On Wed, 02 Aug 2006 14:46:27 +0200, Lawman <law...@invalid.com> wrote: > A test of alternative browsers showed that Opera 9 was > over 3.5 times faster running a Javascript test suite...
This I found in my own web applications as well. Some pages require a build-up of javascripted data, and we can see that Opera9 loads them in about 3 seconds while MSIE7b3 took 22 seconds... for the same functionality.
I also recognised that MSIE is quite sensitive to script optimisation. I managed to rewrite the code I used and bring the loading time down to 7 seconds, 3 times faster. Opera9 would still load it in 3 seconds, more than twice as fast.
One of the optimisations: MSIE does not like dom traversal, so use cached objects wherever suitable.
-- Yours, ΩJr
Using Opera 9.01 build 8543 on Windows 2003 Server
On Wed, 02 Aug 2006 13:46:27 +0100, Lawman <law...@invalid.com> wrote: > A test of alternative browsers showed that Opera 9 was > over 3.5 times faster running a Javascript test suite...
On my system, Firefox 1.5.0.6 does finish in the 2000-2100 range (not too far off from the figures on the site), but Opera 9.01 side-by-side is nowhere near 442. Instead, it ranges between 1700 and 2000. If our Firefoxes finish comparably, why don't our Operas?
Anyone else not seeing a dramatic difference on their system?
Firefox gave 1913-2123. Not only is it slower (never mind benchmarks, the wait in FF felt like a CGI form submission whereas Opera felt responsive like an application) but there seems to be much more fluctuation in FF. It varies by over 2 seconds but there's nothing much going on in the background on this machine.
IE6 gave 1582-1612.
IE5.0 gave 841-861 (and pretty consistently gave 841, with only a couple of 861s and only one 841 inbetween) but it appears it didn't run the array, string and AJAX functions.
Milhouse Van Houten wrote: > "Lawman" <law...@invalid.com> wrote in message > news:Xns98134F14725C9lawmannumogag@140.99.99.130... >> A test of alternative browsers showed that Opera 9 was >> over 3.5 times faster running a Javascript test suite...
> On my system, Firefox 1.5.0.6 does finish in the 2000-2100 range (not too > far off from the figures on the site), but Opera 9.01 side-by-side is > nowhere near 442. Instead, it ranges between 1700 and 2000. If our > Firefoxes > finish comparably, why don't our Operas?
> Anyone else not seeing a dramatic difference on their system?
I'm consistently getting 406ms w/ 9.01 on my Celeron 1.8GHz system w/ 1GB RAM and WinXP SP2. Whereas, FF consistently gives me between 1475ms and 1582ms on the same system.
Below are my results for each particular test in Opera 9.01:
Try/Catch with errors 16 Layer movement 47 Random number engine 46 Math engine 79 DOM speed 46 Array functions 63 String functions 78 Ajax declaration 31 ----------------------- Total Duration 406 ms
So, which of these tests are skewed for your Opera results? Assuming one or more results are significantly larger than the others.
>> On my system, Firefox 1.5.0.6 does finish in the 2000-2100 range (not too >> far off from the figures on the site), but Opera 9.01 side-by-side is >> nowhere near 442. Instead, it ranges between 1700 and 2000. If our >> Firefoxes finish comparably, why don't our Operas?
>> Anyone else not seeing a dramatic difference on their system?
> I'm consistently getting 406ms w/ 9.01 on my Celeron 1.8GHz system w/ 1GB > RAM and WinXP SP2. Whereas, FF consistently gives me between 1475ms and > 1582ms on the same system.
> So, which of these tests are skewed for your Opera results? Assuming one > or more results are significantly larger than the others.
String functions kills it for me every time. Perhaps it doesn't like to be run with a lot of tabs open, which I have at the moment? If so, I don't get why.
Try/Catch with errors 78 Layer movement 63 Random number engine 62 Math engine 110 DOM speed 47 Array functions 62 String functions 1484 Ajax declaration 32 Total Duration 1938 ms
On Thu, 3 Aug 2006 10:16:30 -0700, Milhouse Van Houten <b...@myrealbox.com> wrote:
> String functions kills it for me every time. Perhaps it doesn't like to be > run with a lot of tabs open, which I have at the moment? If so, I don't get > why.
> Try/Catch with errors 78 > Layer movement 63 > Random number engine 62 > Math engine 110 > DOM speed 47 > Array functions 62 > String functions 1484 > Ajax declaration 32 > Total Duration 1938 ms
I opened 28 empty tabs and found that the time for string functions roughly doubled, so you may be on to something. Perhaps it's a memory allocation issue: string operations are the sort of operations that are likely to be allocating and freeing small blocks of memory, and I suspect that the more blocks of memory Opera is dealing with the slower such operations become.
It might be worth comparing the speed of an Opera that has been running for several hours of active use with one that has just been started.
-- Matthew Winn [If replying by email remove the "r" from "urk"]
