@Seiya

Sometimes the browser gets stuck while loading a page, and it`s very frustrating because you cannot eve change tabs or continue browsing from another page/tab, while another is still loading/rendering.
OS4 MUI-OWB will benefit from a JIT optimization.

_________________
Ordell Robbie: Is she dead, yes or no?
Louis: Pretty much.

@Cass

Its stuck because on aos4/aros muiowb done with non-threaded curl => blocking of gui when connection stuck. Once i will add threaded curl (fab already give me the sources of it), there will be no such problems. On morphos its of course alreay threaded.

_________________
aos4/aros/mos projects & articles

If you like any amiga related work i do, feel free to donate

Same Athlon II x2 220 at 2.8ghz

Sunspider

Windows XP

Firefox - Total: 328.9ms 1.2%

Debian Squeeze Live from USB

Epiphany - Total: 452.2ms +/- 0.6%
Iceweasel -Total: 1197.0ms +/- 1.3%

Last edited by Spectre660 on 30-Apr-2012 at 01:13 AM.
Last edited by Spectre660 on 30-Apr-2012 at 01:08 AM.

_________________
Sam440ep-Flex 800Mhz ,1GB.Radeon HD6670 lowprofile via PCI to PCIe Adaper.
Sam460ex 1.15Ghz, 2GB.Radeon HD6670 . Apex TM-302-3 Case

@Tuxedo

Quote:

A code compiled for a different cpu with different hw, different opts and different code core not have any sense imho... We can se where it works faster ok, but the becnhmark nature wasnt that...benchmarks was made to lok where the same works better so giving us wich hw was better in same conditions, testing different things wasnt a benchmark..only that.

Sorry, can't agree here. The goal of any benchmark is finding the better solutions among the available ones. For instance, I'd run a similar benchmark to understand what's the best browser on different hardware platforms, and choose that one for my personal browsing.

Having the same unoptimized routines running on PPC, ARM and X86 processors, even at the same clock speed, just to see where they better perform, is absolutely pointless if in the real world I won't ever use them, but I will choose the optimized ones instead. I have jit-enabled OWB now, running on my amiga-like OS of choise, why the hell should I use the jit-disabled one? The former is faster!

MediaCom SmpartPad 711i 4 GB
Cortex A8 1 Ghz, 1 GB DDR3, Mali 400P

SunSpider:

Opera Mobile 12
4163,0 ms +/- 1,0%

Dolphin HD
3767.3ms +/- 2.8%

V8 Benchmrk Suite:

Opera Mobile 12
Score = 256

Richards: 456
Deltablue: 145
Crypto: 488
Raytrace: 246
EarlyBoyer: 315
RegExp: 131
Splay: 160
NavierStokes: 357

Dolphin HD
score: 626

Richards: 962
Deltablue: 812
Crypto: 863
Raytrace: 661
EarlyBoyer: 1424
RegExp: 234
Splay: 429
NavierStokes: 370


PeacePeeker:

Opera Mobile 12
114 - html5 capabilities 3/7

Dolphin HD
PeacePeaker
196 - html5 capabilities 1/7

Last edited by Seiya on 22-May-2012 at 08:59 PM.
Last edited by Seiya on 22-May-2012 at 08:56 PM.
Last edited by Seiya on 22-May-2012 at 08:30 PM.

_________________

@kas1e

Any news when this non-threaded curl problem will be fixed as its the only thing holding it back from being used in real world?

Got my AmigaOne 500 in the office for another member of staff to work on - and have had to fallback to OWB as its much more reliable - despite its other shortcomings the older OWB is much more stable at daily browsing. Timberwolf is simply out of the question atm as its just so slow in its present form - I can hear you saying 'I told you so' from here - I just didn't expect it to take them so long!

I really do like MUI-OWB, but its only useful atm on Amiga forum sites - anywhere else on the internet and its 50/50 as to whether this threading problem will surface its ugly head. For example, if I go to morphos-files.net (irony of it all) in MUI-OWB the browser will sit there for 1 minute, CPU/HD access maxing out, mouse becoming inoperative, DSI appearing, rinse and repeat 3-5 times... so finally after 5 minutes the page has loaded.

EDIT: Fixed typos.

Last edited by djrikki on 22-May-2012 at 09:19 PM.
Last edited by djrikki on 22-May-2012 at 09:19 PM.
Last edited by djrikki on 22-May-2012 at 09:18 PM.

_________________
Official AmigaOS Website
AmigaOS on Facebook
Lake Marketing and Events

@djrikki

If there's CPU load and that HD is accessed, then it's clearly not about monothreaded DNS issue, so it won't help when it's there. Not to mention your crashes...

Your issue is entirely different, and i have no idea what it could be, really. It certainly didn't happen on MorphOS, even when curl wasn't multithreaded yet.

Last edited by Fab on 22-May-2012 at 09:40 PM.

@djrikki

No such problems here .
MUI-OWB is very stable for me.

_________________
Sam440ep-Flex 800Mhz ,1GB.Radeon HD6670 lowprofile via PCI to PCIe Adaper.
Sam460ex 1.15Ghz, 2GB.Radeon HD6670 . Apex TM-302-3 Case

@fab

I see, well I'll have to post a crash log, if I go to sourceforge, stackoverflow.com, google calendar (if being presented with the log-in screen), morphos-files, twitter and numerous other places I see those symptoms.

_________________
Official AmigaOS Website
AmigaOS on Facebook
Lake Marketing and Events