Re: [mylyn-integrators] Startup help for writing new connector

[Steffen Pingel <steffenp@xxxxxx>]

On Wednesday 16 July 2008, Schenkenfelder Robert wrote:
> I discovered that Mylyn synchronizes my query every - don't no, maybe 10
> or 15 minutes. This means that the query is there, but just not shown in
> the task list. I can also add a new query to my repository with the same
> name of the old query. Shouldn't he at least say something to me?

The framework currently does not guard against creating queries with the same 
name.

> After closing and reopening the test workbench, the synchronizing is
> gone. The query was obviously saved wrong and/or non-persistent.

You could set a breakpoint in TaskList.addQuery() and TaskList.addTask() to 
verify that the elements are added to the task list. If that is not the case, 
look at SynchronizeQueriesJob.TaskCollector.accept(). The query results are 
handled and added to the task list here.

The externalization is implemented in DelegatingTaskExternalizer. You can take 
a look at the tasks.xml.zip file in your workspace/.metadata/.mylyn folder to 
check if your tasks were serialized.

> A point I still don't understand is the difference between Mylyn 2.3 and
> Mylyn 3.0 for query support.
> Following the instructions at
> http://wiki.eclipse.org/Mylyn/Integrator_Reference#Query_support,
> MyTrackerQueryPage.getQuery() should return an instance of my query
> (obviously the Integrator Reference is for Mylyn 2.3). Steffen, you said
> that getQuery() should return null for a new query. Is there any
> documentation available which shows up these differences or how can I
> know such things else?

There is not much documentation beyond the posts on this mailing list. I hope 
that I'll have time to publish a short porting guide in the next few days 
that will explain some of the differences. 

Steffen

-- 
Steffen Pingel - steffenp@xxxxxx - http://steffenpingel.de