Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
OT:OMG! Mozilla enabled their broken rendering engine for Firefox 89.0.
864 views
Skip to first unread message
Rod Pemberton
Jun 3, 2021, 8:21:57 PM6/3/21
to
<OT>
OMG! Mozilla enabled their broken rendering engine for Firefox 89.0.
I know this is off-topic, but some of you may be using the Firefox
browser and have this issue now.
Go to MotionMark link below and run their v1.1 benchmark. Your result
should **NOT** be a low number, like below 10. This takes a few
minutes to run.
https://browserbench.org/MotionMark1.1/
If you get a low number, you'll need to go to https://about:config, and
toggle this Firefox setting to true:
gfx.webrender.force-disabled
Also, make sure that these Firefox settings are still set to false:
gfx.webrender.all
gfx.webrender.enabled
There will be other gfx.webrender.* settings that are now set to true
by default for Firefox 89.0. You can try toggling them one-by-one and
checking the benchmark, if you don't want to fully disable Webrender
with the force disable setting. I'll attempt this at a later time.
E.g., MotionMark1.1 results for my 2009 components computer:
89.0 74+ without Webrender - 100% correctly rendered
89.0 4.33 with Webrender - ~70% correctly rendered
88.0.1 72+ without Webrender - 100% correctly rendered
88.0.1 1.08 with Webrender - ~30% correctly rendered
As you can see, they've improved Webrender from 88.0.1 to 89.0, but it's
still not rendering MotionMark's benchmark correctly or completely.
Webrender is really slow too. The most noticeable rendering which is
still broken is of the circles with lines moving in out about their
radius, as all the lines are missing ...
So, if you've got an older machine, like I do, or if their Webrender is
just broken ... (which it very likely is), you may want to try out the
browser setting changes above.
--
Let's ask the radical leftists something. After Joe Biden's sick
sexual comment about a really young girl, does PizzaGate live?
wolfgang kern
Jun 5, 2021, 6:53:39 AM6/5/21
to
On 04.06.2021 03:23, Rod Pemberton wrote:
> OMG! Mozilla enabled their broken rendering engine for Firefox 89.0.
I could avoid any upgrade until today, but it seems that Mozilla guys
> OMG! Mozilla enabled their broken rendering engine for Firefox 89.0.
think they are gods and force us to take all their new BS every week.
So my Firefox just stopped working without the latest stuff.
Thunderbird upgrades are also frequent and not only redundant but often
make it more worse than it is already. Far away from once user friendly.
do we have other better options ?
__
wolfgang
Rod Pemberton
Jun 5, 2021, 12:05:19 PM6/5/21
to
Which browser feature(s) is most important to you?
For me, other than general web page browsing, I would say the most
important things are bookmarks and video. I want properly working
video without glitches or choppiness or dropped frames, i.e., Youtube,
e.g., 1080p60. I've been having a problem with 60 fps video,
especially 1080p, since Firefox apparently reverted to older code about
version 87, maybe 85, or thereabouts. Firefox fixed this problem once
in the past.
M$ Edge is only available for Windows, as a replacement for Internet
Explorer. It seems to be Internet Explorer 6 to me, without bookmarks.
Chrome is considered to be the leading web browser. However, I was
never able to get Chrome to run on this version of Linux (Slackware
based). There was something wrong with their sand-boxing. It would
crash or abort. They also blocked the root account on Linux from
using it. I understand their security concerns but that decision should
be up to the system administrator, especially if they want to be able
to test it as root.
Firefox has stopped supporting ftp connections as being obsolete or
insecure, but it allows you to pass the link to another browser or
program, which I pass to an old version of Opera. Most ftp-only
websites are available via http/https now. Just change ftp to http or
https. I only have two remaining links that are still ftp-only, and
they're probably mirrored elsewhere.
I used the Opera browser for a long time on Windows 98/SE, after
Internet Explorer 5. I liked it. It was easy to use and had all the
features of Internet Explorer which I used at the time. IIRC, Opera
wasn't available for 64-bit Linux some years ago. Opera seems to be
available for Windows, Mac, and Linux now, but I'm not sure if their
Linux versions work with this Linux, as they're packaged for other
Linux distro's.
