nvaccess / nvda

NVDA, the free and open source Screen Reader for Microsoft Windows
https://www.nvaccess.org/
Other
2.11k stars 637 forks source link

Formatting attributes should be seperated with a comma sign #3283

Open nvaccessAuto opened 11 years ago

nvaccessAuto commented 11 years ago

Reported by k_kolev1985 on 2013-06-14 14:11 I rarely use the NVDA functionality for reporting the formatting attributes with NVDA+F, but I've noticed something about it: the things (attributes) in that message (NVDA+F) are spoken by the synthesizers "in one breath", so to speak. And that makes the formatting information sometimes hard to understand (specially by novice users). This happens because those attributes are not separated one from another with any symbol. My suggestion is to separate them with a comma (,) sign, so the synthesizers will make a small pause between them and that will make them more understandable, in my opinion. What do you think?

Thanks in advance!

LeonarddeR commented 7 years ago

This is probably very easy to fix, and I think I agree.

derekriemer commented 7 years ago

we now have nvda+f twice though.

jcsteh commented 7 years ago

While this is certainly a lot of info to rattle off in one go, inserting commas would make it a lot slower to get through all the info. Note that NVDA+f twice will display each piece on a separate line, so if you need to review it closely, that's probably the better option.

@QChristensen, @MichaelDCurran, thoughts?

Qchristensen commented 7 years ago

I think having NVDA+f twice now makes this less urgent and does mean that those who want to skim through the information to quickly find one attribute can do so. While those who want to go through each piece of information can do that as well now.

There aren't too many other places where we read numerous pieces of information in one block like this. Otherwise perhaps a setting in maybe Voice Settings to "separate grouped information with" and have the options of a comma, full stop or space.

ehollig commented 7 years ago

@feerrenrut, @Qchristensen, coming back to this, I would tend to agree with @jcsteh in https://github.com/nvaccess/nvda/issues/3283#issuecomment-313974730. Thoughts?

derekriemer commented 7 years ago

Why not have it seperated with a PauseCommand of given length? If there's not a pauseCommand, we could make one.

jcsteh commented 7 years ago

We do have a BreakCommand which inserts a pause of specified length (though it's probably not supported by third party synths). Pauses of specific length are problematic with regard to different rates, though.

bhavyashah commented 4 years ago

When I press NVDA +Delete, the three pieces of information are separated by pauses of just the right length. Can we copy the same approach for NVDA+F? I agree that this isn't high priority though given NVDA+F twice.

Qchristensen commented 4 years ago

I agree the specific original case of separating attributes with comma (or other pause) when using NVDA+f is now effectively negated by being able to press that twice to open a window with the information you can move through as you like.

I do think, however, that it would be worth being consistent. As @bhavyashah notes, NVDA+delete puts a comma between pieces of information.

NVDA+TAB (report focus) and NVDA+end (Status bar) both add an extra space in between elements (that doesn't adjust the length of time between reading, at least with eSpeak-NG but could that vary between synthesizers?) Pausing between reading elements specified in NVDA's document formatting when moving through text uses the same (two spaces between elements).

I know there are other examples I haven't thought of I think having the option to choose how much to pause between reading multiple pieces of information would be useful. I expect there would be those who would be happy of the setting in order to remove pauses as much as possible, and those who would be happy because they could insert a longer pause.

Actually I could see the case for having a longer pause between document formatting changes when proof reading in say Word (which could change to anything), compared to element types on a web page. In such cases, having the one option would enable this to be saved in a configuration profile so it could be varied between programs.

XLTechie commented 4 years ago

For what it's worth, I would like to agree with @qChristensen's comments and proposal here.

Adriani90 commented 4 years ago

I think a human similar pause would be enough here to meet every user expectation, the ones who don't like longer pauses could set the speech rate to a higher speed and the pause will be shorter anyway then. I guess a komma between the expressions should suffice here, otherwise the work on this would be really huge if you want to make every pause configurable. I don't know if this is worthy to do. Note that you can also stop the speech with shift and resume it again also with shift, so that you already can adjust the pause between expressions manually without changing any setting.

CyrilleB79 commented 3 years ago

I would also agree to insert commas between each field. Jaws has such pauses for its equivalent script and I remember having difficulty with the absence of pause in NVDA+F when transitioning from Jaws to NVDA.

IMO, with a high rate, it does not lengthen too much the time to get all the information. @jcsteh do you have a precise example where the length of text spoken with and without comas differ a lot? (specific synth, specific rate) Also, it should be noted that punctuation in OneCore were making too long pauses in speech not so much time ago. But this is now fixed.

Regarding the double space question, I guess it is generally used in NVDA to separate things that should be handled independently by the synthesizer. In some case, a single space is not enough, e.g. when space is used as thousand separator when reading number in some languages such as French or German.