about stability (upgrading pear...)

when i tried to upgrade Pear I get this: admin:/usr/local/share/pear # pear upgrade PEAR downloading PEAR-1.3b3.tgz ... ...done: 86,146 bytes 'xmlrpc' PHP extension is not installed PEAR: dependencies failed the strange thing is that xmlrpc support is tagged as experimental. Thus, first question: how is possible to need an experimental PHP library ("Use at your own risk...") when installing main Pear package? At least, it's a strange thing. Largely, I would be able to use Pear in my production servers, but I ever think: is it safe or not? Should I upgrade safety or not? Which extensions are required for the most packages I often use? Maybe -I'm talking as an application builder/mantainer not really as a tester, right?- I would be quite sure there are a group of stable packages which could be safetly plugged within my applications. As I can see, PFC follows this principle, but I feel there's no difference on quality assurance in PFC rather than other pear libraries. Is it right? So my purpose is: could you evidence a "pear core" or something like, that could be safetly used in live applications, maybe in a way of "some-sort-of-guaranteed-package"? Would be useful to test the core in a QA environment, like PHP self use with QA team. I suppose this doesn't happen today, but what I mean about _core_ is slightly different than what's in the manual. How do you feel about it? I'm really interested about this way of developing, but I became dubtful looking at the growing amount of packages that complicate our life since one don't follow much closely Pear development. Well, I think also documentation should help and be written following a "what-user-would-do" method, like you did for DB classes. I mean, it's not a matter of alfa/beta/stable, but of which libraries are "core", intended as "most closely packages" or "most well developed/patched/TESTED" or whatever. It's a nuance that I feel important. If I am wrong in something, please correct me, since I'm not a great PEAR user, but I'd like. However, I appreciate about the work you're doing and I'll go on using Pear! ;-) Bye, Alberto. -- "Imagination is more important than knowledge"