Re: [g-a-devel] RFC: AtkText simplification (take 2)

[Mario Sanchez Prada <mario prada samsung com>]

> On 06/24/2013 02:26 PM, Mario Sanchez Prada wrote:
> > > [...]
> > > So, taking into account that you are already working on this, I
> > > assume that you would prefer to add the generic boundary now, in
> > > order to avoid two updates of the implementation of those APIs.
> Right?

Yes, that would be my preference. However, I just realized that the ATK
version bumped in WebKit has been recently raised to 2.8.0 and so I don't
think raising it again to 2.10 is something likely to happen in the short
term, so I'm afraid we'll have to coexist with the old API for a while
anyway (meaning that I will need to provide implementation for those methods
anyway in the patches I'm currently working on).

> > It's not that I prefer to do that now in order to avoid two updates
> of
> > the implementation of those APIs, but just because I think it's
> > clearer and less confusing from the POV of the new API.
>
> Ok, that is not a valid reason for you, but in any case you didn't
> answer my question. So I will not add the reasons to the question, do
> you (as one of the atk implementors) prefer to add the generic boundary
> now (so probably for next ATK release) or not?

Yes.

Still, as I mention above, we will probably not be able to change WebKitGTK+
yet to match the new API because raising ATK dependency up to 2.10 will
probably not be a good idea right now.

> > Furthermore, I also think it should not be a big deal to have those
> > new boundaries co-exist with the old ones as long as it's clearly
> > stated in the documentation what is deprecated and what is not. Or
> you
> > can use some kind of aliases like what Trevor is suggesting.
>
> Well, it is just less clear, IMHO, having two enums that means the
> same.

I understand your point, but I see it as a temporary evil while we move to
the new API, and not nothing too severe, since we would not be breaking
backward compatibility or the like, just preparing for the future.

> Anyway I agree with Trevor suggestion (will answer him in short).

Me too.

Mario