IIRC, most modern browsers have stopped supporting bookmarks, or they
might be disabled by default. Firefox still supports bookmarks. Most
people only have a few hundred bookmarks. I probably didn't have more
than 500 bookmarks for a very long time. Now, I have about 30 thousand
bookmarked links to about 20 thousand websites. Not being able to use
bookmarks would be a huge disadvantage to me, as I use them like a
large, searchable database of company websites.
--
Scott Lurndal
Jun 5, 2021, 12:40:33 PM6/5/21
to
I've no issues with firefox 89 so far, but I have firefox 38 for a backup.
Rod Pemberton
Jun 5, 2021, 10:17:12 PM6/5/21
to
memory just to display 230MB of web page content for 35 web pages, which
is a ratio of about 12 times. That also causes this version of Linux to
chew up 500KB of swap space.
--
Bernhard Schornak
Jun 6, 2021, 12:56:44 AM6/6/21
to
client, Seamonkey offers both. Firefox is the worst choice - I only use
it to search for things while I'm writing YouTube comments (I hate tabs
- something no one really needs).
Pfüat'Di
Bernhard
wolfgang kern
Jun 6, 2021, 6:18:11 AM6/6/21
to
I like the tabs which allow me to run 8 games in parallel, so I can skip
and don't need to wait for these often long lasting reloads.
Servas,
__
wolfgang
James Harris
Jun 8, 2021, 6:15:42 AM6/8/21
to
On 04/06/2021 02:23, Rod Pemberton wrote:
>
> <OT>
>
> OMG! Mozilla enabled their broken rendering engine for Firefox 89.0.
>
> I know this is off-topic, but some of you may be using the Firefox
> browser and have this issue now.
>
>
> Go to MotionMark link below and run their v1.1 benchmark. Your result
> should **NOT** be a low number, like below 10. This takes a few
> minutes to run.
>
> https://browserbench.org/MotionMark1.1/
>
>
> If you get a low number, you'll need to go to https://about:config, and
> toggle this Firefox setting to true:
>
> gfx.webrender.force-disabled
...
>
> <OT>
>
> OMG! Mozilla enabled their broken rendering engine for Firefox 89.0.
>
> I know this is off-topic, but some of you may be using the Firefox
> browser and have this issue now.
>
>
> Go to MotionMark link below and run their v1.1 benchmark. Your result
> should **NOT** be a low number, like below 10. This takes a few
> minutes to run.
>
> https://browserbench.org/MotionMark1.1/
>
>
> If you get a low number, you'll need to go to https://about:config, and
> toggle this Firefox setting to true:
>
> gfx.webrender.force-disabled
> E.g., MotionMark1.1 results for my 2009 components computer:
>
> 89.0 74+ without Webrender - 100% correctly rendered
> 89.0 4.33 with Webrender - ~70% correctly rendered
>
> 88.0.1 72+ without Webrender - 100% correctly rendered
> 88.0.1 1.08 with Webrender - ~30% correctly rendered
disk at the moment:
6.47 with Webrender enabled
14.96 with Webrender disabled
Firefox 89.0 64-bit on Ubuntu 20.04.02.
Despite that, I've put gfx.webrender.force-disabled back to its default
state so I don't need to remember to change it back for future upgrades.
I don't need any such graphics rendering and the CPU usage in both tests
was about the same.
Have to say, though, that that was fascinating. I was amazed at how fast
the graphics were. And I didn't know a browser could do any such stuff!
--
James Harris
Rod Pemberton
Jun 17, 2021, 12:46:40 PM6/17/21
to
On Thu, 3 Jun 2021 20:23:19 -0500
Rod Pemberton <noe...@basdxcqvbe.com> wrote:
>
> <OT>
>
> OMG! Mozilla enabled their broken rendering engine for Firefox 89.0.
