Hi Patrick,

On Feb 5, 2010, at 11:58 , Patrick Ohly wrote:

>> Ok, I see that the progress events for the server must get more priority on 
>> my todo list!
> 
> It doesn't necessarily have to be on your list, although I admit that I
> am a bit scared of having a go at it myself at this point because it'll
> touch quite a bit of code.

It's something I wanted to do, not in particular the progress events for the 
server, but the cleanup and unification in the architecture for the unilib 
case. There's a bit of a historically grown mess in the client case which has 
some functionality in the appbase/engine (singleton) that really belong into 
the session. Once that's done, we'll get the server progress events for free 
anyway.

How urgent is it in an overall perspective?

> [explanation why not aborting the session with the first datastore - thanks!]

Ok, now I get it. It's about collecting status from all datastores before the 
abort, not continuing sync. Sorry for taking so long to understand :-)

In that light I wonder if it wouln't be easier and cleaner to introduce a new 
session-level <initcompletescript> which would run at the point the 
initialisation package ends (that is, when the <final/> for that package is 
detected), which would have the final say about continuing the session or not. 
That script could check session vars for that decision which could be set from 
datastore level scripts. On that level, I'd also add an additional 
<alertcompletescript> that would be called after processing the alert and 
figuring out if a slow sync is needed. That would be the place to check for the 
unwanted slow sync condition and set a session var flag that 
<initcompletescript> could catch.

How does that sound to you?

Best Regards,

Lukas Zeller ([email protected])
- 
Synthesis AG, SyncML Solutions  & Sustainable Software Concepts
[email protected], http://www.synthesis.ch





_______________________________________________
os-libsynthesis mailing list
[email protected]
http://lists.synthesis.ch/mailman/listinfo/os-libsynthesis

Reply via email to