[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: RE: [ws-rx] More on i113
From: Marc Goodner
[mailto:mgoodner@microsoft.com] I agree there are
more ? than in your post. There are a number of differences that make your
proposal difficult to evaluate though. Specifically there is a break in your
tables that does not occur in WD12 that separates events that cause state
transitions and those that don’t. There are also some events that are in
both the RMS and RMD tables that are only in one or the other in your proposal,
one example is “elapse expires duration”. I only modified RMS table so far –
so that we can discuss first the direction of these updates before updating RMD
accordingly. What motivated the
change of connecting/connected to activating/activated? Activating / active. Seemed more intuitive at first to talk of an
“active [sequence] ” state, as opposed to “connected
[sequence]”, given that this suggests the other side is also connected,
which may not quite be the case. But I can see that “connected”
conveys the fact that the sequence is at least known from both sides. No qualms
keeping connecting/connected. We could use also “established” like
in TCP for describing the normal transfer state. The division of the
events into generate/receive greatly complicates these tables, are you
convinced it really provides additional clarity? I find it rather clarifies the tables…
IMO it certainly helps figure why in some cases we can afford to use “N/A”
while in others you can’t because the RMS/RMD has no control on these
events and must decide what to do in any state it finds itself when getting
these. Also for some events like Faults that can be generated by either RMS or
RMD, it was unclear whether these were either received or generated –
rather confusing. -Jacques Marc Goodner Technical Diplomat Microsoft Corporation Tel: (425) 703-1903 From: Mmmh… I have been using the latest
source provided to me by Matt Lovett I believe. Looking quickly at WD12 pdf, the tables
are pretty much the same as what I started from – actually even more
undetermined (still many “?”) So the changes I proposed (in red in the
RTF) depart indeed enough from the source tables to make it hard to identify the
original ;-) Jacques From: Marc Goodner
[mailto:mgoodner@microsoft.com] Actually, I’m
looking at this now and it looks like it is using an old version of the state
table. I can’t line this up against what is in WD12 at all. Marc Goodner Technical Diplomat Microsoft Corporation Tel: (425) 703-1903 From: Marc Goodner
[mailto:mgoodner@microsoft.com] Jacques, Is this document the
proposed updates you note below? View Document Details: http://www.oasis-open.org/apps/org/workgroup/ws-rx/document.php?document_id=17864 Download Document: http://www.oasis-open.org/apps/org/workgroup/ws-rx/download.php/17864/state-tables-JD-3-diffs.rtf Marc Goodner Technical Diplomat Microsoft Corporation Tel: (425) 703-1903 From: Matt: Most proposed updates (except some in my
#2) still apply to your latest tables – will propose a sample of updated
tables. Also propose the following: - to not
" Fault a Fault", e.g. if RMS receives a Message Rollover Fault for
an unknown sequence, it will
not complain back with "Unknown Sequence Fault". - When
sequence expires: propose it closes rather than terminates: one must still be
able to query it to
get a final Ack. Thanks, Jacques From: Matthew Lovett
[mailto:MLOVETT@uk.ibm.com]
|
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]