Re: New custom NVDA Add-on. Who ask to?


Daniele Casarola
 

Thanks Didier for your suggestions.

I think you are focusing the main points.

Me too, I'm not 100% sure about which is the best solution to manage this issue. I can only try to guess, because until you don't try, you never know indeed.

Reading individual letters while pressing the delete button, is not convenient. Yes you are right, basically because a speed point of view.

I can identify two ways for audio feedback when I delete a word: stroke by stroke or word by word.

1. Delete stroke by stroke. Which is the default I'm using now. As I said, it works like an "undo" or "ctrl+z", for the group of letters you type. In practice what I hear when I press the delete button, should be part of a word, red in the opposite direction. For example if I type "telefono" in 3 strokes "te, le, fono". Deleting I hear "fono, le, te".
It is comfortable? I don't know. I should try. Maybe is confusing, maybe not. Maybe with a bit of exercise is ok. I don't know.

2. Delete word by word. This gives to me the opportunity to introduce a nice Plover feature.
Plover's core is basically the dictionary. A dictionary is a text file where each line corresponds to an entry, with this simple syntax:
"Steno Machine Strokes": "output on the screen",
An example:
"Cp": "Codice Penale",
So if I press the keys "C" and "p" together on the keyboard, on the screen I read the words "Codice Penale".
N.B. In this case if I press the delete button, both the word are deleted, "Codice" and "Penale", because they has been created with one stroke.

Instead of a text output, I can assign to Steno Strokes also a QWERTY Keyboard combination of keys, like in this example:
"Cp": "{#Ctrl_L(Backspace)}",
In this case, pressing on the steno keyboard "C" and "p", it is the same as if I press on the QWERTY keyboard "CTRL+Backspace".

Basically I can simulate each combination of a standard QWERTY keyboard, using my steno machine.

I tell you more. If I check the option “Handle keys from other applications”, I also can hear the audio feedback when I simulate CTRL+Backspace with the steno keyboard. So I hear the whole word that I delete.

As you can imagine, there are some problems with this set, otherwise we would not be here.

Mainly I've found three issues:

1. The first depends of punctuation. Plover automatically add the space between words. I don't have to press any spacebar. The space is automatically added after the word or punctuation. If I press CTRL+Backspace, an empti space is left with the previous word. So I can continue to type with any problem, the new word I will type after pressing CTRL+Backspace will not be attached to the previous one.
But in case I delete a punctuation like a comma, or a period, or a question mark, using CTRL+Backspace, the cursor stops attached to the previous word, so the word I will type will be attached to the previous one. You can try on a Wordpad application.

2. The second is basically the same as the previous one. It regards specifically Microsoft Word. I don't know why, anybody knows why, probably also the Microsoft Team knows why, but if you press CTRL+Backspace in Microsoft Word (the last versions), sometime the empty space is left with the previous word, sometime it isn't.
Looking on the web it seems a problem other people had, but I didn't found the solution. I mean it not only depends of punctuation, like in a normal case, but also between words with no punctuation.
Unfortunatly I cannot bypass Microsoft Word. It is the application all companies use. There are some macro inside we use to produce the final document.

3. This problem appears only if “Handle keys from other applications” option is checked. If I uncheck I solve the problem, but of course NVDA doesn't read in case of a CTRL+Backspace simulation,.
Before to type what a speaker says, I have to insert his qualification and name and surname. This is an example in case a witness called "Mario Rossi" speaks (in italian witness is "testimone"):

TESTIMONE ROSSI M. - Bla bla bla

I have a specific stroke to type fast the speaker's name/qualification. This is the syntax on the Plover dictionary:
"I/Tshr": "TESTIMONE ROSSI M. -",

So if I type in the first stroke the letter "I", and in the second the four keys "Tshr", the output on the screen is "TESTIMONE ROSSI M. -"

I've noticed that using this kind of entries, basically when the output is a long serie of characters instead of one or few syllables, in this case the output can be modified. It happens randomly. Without a specific logic I can understand. For example the output can be:

TESTIMONEEE ROSSI M. -
or
TESTIMONE ROSSI_M. -
or
TESTIMONE ROSSIII M... -

It is not a big mistake. But I cannot deliver an official document in this way. Sometimes the spellchecker tool is not able to find them. So “Handle keys from other applications” checked, modifys a little bit the output of my machine.

That's all.
In my opinion, maybe it is better to investigate about the second solution. It means to delete "word by word", using a "CTRL+Backspace" simulation, as also Sandra suggested. I think is more comfortable for audio feedback, instead of hear "syllable by syllable" which could be a bit confusing. About speed, I don't think there is a big waste of time, in some cases it could be also faster.
The wall is the three issues listed above.

Thanks again for reading.

Daniele.

Join nvda@nvda.groups.io to automatically receive all group messages.