Re: Futore porting of NVDA to Python 3.X. Why?

Mário Navarro

Yes, narrator is the future on windows.

more and more I am using narrator.

narrator is becoming an excellent screen reader.
microsoft is having a fantastic behavior with accessibility.
Às 21:27 de 26/12/2017, Shaun Everiss escreveu:

Having a student version of windows which is what s is is well interesting.

However for a full user you will either want to upgrade at cost to pro or avoid s completely.

But for a student or something that just wants a tablet and really doesn't need the extra stuff maybe an old person this may work.

Its just that we are all normal users and geeks here.

I for one will be getting nvda from the usual place.

On 26/12/2017 11:05 p.m., Brian's Mail list account via Groups.Io wrote:
Yes, But what is the actual point of having nvda in the Windows store in the first place?
From what I can see, most users will need to run add ons, which will mean running it as a normal app, and in that case they might as well get it from nvaccess.
The only people who will benefit from it being in the store are those running Windows 10S, the rubbish restricted hands tied behind back version of the operating system. Surely in this context, Narrator itself is going to have better access, being part of the operating system, albeit a bit slow and clunky.
I also don't understand how, for example, Google Chrome can actually be allowed to have a mere stub to load in normal Chrome in the Microsoft store.
If that is a way around it, why are we bothering to do all the work? One assumes Chrome is going to be the same as nvda, in that it can either run with its hands tied or normally in the same code. I'm sure this whole mess with Firefox is due to the same strange concept of Microsofts.
As I said in a previous post, most people are distrustful of Microsofts real agenda here, and so many would choose to continue with normal Windows, Normal Office and normal apps.
The other down side of so called universal apps and the restrictions of the system is that they seem to have abandoned a user interface that is the same for all. Its a ruddy free for all in this area, making apps that run under it really hard to learn as one has to throw all knowledge out and then try to understand what a particular author has in fact done.

I'm not having a go at nvda here, as we have to live in the world as it is, its just that for many blind people we feel that the rug is being pulled out from below us just as we are getting equal access.
Sent via blueyonder.
Please address personal email to:-, putting 'Brian Gaff'
in the display name field.
----- Original Message ----- From: "Joseph Lee" <>
To: <>
Sent: Monday, December 25, 2017 11:51 AM
Subject: Re: [nvda] Futore porting of NVDA to Python 3.X. Why?

Hi everyone,
Ah, an interesting question on the morning of Christmas (where it is past 3
AM my time)...
A bit of explaining is in order:
Regarding dropping support for Windows releases prior to 7 SP1: for a long
time, folks using Firefox and other web browsers experienced a problem where
NVDA's browse mode functionality wouldn't work when you restart NVDA while
focused on the browser window. In order to fix this, NV Access turned to
using some things from Windows API that isn't part of old Windows releases.
Because of this and other factors outlined below, NV Access wrote in August
2017 that NVDA 2017.3 will be the last release to support old Windows
Another factor is Windows OneCore rate boost issue. Currently in order to
use OneCore voices (on Windows 10 only) with faster speech rate, you have to
go to Settings, go to Ease of Access/Narrator and change the speech rate to
a faster value. A fix is now available but only on Windows 10 Version 1709
(Fall Creators Update), and incorporating the fix requires us (NV Access and
other developers) to use latest Windows 10 SDK, which will work only on an
update to Visual Studio 2017. Unfortunately, this meant giving up ability to
compile NVDA so it can run on old Windows releases.
Last one for now: a few days ago, you may recall a message where I told some
people to "shhh for now" over something under active development, and I
hinted on Twitter that you'll meet NVDA on a new outlet. For those who
solved the puzzle, great. For the rest of you: one day, you'll find yourself
opening Microsoft Store app on your Windows 10 S computer, searching for and
installing a Windows Store (aka Project Centennial) version of NVDA. This
also answers a question some of you may have had: yes, the Windows Store
version of NVDA CANNOT run add-ons at this time, but that could change as
development progresses. I won't go into details on mechanics of how this can
be done, but suffice to say that those running latest next branch snapshots
are already running a modified code that lets NVDA detect if it's running
inside a modified container. Fortunately for now, the old desktop version
code still lives, but once the Store version of NVDA ships, this will mean
saying goodbye to old technologies that were used on old Windows releases
(and the Store version and the desktop edition will still be together).
This is sort of an interesting segue to the question at hand: why Python 3?
The biggest advantage is ease of making NVDA speak and understand more
languages through extensive use of Unicode. One of our goals (developers,
and in extension, the community at large) is to let more blind people taste
what it is like to work and play with minimal or no financial barriers, and
internationalization is the key (this is why I kept asking for folks to help
out with translations). Python 3.x changes the game by shipping with
built-in support for Unicode, something Python 2 does not do well (hence the
need to use the Unicode function when needed).
Of course upgrading to Python 3 comes with downsides. Although we'll gain
native Unicode support, code must be edited and checked to make sure things
are working for folks as before. Unfortunately, there is a dark cloud over
us: add-ons, and I and community leaders are mostly to blame: we lost
contact with creators of some prominent add-ons, there are add-ons installed
on many NVDA installations that weren't updated in a very long time, and
add-on repositories are scattered all over the internet. As much as add-ons
are the sauce that binds the community together (among other things), our
lack of coordination, coupled with ones that won't be ported to Python 3
easily saddens me, knowing that this will be our undoing. Thankfully, some
in the add-ons community have recognized this early and are working
tirelessly to make sure that our add-ons are Python 3 ready.
Regarding Python 3 readiness of add-ons: mostly for add-ons community, but
effective March 1, 2018, any add-on I'll be reviewing must show that it is
python 3 ready, otherwise I'll ask authors to "transform" their code before
asking for another round of reviews. As for details, I'll post on the
add-ons mailing list, as it mostly concerns source code edits. For users,
this is so that your favorite add-ons can run on future NVDA versions
powered by Python 3.

-----Original Message-----
From: [] On Behalf Of Brian's
Mail list account via Groups.Io
Sent: Monday, December 25, 2017 2:34 AM
Subject: [nvda] Futore porting of NVDA to Python 3.X. Why?

Hi folks. I see a lot of issues and chat about doing this on github and
other places, however what seems to be lacking in these mostly technical
discussions, is a reason for doing it.
Is there anybody out there who understands the reasons well enough to
explain to the user who has probably not go a lot of understanding of
computer languages, exactly why this seems to be being prioritised over
getting nvda to work better as it stands.
I ask as to me at least, unless there is a need to rewrite whole sections
for some reason, it does seem a lot of work and will obviously slow down
development and indeed create bugs or remove functions accidentally.
We have just been through the dropping of support for XP, a decision not
universally popular from what I have heard, but obvious when you here
somebody explain why in as plain a language as one can. If somebody could
do the same for this major move it would I think go a long way toward
calming the frustrations some feel at the moment.
Oh and don't shoot me for saying this, its the season of Good Will you know!

Sent via blueyonder.
Please address personal email to:-, putting 'Brian Gaff'
in the display name field.



Join to automatically receive all group messages.