>
> [snip]
Rod Pemberton <noe...@basdxcqvbe.com> wrote:
>
> <OT>
>
> OMG! Mozilla enabled their broken rendering engine for Firefox 89.0.
>
>
> E.g., MotionMark1.1 results for my 2009 components computer:
>
> 89.0 74+ without Webrender - 100% correctly rendered
> 89.0 4.33 with Webrender - ~70% correctly rendered
>
> 88.0.1 72+ without Webrender - 100% correctly rendered
> 88.0.1 1.08 with Webrender - ~30% correctly rendered
>
>
> [snip]
> E.g., MotionMark1.1 results for my 2009 components computer:
>
> 89.0 74+ without Webrender - 100% correctly rendered
> 89.0 4.33 with Webrender - ~70% correctly rendered
>
> 88.0.1 72+ without Webrender - 100% correctly rendered
> 88.0.1 1.08 with Webrender - ~30% correctly rendered
>
>
Update:
89.0.1 76.18 without Webrender
89.0.1 4.19 with Webrender
AMD RS780 (DRM 2.40.0 / 3.18.16, LLVM 9.0.1)
3.0 Mesa 19.2.8
mesa/r600
--
What is hidden in the ground, when found, is hidden there again?
Rod Pemberton
Jun 28, 2021, 8:11:18 AM6/28/21
to
On Thu, 17 Jun 2021 12:48:01 -0500
89.0.2 71.33 without Webrender
89.0.2 4.44 with Webrender
Complete results:
FF Webrender OFF ON
MotionMark 71.33 4.44
Multiply 1.00 pass 1.00 fail
Canvas Arcs 352.20 pass 1.00 fail
Leaves 69.48 pass 84.13 pass
Paths 1156.07 pass 1.00 fail
Canvas Lines 1117.47 pass 1.00 fail
Focus 18.89 pass 19.80 pass
Images 41.47 pass 2.00 pass
Design 36.51 pass 2.74 pass
Suits 52.83 pass 73.29 pass
The higher the score, the more accurate and more complete the rendering
appeared (subjective).
The tests with Webrender on that had higher scores than with Webrender
off actually performed better visually (IMO). Unfortunately, many tests
with Webrender on failed to render properly. A couple of those with
Webrender on that passed weren't rendered quite as well as with
Webrender off. Obviously, those with very high scores with Webrender
off had serious defects or failure with Webrender on.
Now, I'm wondering if they're just optimizing Firefox for a few types
of graphics situations with Webrender, e.g., text and static images,
because for every test with Webrender on which had a high score,
Webrender on scores were slightly better than Webrender off. It's just
that all the other tests either failed or performed poorly with
Webrender on ...
Rod Pemberton <noe...@basdxcqvbe.com> wrote:
> On Thu, 3 Jun 2021 20:23:19 -0500
> Rod Pemberton <noe...@basdxcqvbe.com> wrote:
>
> >
> > <OT>
> >
> > OMG! Mozilla enabled their broken rendering engine for Firefox 89.0.
> >
> > [snip]
> >
> > E.g., MotionMark1.1 results for my 2009 components computer:
> >
> > 89.0 74+ without Webrender - 100% correctly rendered
> > 89.0 4.33 with Webrender - ~70% correctly rendered
> >
> > 88.0.1 72+ without Webrender - 100% correctly rendered
> > 88.0.1 1.08 with Webrender - ~30% correctly rendered
> >
> >
> > [snip]
>
> Update:
>
> 89.0.1 76.18 without Webrender
> 89.0.1 4.19 with Webrender
>
> AMD RS780 (DRM 2.40.0 / 3.18.16, LLVM 9.0.1)
> 3.0 Mesa 19.2.8
> mesa/r600
>
Update:
> On Thu, 3 Jun 2021 20:23:19 -0500
> Rod Pemberton <noe...@basdxcqvbe.com> wrote:
>
> >
> > <OT>
> >
> > OMG! Mozilla enabled their broken rendering engine for Firefox 89.0.
