals etc?

Pete nope. I was actuall really happy to get a firm decision on the need for the common primitive layer. once that's in place, the other stuff [lots of arm waving here <g>] can be worked on instead of thought about.