Re: [g-a-devel] How to choose the suitable event in AT-SPI

[Michael Pozhidaev <msp altlinux ru>]

Hello,

Magdalen Berns writes:

> What is it you mean by "most preferable event"? Can you clarify that a bit
> further?

Some events in itself may be sufficient to reflect a particular user action
(e.g. text-caret-move notifies that the user is moving the cursor). But
it is the case only if there are no other events. If text-caret-move
comes in conjunction with text-change event it is a completely different
situation implying that the user is typing a new text.

If we would like to announce the exact user action  in one message we
must be sure that know the _*all*_ events which were produced due to some
user step.

The question is how I can be sure that there will no events to come next
if I already is considering one? Alexander suggested that probably I
should to wait a bit and if there are no events within some delay I must
to take a look at a whole number of events which were accumulated. I am
wondering if there are some AT-SPI features which could help to handle
this situation without a delay which obviously not a perfect solution.

> better. Take a look at Alejandro's examples  github repository in if that seems
> like it might be useful:  https://github.com/infapi00/at-spi2-examples

Yes, thank you a lot for this link, I didn't know about it and will take
a look!

-- 
Michael Pozhidaev. Tomsk, Russia.
Russian info page: http://www.marigostra.ru/
English info page: http://www.marigostra.com/