> >
> > [snip]
> >
> > E.g., MotionMark1.1 results for my 2009 components computer:
> >
> > 89.0 74+ without Webrender - 100% correctly rendered
> > 89.0 4.33 with Webrender - ~70% correctly rendered
> >
> > 88.0.1 72+ without Webrender - 100% correctly rendered
> > 88.0.1 1.08 with Webrender - ~30% correctly rendered
> >
> >
> > [snip]
>
> Update:
>
> 89.0.1 76.18 without Webrender
> 89.0.1 4.19 with Webrender
>
> AMD RS780 (DRM 2.40.0 / 3.18.16, LLVM 9.0.1)
> 3.0 Mesa 19.2.8
> mesa/r600
>
89.0.2 71.33 without Webrender
89.0.2 4.44 with Webrender
Complete results:
FF Webrender OFF ON
MotionMark 71.33 4.44
Multiply 1.00 pass 1.00 fail
Canvas Arcs 352.20 pass 1.00 fail
Leaves 69.48 pass 84.13 pass
Paths 1156.07 pass 1.00 fail
Canvas Lines 1117.47 pass 1.00 fail
Focus 18.89 pass 19.80 pass
Images 41.47 pass 2.00 pass
Design 36.51 pass 2.74 pass
Suits 52.83 pass 73.29 pass
The higher the score, the more accurate and more complete the rendering
appeared (subjective).
The tests with Webrender on that had higher scores than with Webrender
off actually performed better visually (IMO). Unfortunately, many tests
with Webrender on failed to render properly. A couple of those with
Webrender on that passed weren't rendered quite as well as with
Webrender off. Obviously, those with very high scores with Webrender
off had serious defects or failure with Webrender on.
Now, I'm wondering if they're just optimizing Firefox for a few types
of graphics situations with Webrender, e.g., text and static images,
because for every test with Webrender on which had a high score,
Webrender on scores were slightly better than Webrender off. It's just
that all the other tests either failed or performed poorly with
Webrender on ...
Rod Pemberton
Jul 27, 2021, 12:02:11 AM7/27/21
to
On Mon, 28 Jun 2021 08:12:38 -0500
Firefox 90.0.2 64-bit Linux
(webrender off on)
MotionMark 1.1 70.81 4.93
Multiply 1.00 1.00
Canvas Arcs 396.93 3.00
Leaves 73.96 9.00
Paths 1187.56 2.00
Canvas Lines 996.20 1.00
Focus 23.74 20.95
Images 39.50 2.00
Design 27.28 6.40
Suits 50.41 119.73
Well, they're clearly still having issues with this. Their "Leaves"
test performance completely died, while their "Suits" test performance
improved immensely. They had moderate improvements on other tests.
--
Liberals preach diversity, equity, and inclusion, but engage in
misandry, are racist against whites, promote hatred of conservatives.
(webrender off on)
MotionMark 1.1 70.81 4.93
Multiply 1.00 1.00
Canvas Arcs 396.93 3.00
Leaves 73.96 9.00
Paths 1187.56 2.00
Canvas Lines 996.20 1.00
Focus 23.74 20.95
Images 39.50 2.00
Design 27.28 6.40
Suits 50.41 119.73
Well, they're clearly still having issues with this. Their "Leaves"
test performance completely died, while their "Suits" test performance
improved immensely. They had moderate improvements on other tests.
--
Liberals preach diversity, equity, and inclusion, but engage in
misandry, are racist against whites, promote hatred of conservatives.
Rod Pemberton
Aug 26, 2021, 9:21:41 PM8/26/21
to
On Tue, 27 Jul 2021 00:03:28 -0500
MotionMark 1.1 71.15 6.84
Multiply 1.00 1.00
Canvas Arcs 344.58 1.00
Leaves 73.29 122.67
Paths 1367.65 3.00
Canvas Lines 993.32 1.00
Focus 21.97 24.50
Images 37.61 4.00
Design 30.19 5.56
Suits 54.57 163.36
They restored and improved their "Leaves" performance for webrender.
They improved their "Suits" performance. And, "Focus" improved a bit.
Other tests have slight improvements and slight declines. Now, I'm
wondering if only "Leaves" and "Suits" are what they're focused upon?
