[orca-list] Possible to improve continuous read in VSCode?

[Nolan Darilek <nolan thewordnerd info>]

I know, old version of Electron, etc. Hard to know whether this has 
since been fixed in Chromium and will be a non-issue in a few months, or 
if it slipped beneath the radar.


I like how VSCode restores previous sessions, along with positions in 
files. I started experimenting with it as a way to read text-based 
technical documents converted with pdftotext, since that makes a mess of 
the formatting. The idea is that, if I want better insight into a piece 
of code, I can copy it straight into another tab and have it 
automatically formatted on paste. I can also run it directly using 
extensions.


I notice that continuous read in the editor doesn't quite work. It works 
well enough once it starts, but it often jumps far up in the text, then 
starts reading from that point. I have Orca set to read by sentences, 
and it jumps further up than the previous sentence boundary. It does 
continue reading normally, though.


I've also set it to read by line. In this case it works better in that 
it doesn't jump, but it speaks the name of the editor between each line.


To reproduce:


1. Open VSCode. (I've only tested with latest stable, not Insiders, 
though I have no reason to think Insiders would behave differently.)

2. Open a long text document in the editor.

3. Focus on the editor and switch to Focus mode.

4. Arrow to a position lower down in the text and start continuous-read.

5. For me, reading begins far above the cursor if Sentence mode is used. 
It begins on the same line as the cursor when Line mode is active, but 
the filename followed by "editor" is spoken between each line.


Thanks a bunch.