From: stephan b. <st...@s1...> - 2004-12-25 03:04:12
|
Returning to Marc's comments about s11n support on IRC today (yesterday?): How big of a role would you like s11n to play in the P distribution? We have at least these options: a) i can integrate P::s11n directly into P's tree. This requires a dependency on libs11n. Since s11n will be 1.0.0 in a few days, stability shouldn't be too much of a concern. However, s11n isn't yet ported to Windows, and i don't know how difficult it would be to port it (it uses very, very little Unix-only code, but porting build process would be a lot of work). b) Add optional support for s11n directly into the P tree, so the dep on s11n is optional. i think this will turn into a huge #ifdef problem, though, and probably cause dev problems for developers not using the s11n support. c) Provide no s11n-related support in the main P distribution, and i can provide P::s11n as an add-on. IMO this is the most practical, cleanest approach, with the least pain for the developers. We can keep P::s11n either in the s11n CVS tree or the P CVS tree, whichever you prefer. ??? Right now i'm more focused on seeing P2 reach a usable state than i am in extending the current P::s11n code (it works with P1, CVS version - but only with the hacked copy on my box). Once the IOManager is working and has at least one supported protocol i can really start with the P2::s11n support. i've already started some minor architectural changes in s11n 1.1-dev to make the P::s11n implementation easier to do. To be clear, s11n can be used in lots of P contexts other than IO/Net, like App or Pref serialization, but it is the IO/Net options which are ultimately the most interesting. Why? Because we can already serialize objects, but we can't yet do it over the net. i really want to see the network channels working :). The prototype code works, and so i know P can do this, but P2 isn't yet in a state to support it. Soon... soon... -- ----- st...@s1... http://s11n.net "...pleasure is a grace and is not obedient to the commands of the will." -- Alan W. Watts |