New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
To have braille insure that lists and such consistently present the start of items in the same place on the display #217
Comments
Comment 1 by jteh on 2008-10-31 13:26
|
Comment 2 by valiant8086 (in reply to comment 1) on 2008-10-31 15:53
I'll try to think up some situations here max battery li choose a power scheme tx power options dlg Replying to jteh:
|
CC @LeonarddeR, @dkager. I'd appreciate your thoughts on #217 (comment). There's also another option: determine what to focus at the left of the display based on new focus ancestry, just as we do for speech. For example, if you open a simple dialog, you might see "Prompt dlg Yes btn", but if you press tab, "No btn" would be all you see; you'd have to scroll back to see the dialog. This means you do know that context has changed when it's significant, but you wouldn't have to deal with items constantly moving about in a menu, for example. |
I probably missed this issue since it is so old, so sorry for the duplicate. Anyway
I agree with this and would be happy to find out how this can be realised.
I agree, but the proposed behaviour isn't strange. It is actually the default behaviour in JAWS. It shows everything in reversed order, so first you see the button, followed by the message and the dialog title. It has its charms, but not more than that, and I don't think we should consider this at this point. @dkager, any thoughts about this? |
Hmm, I really like this one. It is kind of an intermediate between always showing the ancestry and always having to scroll back to see it. So either I'd go for this one and the current behaviour, or allow the users to choose between the three different behaviours. |
So I like seeing context and I also like my focus text to start in the same, predictable place. Based on this, I have one more insane option. I suppose it's like status cells. |
I also like the intelligent approach of placing focus hard left if the ancestry hasn't changed. This requires more training time for it to become intuitive, but after that it is probably the most efficient (and novel, which doesn't hurt). |
@leonardder commented on [May 25, 2017, 4:59 PM GMT+10]
I assume you'd still want it to show as much of the focused control as possible as we do now? For example, if a dialog appeared and focused a button and both were too long to fit on the display, you'd still expect to have to scroll back to see the full dialog text? |
@leonardder commented on May 25, 2017, 4:55 PM GMT+10:
I think this should be fairly simple. In In contrast, now that I've grabbed your interest with this other idea of scrolling based on ancestry changes, I'm not sure how that would be implemented. :( |
@dkager commented:
Honestly, I wouldn't use such functionality since it will probably always force me to scroll somehow. Would you prefer this one over the three other approaches mentioned here, the current approach, focus to hard left and the hybrid of these? @jcsteh commented:
This is correct indeed. |
There is a parallel here. Do you always listen to everything speech has to say? That was my motivationfor suggesting the midway point. In a GUI that you know reasonably well, you should have enough with a bit of context and a bit of focus. Note that I haven't tested this, maybe it really doesn't work well.
In my order of preference:
I'm not sure where the midway suggestion fits in without trying it first. Now, if you'd asked me last year I would probably have put the hard left solution up much higher. But that's mostly because I liked SuperNova's "physical mode", which is still more verbose than only seeing the focus. Long story short: can't really decide without trying the all for a month or so. |
Reported by valiant8086 on 2008-10-31 12:23
When you arrow through a list of items and are reading the braille display, the start of each list item may begin in a different place for one list item than it is for another. This makes it more difficult to easily arrow down through a list in search of something, because you have to stop and read the display to be able to find the list item. Right now, NVDA tries to make sure the end of the text in such a situation is displayed that is, the display will be presented with the right end of the text. Any text not on the display for that line will usually be to the left, in theory, and the user would reverse advance to read it. I want to try and switch things around a little to see if we can at least do something to where the beginnings of list items and things don't move left and right depending on how long the text is. For example, Maybe we could show li on the left end of the display to indicate this is a list item. After that there would be one space, then the text of that list item. If I were to arrow up and down through the list, that li at the left end would not move, and the text of the list item would consistently begin in the same place on the display. This may mean to do this that braille will need to either not show things like dialogue names, or show them only after the currently focused control. Having this functionality is very important to me. I really feel strongly that it will make doing things useing braille much much more reasonable. I have thought long about whether we would start missing out on information by doing this and I don't think so. If we can just re-arange where things are displayed, we should be ok. This functionality should appear in all possible situations, list items, folder view list views, tree views, combo boxes in and out of virtual buffers, menu items, start menu, etc. Heck i think it should do that for buttons and checkboxes and things too. the idea is present the control that currently has focus on the left end, and put the rest of the information on the right of that. if I want to read which dialogue i'm in etc, I can advance to take a look, assuming the display didn't have enough room to show that before hand.
Blocking #4659
The text was updated successfully, but these errors were encountered: