summary of last call and a few things since

[rkhalaf <rkhalaf@watson.ibm.com>]

Hi all,

I didn't get a chance to summarize our discussion at the full TC call 
since the IP discussion drove it into overtime and I had to drop off :(

I thought I'd summarize to this list before I forget what happened ;) 
and see whether you guys agree.
status:
-A few people tried the samples except the simple interop (echo) one. It 
went smoothly, but a few errors were pointed out that led to some 
interesting discussions.
-We would like to get samples that address some of the issues from main 
TC, and are looking for people to volunteer to submit samples to us that 
represent these issues so we can do some test runs and open a discussion 
on the behaviors.
-Discussions from the simple samples:
    -discussion on static analysis. To what level does one check? One 
engine may reject a process as invalid, while another more forgiving one 
won't catch the errors till runtime. Examples: error in the string in an 
xpath, is a fault actually defined, etc.
    -discussion on faults and in particular communication faults and 
runtime faults. are they caught by catchAll. The conversation leaned 
towards don't catch runtime faults, but consider catching communication 
faults. Another questions was regarding notification of partners on 
process failure due to a runtime error? favorite fault here? Ivana sent 
mail summarizing.
    -The bpws prefix should be defined separately and probably also add 
a prefixes section at the beginning of the spec, contrary to what is 
said in issue 102. The samples in the spec also need to define it. Chris 
Keller will send mail to the TC, and someone (I don't remember whom) was 
going to open an issue to the main list.

Tks,
Rania