Active Topics
-
The N900 is now 15 years old! (7)
to Nokia N900 by Danct12 - 1 day ago -
Sailfish OS for the Motorola Moto G7 Power (XT1955-5) - (ocean) (4)
to SailfishOS by Maemish - 1 day, 20 hrs ago -
F(x)tec Pro1x (QWERTY) for sale (0)
to Buy & Sell by aln00000 - 4 days, 5 hrs ago - more...
coffeedrinker |
2008-01-13
, 22:48
|
|
Posts: 168 |
Thanked: 51 times |
Joined on Jun 2007
|
#11
|
Actually General, I was referring the work that Mozilla was doing trying to get a mobile browser going. We'll definitely gain from that, too. I am not discounting the work of the microb team.
| The Following User Says Thank You to coffeedrinker For This Useful Post: | ||
 |
|
2008-01-14
, 00:07
|
|
Posts: 693 |
Thanked: 502 times |
Joined on Jul 2007
|
#12
|
Although i like current browser much better and have no stability issues, I did look into ways to increase performance.
Back in the opera days someone showed how you could increase simultaneous connections to server to increase performance.
Mozilla seems to have similar values so i am using them now and its hard to tell how much of a difference it makes but perhaps other itt users can post their results.
Mainly for mozilla the values seem to be kept in :
/usr/lib/microb-engine/greprefs/all.js
Its a large file but skip down near the middle (40%) of the file and find settings for :
pref("network.http.max-connections", 24) Increase this to maybe 32
pref("network.http.max-connections-per-server", 8) Increase this to maybe 16
pref("network.http.max-persistent-connections-per-server", 4) Scary maybe 6?
and further down
pref("network.http.pipelining.maxrequests", 8) Increase this to maybe 16
Although setting these values too high could saturate the server or your connection, i think the servers will limit you if you exceed their limits.
I would guess this would help me more when i'm on higher latency cell phone, where i request only 8 web page images/icons at a time at probably 250ms latency per batch... i'd rather go ahead and request a bunch more and pay for latency in larger batches. Even if you are on a wifi cable/dsl, your connection to server might be higher latency so this would also help.
Feel free to reply if this makes a noticable difference to you, i'm still trying out.
Oh and whatever you do, don't use maemo.org as a benchmark for browser performance, you'd think they were running that webserver out of someones garage
Last edited by pipeline; 2008-01-14 at 00:11.
Back in the opera days someone showed how you could increase simultaneous connections to server to increase performance.
Mozilla seems to have similar values so i am using them now and its hard to tell how much of a difference it makes but perhaps other itt users can post their results.
Mainly for mozilla the values seem to be kept in :
/usr/lib/microb-engine/greprefs/all.js
Its a large file but skip down near the middle (40%) of the file and find settings for :
pref("network.http.max-connections", 24) Increase this to maybe 32
pref("network.http.max-connections-per-server", 8) Increase this to maybe 16
pref("network.http.max-persistent-connections-per-server", 4) Scary maybe 6?
and further down
pref("network.http.pipelining.maxrequests", 8) Increase this to maybe 16
Although setting these values too high could saturate the server or your connection, i think the servers will limit you if you exceed their limits.
I would guess this would help me more when i'm on higher latency cell phone, where i request only 8 web page images/icons at a time at probably 250ms latency per batch... i'd rather go ahead and request a bunch more and pay for latency in larger batches. Even if you are on a wifi cable/dsl, your connection to server might be higher latency so this would also help.
Feel free to reply if this makes a noticable difference to you, i'm still trying out.
Oh and whatever you do, don't use maemo.org as a benchmark for browser performance, you'd think they were running that webserver out of someones garage
Last edited by pipeline; 2008-01-14 at 00:11.
| The Following 2 Users Say Thank You to pipeline For This Useful Post: | ||
|
|
2008-01-14
, 00:34
|
|
Posts: 566 |
Thanked: 150 times |
Joined on Dec 2007
|
#13
|
Thanks. here are some more tips:
http://www.tweakfactor.com/articles/...oxtweak/4.html
I use the 'general' tweaks and the ones for 'Slow Computer, Fast Connection' mentioned in the article and it speeds up things quite a bit, or gives a faster 'feel' as with the glayout.initialpaint.delay setting. I assume these settings do have some tradeoff/downsides or else they would be default.
Last edited by iamthewalrus; 2008-01-14 at 01:08.
http://www.tweakfactor.com/articles/...oxtweak/4.html
I use the 'general' tweaks and the ones for 'Slow Computer, Fast Connection' mentioned in the article and it speeds up things quite a bit, or gives a faster 'feel' as with the glayout.initialpaint.delay setting. I assume these settings do have some tradeoff/downsides or else they would be default.
Last edited by iamthewalrus; 2008-01-14 at 01:08.
 |
|
2008-01-14
, 02:18
|
|
Posts: 861 |
Thanked: 734 times |
Joined on Jan 2008
@ Nomadic
|
#14
|
I thought that I'd chime in here. One of the things that I liked best about the IT when I reviewed it this time last year was the Opera browser. It was not just fast, but rendering of sites was very well done (and led to a good deal of bug reports to developers to fix their sites with correct code).
That being said, I am liking MicroB just a touch more because of the expanded features support. I do wish that it were more like the FF3 beta and load HTML lighting fast, but it does ok in that dept.
One thing that I have not see much a definitve answer on has been the use of browser extensions. It would seem to me that on such a device that the use of extensions and widgets would be really good. But other than a few GreaseMonkey scripts, I do not see any. If that area of things could pic up (along with the rendering performance), I'd be happier with it.
