[PATCH] KIO::SlaveBase and the event loop

[Andreas Hartmetz]

On Monday 14 July 2008 18:27:26 Sebastian Trüg wrote:
> Let me open with a question: Is there any technical problem with enabling
> event handling in KIO::SlaveBase?
>
> The attached patch is not really pretty as it uses
> QCoreApplication::processEvents(). A proper implementation (which I would
> do once this issue is clarified) would be based on signal/slot
> communication. But it works and allows to test the event handling.
>
> As the event handling is disabled by default, nothing changes for existing
> kio slaves. But it allows my rebooted nepomuk search kio slave to update
> the folder contents when the database changes.
>
Is it for some reason impossible to use direct signal/slot connections? They 
don't need the event loop to run.
kioslaves are already somewhat complex in some aspects and I'd like to avoid 
adding one more state that has to be tested and maintained in the future.
Somebody else (David, Thiago) will need to say something here, too. I am still 
relatively new to the KIO thing.

-- 
- Giant rat guys! Weird, interesting.
- And somehow appropriate. Shall we beat the living crap out of them?
- Can't think of a reason not to.