Or, if "Focus" is also a base requirement, as "Focus" test mostly
matches hardware performance, but the other two now outperform hardware.
--
Why is it that the most liberal tech companies become the most
authoritarian?
Multiply 1.00 1.00
Canvas Arcs 344.58 1.00
Leaves 73.29 122.67
Paths 1367.65 3.00
Canvas Lines 993.32 1.00
Focus 21.97 24.50
Images 37.61 4.00
Design 30.19 5.56
Suits 54.57 163.36
They restored and improved their "Leaves" performance for webrender.
They improved their "Suits" performance. And, "Focus" improved a bit.
Other tests have slight improvements and slight declines. Now, I'm
wondering if only "Leaves" and "Suits" are what they're focused upon?
Or, if "Focus" is also a base requirement, as "Focus" test mostly
matches hardware performance, but the other two now outperform hardware.
--
Why is it that the most liberal tech companies become the most
authoritarian?
wolfgang kern
Aug 28, 2021, 5:46:37 AM8/28/21
to
On 27.08.2021 04:23, Rod Pemberton wrote:
> On Tue, 27 Jul 2021 00:03:28 -0500
> On Tue, 27 Jul 2021 00:03:28 -0500
>> On Mon, 28 Jun 2021 08:12:38 -0500
>>> On Thu, 17 Jun 2021 12:48:01 -0500
I had to upgrade ThunderBird just now (it wouldn't load w/o).
Every update makes the whole Mozilla story worse in terms of size and
start load time [TB takes up to 30sec to open].
I didn't check on details, so I dunno how to disable or alter any bit.
Where would I have to start ? in this 'not recommended' config ?
__
wolfgang
Kerr-Mudd, John
Aug 28, 2021, 7:15:20 AM8/28/21
to
On Sat, 28 Aug 2021 11:44:34 +0200
wolfgang kern <now...@never.at> wrote:
[]
--
Bah, and indeed Humbug.
wolfgang kern <now...@never.at> wrote:
[]
>
> Thanks for keep track on it, I didn't upgrade to latest FireFox yet, but
> I had to upgrade ThunderBird just now (it wouldn't load w/o).
> Every update makes the whole Mozilla story worse in terms of size and
> start load time [TB takes up to 30sec to open].
>
TB stands for TeraByte.
> Thanks for keep track on it, I didn't upgrade to latest FireFox yet, but
> I had to upgrade ThunderBird just now (it wouldn't load w/o).
> Every update makes the whole Mozilla story worse in terms of size and
> start load time [TB takes up to 30sec to open].
>
--
Bah, and indeed Humbug.
wolfgang kern
Aug 28, 2021, 1:05:46 PM8/28/21
to
On 28.08.2021 12:32, Kerr-Mudd, John wrote:
> TB stands for TeraByte.
yeah, just lazy but maybe understandable within con-text :)
__
wolfgang
> TB stands for TeraByte.
yeah, just lazy but maybe understandable within con-text :)
__
wolfgang
Kerr-Mudd, John
Aug 28, 2021, 4:05:18 PM8/28/21
to
wolfgang kern
Aug 29, 2021, 9:19:17 AM8/29/21
to
On 28.08.2021 22:05, Kerr-Mudd, John wrote:
>>> TB stands for TeraByte.
>> yeah, just lazy but maybe understandable within con-text :)
> It was meant as a joke about TB getting bloated!
TB+FF are "only" about 450 MB but both seem to always check on start for
>>> TB stands for TeraByte.
>> yeah, just lazy but maybe understandable within con-text :)
> It was meant as a joke about TB getting bloated!
everything on my desktop screen, and there's many to see, so it takes a
lot of time.
I gave them a something to read: "Sucks-Ass by Mozilla Shitheads.rtf".
(just a collection of nasty curses in return for annoying behaviour)
__
wolfgang
Rod Pemberton
Sep 7, 2021, 6:38:05 PM9/7/21
to
On Thu, 26 Aug 2021 21:23:00 -0500
Well, it would appear (*) that Firefox 92.0 finally fixed webrender, at
least for Linux!
