[Webkit-unassigned] [Bug 132350] AX: [PATCH] procedures for automated atk text and caret tests

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=132350


Mario Sanchez Prada <mario at webkit.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |apinheiro at igalia.com




--- Comment #8 from Mario Sanchez Prada <mario at webkit.org>  2014-05-01 07:21:40 PST ---
(In reply to comment #6)
> Sorry if I make a conceptual mistake again, but...
> 
> When we had webkit1, testatk.c did all atk (and spi2 in other words) tests, 
> including caret-moved tests. It served whole webkit, including webkit2, by 
> testing accessibility scenarios. Now that we don't have testatk.c, what 
> takes its role of testing atk related issues? Shouldn't we simply port 
> testatk.c to webkit2 and use it then as before?

See my previous comment for further details but, trying to reply shortly here:

 - testatk.c only tested things using ATK, no AT-SPI2 was involved at all
 - testatk.c only served the WebKit1 version of WebKitGTK+. Webkit2GTK+ or EFL, for instance, were not covered by it
 - Porting testatk.c to WebKit2 would be possible, but then we would need to call it testatspi2.c instead, because we could no longer use ATK API to test things, but AT-SPI2 instead, since the unit tests and the ATK world would be living in different processes.
 - Converting those tests into layout tests is a much better approach (see my previous comment)

I recommend you to read a post I wrote last year (and not because of self-promotion purposes), I guess you might find it handy to understand some of the messy concepts here :). Here's the link: http://mariospr.org/2013/02/03/accessibility-in-webkitgtk/

> Wording it differently: testatk.c (and other files from this directory) 
> covered a set of tests, that were not covered by other tests in webkit.

That is right but, as mentioned in my previous comment, the best idea would be to continue Joanie's work on converting those leftovers in testatk.c into layout tests, so we can keep testing those cases in upstream ports based in WebKit2, such as WebKit2GTK+ and EFLWebKit.

> Do we want to drop this set?
> 
Of course not, and I would have gladly worked on this myself if I could, but has not been possible so far :(

And that's part of the reason why I'm so happy to help you with this by giving you the right pointers, because if you could help with that that would be really great (and I'm sure Joanie would agree with that too).

> If I don't make a mistake, it may be better to answer this on webkit-gtk.

I'm fine if you want to move the discussion to webkit-gtk, yet I'm fine as well with keeping discussing here too. Joanie is on CC in this bug anyway and I'm adding apinheiro (ATK maintainer) now, to keep him in the loop

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.