I may reply to John's very detailed post later, particularly focusing
on areas where the application is limited by HTTP at the moment (and
hence some skepticism as to whether the whole application should be
exportable as web services) but for now I have three questions:

1)  What is the scope we want to give to the web service interface?

2)  How many of these areas would the database-level API structure we
are working on be a better or worse fit?

3)  As far as state handling, do we really need "stateful web
services" or should we look at a re-usable API which could be
encasulated over other protocols (for example XMPP) where we need a
richer interface/things like state handling?

I will say clearly that in an ideal world, everything in LedgerSMB
would be running over a fully stateful protocol, but that's not a
model we have inherited and so we do a lot of work in some areas
getting around the fact that we are using the adjustable wrench that
HTTP is to pound nails.

Best Wishes,
Chris Travers

------------------------------------------------------------------------------
All the data continuously generated in your IT infrastructure 
contains a definitive record of customers, application performance, 
security threats, fraudulent activity, and more. Splunk takes this 
data and makes sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-novd2d
_______________________________________________
Ledger-smb-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ledger-smb-devel

Reply via email to