(I.e., wolfgang, you probably don't need to do anything now that 92.0
has been released. But, you can run MotionMark1.1 to check. Sorry that
I didn't reply sooner, but I've been really busy lately. Ditto to
stalled conversations with James, Paul, and others.)
https://browserbench.org/MotionMark1.1/
The visual quality of the rendering with MotionMark for webrender is now
on par with hardware/opengl. The quality is good enough that it almost
makes me suspect that Firefox might be falling back and using hardware,
somehow, or at least using it.
The old setting to enable opengl has been disabled, should be deleted
from about:config, and a new equivalent setting with a different name
has been added for Firefox 92.0.
The old setting that has been disabled and should be deleted from
about:config is:
gfx.webrender.force.disabled
This new setting that has been added as it's replacement:
gfx.webrender.force-legacy-layers
To test OpenGL you must change
gfx.webrender.force-legacy-layers
from false to true and restart Firefox.
opengl webrender
MotionMark 78.07 62.98
multiply 1.00 1.00
canvas arcs 377.68 181.02
leaves 94.76 133.57
paths 1378.52 648.23
canvas lines 1169.99 567.70
focus 32.70 26.92
images 42.57 40.46
design 23.79 8.93
suits 56.38 180.25
Huge improvements for webrender all around, with "leaves" and "suits"
outperforming hardware/opengl. The "canvas arcs", "paths", and "canvas
lines" have reached 50% of hardware/opengl. That's phenomenal if this
truly is purely software rendering. So, it raises my suspicions as to
hardware fallback, but that wouldn't explain "leaves" and "suits", only
the others.
I also tried toggling
gfx.webrender.reject-software-drivers
from true to false.
opengl webrender
MotionMark 78.10 95.13
multiply 1.00 39.00
canvas arcs 390.34 180.58
leaves 80.76 131.05
paths 1237.11 656.09
canvas lines 1256.80 548.84
focus 29.38 27.33
images 42.57 45.40
design 28.84 8.73
suits 61.19 177.23
I see a huge improvement in "multiply" for webrender, when
gfx.webrender.reject-software-drivers is set to false. I've not seen
any "multiply" result over 1.00 previously. ... Suspicious?
Othwerwise, the numbers are on par with prior tests above.
(*) FYI, I've been updating and/or recompiling numerous graphics
libraries, video drivers, and math related software packages recently,
to obtain better performance on this machine. So, I'm not entirely
100% certain if this fix was exclusively a Firefox fix or something
else, unfortunately.
--
Democrats on abortion: We claim bodily autonomy.
Democrats on Covid: We reject bodily autonomy. Forced vaccinations!
least for Linux!
(I.e., wolfgang, you probably don't need to do anything now that 92.0
has been released. But, you can run MotionMark1.1 to check. Sorry that
I didn't reply sooner, but I've been really busy lately. Ditto to
stalled conversations with James, Paul, and others.)
https://browserbench.org/MotionMark1.1/
The visual quality of the rendering with MotionMark for webrender is now
on par with hardware/opengl. The quality is good enough that it almost
makes me suspect that Firefox might be falling back and using hardware,
somehow, or at least using it.
The old setting to enable opengl has been disabled, should be deleted
from about:config, and a new equivalent setting with a different name
has been added for Firefox 92.0.
The old setting that has been disabled and should be deleted from
about:config is:
gfx.webrender.force.disabled
This new setting that has been added as it's replacement:
gfx.webrender.force-legacy-layers
To test OpenGL you must change
gfx.webrender.force-legacy-layers
from false to true and restart Firefox.
opengl webrender
MotionMark 78.07 62.98
multiply 1.00 1.00
canvas arcs 377.68 181.02
leaves 94.76 133.57
paths 1378.52 648.23
canvas lines 1169.99 567.70
focus 32.70 26.92
images 42.57 40.46
design 23.79 8.93
suits 56.38 180.25
Huge improvements for webrender all around, with "leaves" and "suits"
outperforming hardware/opengl. The "canvas arcs", "paths", and "canvas
lines" have reached 50% of hardware/opengl. That's phenomenal if this
truly is purely software rendering. So, it raises my suspicions as to
hardware fallback, but that wouldn't explain "leaves" and "suits", only
the others.
