I guess there must be a performance penalty from polling the mouse so often (can't imagine why that setting exists if there isn't one) and that value is in milliseconds, so 1/1000 of a second. Assuming a screen refresh rate of 60Hz each screen refresh takes 16ms so I think if the setting is put at 15ms then a mouse poll will happen during every screen refresh cycle. From my testing I can't tell the difference between a setting of 15 and a setting of 1. Whatever the performance hit is (I couldn't detect one but a low performance netbook might struggle) I would imagine that a setting of 15 would be easier to argue to move to than 1 because its only a change of about a factor of 3 rather than a factor of 50.
Is 15 as good as 1 for you?

Alan.

On 26/01/12 02:01, Robert cole wrote:
Hello, everyone.

Well, I did a slight bit of tinkering with the GNOME Shell Magnifier settings today (I love that I can tinker in Linux!).

I am not trying to complain by any means, so I hope that this does not sound as such. I have been trying out Unity2D in Ubuntu 11.10, which I recently installed onto my desktop system to replace Linux Mint 12. I love how the GNOME Shell Magnifier is coming along, but at the present time it does not pan as smoothly as I am used to, which makes it slightly more difficult to read and to navigate my system. The same can be said for panning using Compiz's eZoom plugin on a fresh install of Ubuntu--it does not pan so smoothly either.

Well...I learned through experience that I could change the value for Compiz's Mouse position polling plugin to smoothen things out. (Just for the record, the GNOME Shell Magnifier pans much more smoothly by default as compared to Compiz's default settings under Unity.) Anyway, I learned that if I decrease the Mouse Poll Interval from 40 to 1 panning was very smooth...well...just beautiful to me.

So today, rather than bother the list by asking where I could find the settings files for the Zoom feature, I figured I'd do some exploration. I ran the basic command:

locate magnifier

and this file stood out to me:
/usr/share/gnome-shell/js/ui/magnifier.js

I am not a professional programmer, but I figured I'd take a look at this file. I fired up nano (after making a backup of the file to my /home directory) and began reading. I found the following line under the first set of functions at the top of the file (section comment below):

// Keep enums in sync with GSettings schemas
[...]
const MOUSE_POLL_FREQUENCY = 50;

So I contemplated whether I should change this value from 50 to 1. I gave in and did it!

I made the changes from within Unity2D, so I logged out and then logged into GNOME Shell. I crossed my fingers and then hit CTRL+ALT+M, the keyboard shortcut I set to activate the magnifier...and...VOILA! Super smooth panning! This made me love GNOME Shell even more than I already do!

My questions as concerning this thread, though, is this: How likely is it that decreasing this value from 50 to 1 will cause possible instabilities? Is it set to 50 for any given reason (I am sure there has to be a reason)? I ask this because it is the same way with Compiz, but it seems (in either case) that the lower the Poll Interval (Compiz) or Poll Frequency (GNOME shell Magnifier) the more smoothly panning is in fullscreen magnification mode.

In any case, I just wanted to put this on the list to get some input. As I mentioned earlier, I am definitely not a professional programmer. I took classes in C++, Visual Basic, and Python throughout the years, but I the class content was quite basic.

thanks for any input regarding this subject. I'm really looking forward to GNOME 3.4!

Take care.
_______________________________________________
gnome-accessibility-list mailing list
gnome-accessibility-l...@gnome.org
http://mail.gnome.org/mailman/listinfo/gnome-accessibility-list


--
The Open Learning Centre is rebranding, find out about our new name and look at 
http://libertus.co.uk


--
Ubuntu-accessibility mailing list
Ubuntu-accessibility@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-accessibility

Reply via email to