[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: [orca-list] Keybindings for ARIA landmark roles



On Fri, 2009-01-30 at 13:39 -0500, Willie Walker wrote:
> Hi Krishnakant:
> Hi Will:
> Even discovering if there is one landmark on a page can be a very 
> expensive operation.  What has to be done is that Orca needs to query 
> each accessible object for its attributes and then look at each 
> attribute to see if it matches one of the ARIA landmark roles.
> 
I agree, however given the technical specs you just provided I feel it
would be a relative asumtion that performance wil be always hit.

In the case of counting how many landmarks are there on the page, you
will have to loop through all the objects and find for landmarks.

But in the case of what I had suggested, the querying loop will
terminate on the first occurance of the landmark.

happy hacking.
Krishnakant.
> Will
> 
> Krishnakant wrote:
> > Hi Will,
> > If performance is going to be so much of an issue with landmarks, then
> > out of my accessibility research i would suggest that orca at least say
> > "page has landmarks " as a message.
> > 
> > What is important is that the user knows that there are one or more
> > landmark on the page the facility will be usefull with reference to this
> > page.
> > 
> > I think at least this might not be a payoff on performance?
> > 
> > happy hacking.
> > Krishnakant.
> >  
> > On Fri, 2009-01-30 at 13:25 -0500, Willie Walker wrote:
> >> Hi All:
> >>
> >> We hear your desires loud and clear and are looking at ways to make this 
> >> performant.  Right now, it can take a very long time to gather this 
> >> information, however, and we have decided to err on the side of trying 
> >> to reduce the number of things that slow Orca down.
> >>
> >> BTW, counting up all the landmarks on a page has the potential to be a 
> >> very expensive operation -- much worse than counting up thing like 
> >> links, headings, etc.
> >>
> >> As I mention, however, we are looking at ways to potentially make things 
> >> more performant in Orca, which could give us a performance boost overall 
> >> and open the door for other things like the above.
> >>
> >> Will
> >>
> >> Steve Holmes wrote:
> >>> -----BEGIN PGP SIGNED MESSAGE-----
> >>> Hash: RIPEMD160
> >>>
> >>> I agree; I would not only want to know some characteristics of the
> >>> page but also be told when I navigate into things like forms or
> >>> tables.  Especially tables because then you know you can start using
> >>> the table navigation features.  But this probably  slightly off topic
> >>> from this thread but since the page contents issue came up...
> >>>
> >>> I have run into Krichnakant's situation numerous times when trying to
> >>> find headings.  I guess one could press the numpad-enter key twice to
> >>> hear the counts and your landmarks count could go in there.  But I
> >>> forget half the time to hit that key sequence but that does get us
> >>> around having to have it built in the page loading sequence.
> >>>
> >>> On Fri, Jan 30, 2009 at 11:25:57PM +0530, Krishnakant wrote:
> >>>> hi will,On Fri, 2009-01-30 at 10:14 -0500, Willie Walker wrote:
> >>>>> Hi Krishnakant:
> >>>>>
> >>>>> Determining if there are landmarks can be an expensive operation and can 
> >>>>> slow performance down.  So, it might be best to only do landmark 
> >>>>> operations when the user explicitly requests them.
> >>>>>
> >>>> But how does user know if there are area landmarks defined by the
> >>>> designer of that page in the first place?
> >>>>
> >>>> The first rule of Accessibility with Human Computer Interface Design is
> >>>> to facilitate user's anticipation about the structure of the page.
> >>>>
> >>>> For example Imagine if there is a large page and I want to know if there
> >>>> are headings.  Now, i press h and i wait for a few secons or more
> >>>> depending on the size of the page, just to hear "wrapping to top, no
> >>>> more headings found ".
> >>>>
> >>>> Had I known that the page has 0 headings, I won't try the command in the
> >>>> first place which any ways gave orca a performance hit.
> >>>>
> >>>> Same thing will apply for landmarks and I believe every one agrees that
> >>>> on a big page, orca will take a performance hit and would slow down the
> >>>> system and overall response because orca will scann the entire page for
> >>>> the structural element which the user are trying to search.
> >>>>
> >>>> happy hacking.
> >>>> Krishnakant.
> >>>>> In any case, what I'm hearing is that you are joining the list of users 
> >>>>> who want default keybindings.  Do you think the 'm' key is reasonable 
> >>>>> for this?
> >>>>>
> >>>>> Will
> >>>>>
> >>>>> Krishnakant wrote:
> >>>>>> On Fri, 2009-01-30 at 09:46 -0500, Willie Walker wrote:
> >>>>>>> Hi Krishnakant:
> >>>>>>>
> >>>>>>> The content provider provides the landmarks as part of the markup.  What 
> >>>>>>> the Orca feature does is allow the users to navigate between them quickly.
> >>>>>>>
> >>>>>>> Will
> >>>>>>>
> >>>>>> hi will,
> >>>>>> this works perfectly.
> >>>>>> Just a quick suggestion, as soon as the user lands on a certain web
> >>>>>> page, we must announce if there are landmarks set by the page designer
> >>>>>> so that the user knows that the move to landmark (next and previous )
> >>>>>> can be useful on this page.
> >>>>>>
> >>>>>> Since orca is locating landmarks, i believe orca knows that there are
> >>>>>> defined landmarks on the page.
> >>>>>> I believe this is the way the scrip for mozila works?  or is it that the
> >>>>>> first time user presses the move to next landmark key that orca finds if
> >>>>>> there are any landmarks?
> >>>>>>
> >>>>>> happy hacking.
> >>>>>> Krishnakant.
> >>>>>>
> >>>> _______________________________________________
> >>>> Orca-list mailing list
> >>>> Orca-list gnome org
> >>>> http://mail.gnome.org/mailman/listinfo/orca-list
> >>>> Visit http://live.gnome.org/Orca for more information on Orca
> >>> -----BEGIN PGP SIGNATURE-----
> >>> Version: GnuPG v1.4.9 (GNU/Linux)
> >>>
> >>> iEYEAREDAAYFAkmDQaQACgkQWSjv55S0LfGpvwCfS4jOESZoql35QF9Cx2aRGDZd
> >>> IocAnR7Ot8V99ufl1NgK+fbcDjXBv/Jk
> >>> =0VDj
> >>> -----END PGP SIGNATURE-----
> >>> _______________________________________________
> >>> Orca-list mailing list
> >>> Orca-list gnome org
> >>> http://mail.gnome.org/mailman/listinfo/orca-list
> >>> Visit http://live.gnome.org/Orca for more information on Orca
> >> _______________________________________________
> >> Orca-list mailing list
> >> Orca-list gnome org
> >> http://mail.gnome.org/mailman/listinfo/orca-list
> >> Visit http://live.gnome.org/Orca for more information on Orca
> > 
> 



[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]