I also tried toggling
gfx.webrender.reject-software-drivers
from true to false.
opengl webrender
MotionMark 78.10 95.13
multiply 1.00 39.00
canvas arcs 390.34 180.58
leaves 80.76 131.05
paths 1237.11 656.09
canvas lines 1256.80 548.84
focus 29.38 27.33
images 42.57 45.40
design 28.84 8.73
suits 61.19 177.23
I see a huge improvement in "multiply" for webrender, when
gfx.webrender.reject-software-drivers is set to false. I've not seen
any "multiply" result over 1.00 previously. ... Suspicious?
Othwerwise, the numbers are on par with prior tests above.
(*) FYI, I've been updating and/or recompiling numerous graphics
libraries, video drivers, and math related software packages recently,
to obtain better performance on this machine. So, I'm not entirely
100% certain if this fix was exclusively a Firefox fix or something
else, unfortunately.
--
Democrats on abortion: We claim bodily autonomy.
Democrats on Covid: We reject bodily autonomy. Forced vaccinations!
John Smith
Oct 22, 2021, 10:19:41 AM10/22/21
to
> IIRC, most modern browsers have stopped supporting bookmarks, or they
> might be disabled by default. Firefox still supports bookmarks.
>
I don't know about all modern browsers, but Chrome and Firefox, which both cover most likely at least 70% of all users on the planet (Chrome alone covers billions of users) both support bookmarks.
> might be disabled by default. Firefox still supports bookmarks.
>
I have the recent Chrome version and it has bookmarks.
Rod Pemberton
Nov 3, 2021, 5:14:32 AM11/3/21
to
On Tue, 7 Sep 2021 18:39:25 -0500
I was never able to figure out how to enable "opengl" for Firefox 93.
They changed the variables or fully disable it. After testing numerous
settings, some current, some obsolete, I was only able to get webrender
level performance (above), which was half of the opengl setting.
For Firefox 94 on Linux, WebGL detection fails for my system, which
worked correctly in Firefox 93 ... If this happens for you, you should
set webgl.force-enabled to true.
While the standalone version of glxtest correctly identifies my GPU
hardware, Firefox 94's internal glxtest is not detecting my GPU
properly ... I don't currently see any problems because of this.
Apparently, some features, like EGL, are only available on Linux by
default if you have Mesa 21 or later. I managed to recently update
Mesa to 19.2.8 ... and it appears that heavy system upgrades are
required to compile later versions, e.g., gcc, g++, glibc, pthreads,
LLVM, etc. EGL has never been reported as working, so far. If I'm
able to somehow update to Mesa 21, I'll check again.
--
Why does Mr Zuckerberg's metaverse avatar appear as if it's being
strangled by an invisible noose? chest out, shoulders back, ...
They changed the variables or fully disable it. After testing numerous
settings, some current, some obsolete, I was only able to get webrender
level performance (above), which was half of the opengl setting.
For Firefox 94 on Linux, WebGL detection fails for my system, which
worked correctly in Firefox 93 ... If this happens for you, you should
set webgl.force-enabled to true.
While the standalone version of glxtest correctly identifies my GPU
hardware, Firefox 94's internal glxtest is not detecting my GPU
properly ... I don't currently see any problems because of this.
Apparently, some features, like EGL, are only available on Linux by
default if you have Mesa 21 or later. I managed to recently update
Mesa to 19.2.8 ... and it appears that heavy system upgrades are
required to compile later versions, e.g., gcc, g++, glibc, pthreads,
LLVM, etc. EGL has never been reported as working, so far. If I'm
able to somehow update to Mesa 21, I'll check again.
--
Why does Mr Zuckerberg's metaverse avatar appear as if it's being
strangled by an invisible noose? chest out, shoulders back, ...
0 new messages