Re: Hello/Still Noticing Slight Performance Decrease When Navigating with NVDA 2018.4


Tony Malykh
 

David,
I believe speech refactoring project is not going to improve latency.
It will rather add some new features to speech API of NVDA.
--Tony

On 12/19/18, David Goldfield <david.goldfield@outlook.com> wrote:
Tony,

Thank you for confirming what I have noticed. The performance hit I'm seeing
is significant enough that I really prefer what the previous version.
However, I am impressed with the experimental UIA support which I enabled in
my NVDA.ini file for use with Word. I definitely notice the difference,
particularly on my machine which is not the fastest machine around. What I'd
really like to see most in NVDA is a refactoring of the speech to improve
latency and, dare I make this comparison, to make it behave more like JAWS.
Specifically, I'm referring to the responsiveness when navigating using the
cursor and when typing.

David Goldfield, Assistive Technology Specialist
WWW.David-Goldfield.Com<http://WWW.David-Goldfield.Com>
On 12/18/2018 1:53 PM, Tony Malykh wrote:

I noticed this in Chrome too - 2018.4 is slightly slower than
2018.3.2. All the betas of 2018.4 were slower as well. However, with
2018.3.2 I had a problem of weird behavior with missing keystrokes in
browse mode. So it seems to me some bugs were fixed that both solved
my problem and made chrome a bit slower.

On 12/17/18, David Goldfield
<david.goldfield@outlook.com><mailto:david.goldfield@outlook.com> wrote:


My guess is that nobody else is experiencing this issue as I'm seeing no
other posts about this.

When I navigate by links with K or shift K in Thunderbird and even, to an
extent, in Chrome I notice a slight slowdown in responsiveness compared to
2018.3. For a short time I thought it was my computer as I was sometimes
seeing this occur in 2018.3.2 but, in general, 2018.3.2 is still somewhat
faster. I sent Quentin a debug file with all addons disabled but I'm
assuming there was nothing obvious in the output as I haven't heard back. I
notice this with Espeak. On a faster machine I do not notice this but on a
slower Athlon 2.9 GHZ there is a slight but noticeable difference. Just
wondering if anybody has noticed this and if something changed in the code
regarding navigation?

--
David Goldfield, Assistive Technology Specialist
https://nam05.safelinks.protection.outlook.com/?url=WWW.David-Goldfield.Com&;data=02%7C01%7C%7C02baf334e89b435b7a8008d6651a2b34%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C636807560394078993&amp;sdata=VQyb5%2B5iHZPxlsV9qvHkN5yCHwYNnI1uT%2FTfEq%2B8qkA%3D&amp;reserved=0<https://nam05.safelinks.protection.outlook.com/?url=http%3A%2F%2FWWW.David-Goldfield.Com&;data=02%7C01%7C%7C02baf334e89b435b7a8008d6651a2b34%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C636807560394078993&amp;sdata=suBnh5KjAV4okePbqYQnpsXB6sIO691ItYJEhhM1enU%3D&amp;reserved=0><https://nam05.safelinks.protection.outlook.com/?url=http%3A%2F%2FWWW.David-Goldfield.Com&;data=02%7C01%7C%7C02baf334e89b435b7a8008d6651a2b34%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C636807560394078993&amp;sdata=suBnh5KjAV4okePbqYQnpsXB6sIO691ItYJEhhM1enU%3D&amp;reserved=0>













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