That being said, I am liking MicroB just a touch more because of the expanded features support. I do wish that it were more like the FF3 beta and load HTML lighting fast, but it does ok in that dept.
One thing that I have not see much a definitve answer on has been the use of browser extensions. It would seem to me that on such a device that the use of extensions and widgets would be really good. But other than a few GreaseMonkey scripts, I do not see any. If that area of things could pic up (along with the rendering performance), I'd be happier with it.
 |
|
2008-01-14
, 02:30
|
|
Posts: 469 |
Thanked: 88 times |
Joined on Sep 2007
@ Montana
|
#15
|
Originally Posted by ARJWright
Well there is also the AdBlock extension. It does slow down the browser when it initially launches (a few seconds?), and takes some memory, but it can really speed up some ad heavy sites. Especially if they are riddled with flash ads.
.....It would seem to me that on such a device that the use of extensions and widgets would be really good. But other than a few GreaseMonkey scripts, I do not see any. If that area of things could pic up (along with the rendering performance), I'd be happier with it.
 |
|
2008-01-14
, 02:50
|
|
Posts: 5,478 |
Thanked: 5,222 times |
Joined on Jan 2006
@ St. Petersburg, FL
|
#16
|
Originally Posted by ARJWright
Lies! Opera rendered sites very poorly. It was an old and buggy rendering engine that was very much non-standards–compliant.
I thought that I'd chime in here. One of the things that I liked best about the IT when I reviewed it this time last year was the Opera browser. It was not just fast, but rendering of sites was very well done (and led to a good deal of bug reports to developers to fix their sites with correct code).
Originally Posted by ARJWright
MicroB would be lightning-fast, too, if you ran it on a 2.0GHz Core 2 Duo.
That being said, I am liking MicroB just a touch more because of the expanded features support. I do wish that it were more like the FF3 beta and load HTML lighting fast, but it does ok in that dept.
Originally Posted by ARJWright
MicroB does not use the same interface elements and scripting as Firefox. Because of this, GUI stuff has to be specifically ported. More plugins will come with time.
One thing that I have not see much a definitve answer on has been the use of browser extensions. It would seem to me that on such a device that the use of extensions and widgets would be really good. But other than a few GreaseMonkey scripts, I do not see any. If that area of things could pic up (along with the rendering performance), I'd be happier with it.
|
|
2008-01-14
, 03:30
|
|
Posts: 85 |
Thanked: 10 times |
Joined on Dec 2007
|
#17
|
i have had good luck with the os2008 browser so far...
except some streaming content fails to play,
and java apps won't load from a web page (from my broker).
it is a better browser overall than on my palm or cell phone...
as soon as someone figures out how to run java apps from
a web page, let me know!
except some streaming content fails to play,
and java apps won't load from a web page (from my broker).
it is a better browser overall than on my palm or cell phone...
as soon as someone figures out how to run java apps from
a web page, let me know!
 |
|
2008-01-14
, 05:45
|
|
Posts: 107 |
Thanked: 14 times |
Joined on Dec 2007
|
#18
|
Originally Posted by pipeline
May be some body can do a Faster Fox like add-on for MicroB???
Although i like current browser much better and have no stability issues, I did look into ways to increase performance.
Back in the opera days someone showed how you could increase simultaneous connections to server to increase performance.
Mozilla seems to have similar values so i am using them now and its hard to tell how much of a difference it makes but perhaps other itt users can post their results.
Mainly for mozilla the values seem to be kept in :
/usr/lib/microb-engine/greprefs/all.js
Its a large file but skip down near the middle (40%) of the file and find settings for :
pref("network.http.max-connections", 24) Increase this to maybe 32
pref("network.http.max-connections-per-server", 8) Increase this to maybe 16
pref("network.http.max-persistent-connections-per-server", 4) Scary maybe 6?
and further down
pref("network.http.pipelining.maxrequests", 8) Increase this to maybe 16
Although setting these values too high could saturate the server or your connection, i think the servers will limit you if you exceed their limits.
I would guess this would help me more when i'm on higher latency cell phone, where i request only 8 web page images/icons at a time at probably 250ms latency per batch... i'd rather go ahead and request a bunch more and pay for latency in larger batches. Even if you are on a wifi cable/dsl, your connection to server might be higher latency so this would also help.
Feel free to reply if this makes a noticable difference to you, i'm still trying out.
Oh and whatever you do, don't use maemo.org as a benchmark for browser performance, you'd think they were running that webserver out of someones garage
Regards,
Taguapire.
 |
|
2008-01-14
, 05:54
|
|
Posts: 11,700 |
Thanked: 10,045 times |
Joined on Jun 2006
@ North Texas, USA
|
#19
|
IE7 crashes much, much more often for me than microB. It's not even close.
__________________
Nokia Developer Champion
Different <> Wrong | Listen - Judgment = Progress | People + Trust = Success
My personal site: http://texrat.net
Nokia Developer Champion
Different <> Wrong | Listen - Judgment = Progress | People + Trust = Success
My personal site: http://texrat.net
| The Following User Says Thank You to Texrat For This Useful Post: | ||
|
|
2008-01-14
, 06:32
|
|
Posts: 21 |
Thanked: 3 times |
Joined on Mar 2007
|
#20
|
Originally Posted by Texrat
My IE7 crashed just after I read this... must have been out of spite!!
IE7 crashes much, much more often for me than microB. It's not even close.