[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
[
List Home]
|
RE: [mylar-dev] connection timeouts
|
Yes, our offline support should make even very slow repositories usable. As
of the fixes to 0.9.1 I haven't seen a single timeout, and I've been on some
very slow connections while traveling in Europe. If others are still seeing
timeouts please post.
What I'm quite impressed with now is how much better we are for offline work
than during my last trip with sporadic connectivity (June). When I get to a
hotspot I just hit the synch button on the toolbar, then on the plane I read
all the incoming changes and write comments. The main thing that's still
missing is having the task data stored for the new hits. The manual
work-around I do now is to schedule all the hits that came in for today.
Mik
> -----Original Message-----
> From: mylar-dev-bounces@xxxxxxxxxxx [mailto:mylar-dev-bounces@xxxxxxxxxxx]
> On Behalf Of Eugene Kuleshov
> Sent: Thursday, November 23, 2006 11:22 PM
> To: Mylar developer discussions
> Subject: [mylar-dev] connection timeouts
>
>
> It looks like current bugzilla slowness on eclipse.org highlighted
> several issues including some inconsistent error handling and absence of
> retry.
>
> I already pointed out earlier that not loading task info in the
> background makes user to wait when server roundrip will be completed on
> editor opening. Which is quite noticeable, but also adds some confusion
> if task has the stale data and refresh failed. I've seen that rich
> editor has missing comment comparing to browser tab in the same editor.
>
> What is even worse is that failures on submission. In this case
> request timed out with all kind of errors. So, task is staying in
> "outgoing" status, but in fact comment actually made to the server and
> on second submit you can get either mid air collision, or for new task
> duplicates can be created.
>
> One of the great features of Mylar is the offline task support, but
> those issues are killing all the benefits. I wonder if there could be
> some easy fix for that stuff, perhaps increase some timeouts or make
> them configurable...
>
> regards
> Eugene
>
>
> _______________________________________________
> mylar-dev mailing list
> mylar-dev@xxxxxxxxxxx
> https://dev.eclipse.org/mailman/listinfo/mylar-dev
