What`s On My PII and How fast.

realgeorge · Post by **realgeorge** » March 28th, 2004, 1:02 pm

I have 4 builds installed on my PII: The JTw 3-27-04, MozJF 3-26-04, Djeter 3-25-04, and Nova 2-16-04. I just ran some practical speed tests using http://www.numion.com/Stopwatch/index.html? to check the page loading of my four most used sites. The results shown are the average of the four sites in order of fastest first: MozJF 4.360 sec, Nova 5.012 sec , Djeter 5.056 sec, and JTw 5.184 sec. All have the same Pipelining settings recommended by Laszlo`s post at http://snipurl.com/5d2p, and used the same dialup connection @ 44 Kbps measured at http://reviews.cnet.com/7004-7254_7-0.html. Comments?

frease · Post by **frease** » March 28th, 2004, 7:22 pm

Yes, when I was still using my P2 machine (before the motherboard blew a month ago

), I find that MozJF's builds are the fastest and most stable. That's why I highly recommend using his builds on a P2.

realgeorge · Post by **realgeorge** » March 28th, 2004, 8:31 pm

I certainly agree with that frease. It`s hard to beat MozJF`s builds.

alex-the-cat · Post by **alex-the-cat** » March 28th, 2004, 9:02 pm

how about the standard build?

frease · Post by **frease** » March 28th, 2004, 10:33 pm

The standard build is had no special optimisations but MozJF's has (same goes for other processor-specified builds).

Lohvarn · Post by **Lohvarn** » March 29th, 2004, 1:01 pm

Honestly the thing that would cause the greatest variations here is your Internet connection, not the optimization flags...

MozJF · Post by **MozJF** » March 29th, 2004, 1:43 pm

I completely agree, but having a lightly speeder build could help sometimes ;o)

Lohvarn · Post by **Lohvarn** » March 29th, 2004, 1:48 pm

Well of course it was not to say that MozJF's builds are bad, just that testing over an Internet connection (especially modem) will normally yield unreliable results

realgeorge · Post by **realgeorge** » March 29th, 2004, 2:18 pm

Lohvarn wrote:Honestly the thing that would cause the greatest variations here is your Internet connection, not the optimization flags...

I do and did all my testing using the same dialup connection which does not have the wider fluctuations as having a new connection for each test. It`s about as good as I can do on dialup.

frease · Post by **frease** » March 29th, 2004, 4:20 pm

I have broadband 512kps and I find that having optimised builds do help, even though I do not have the statistics with me.

Lohvarn · Post by **Lohvarn** » March 29th, 2004, 6:29 pm

realgeorge wrote:
Lohvarn wrote:Honestly the thing that would cause the greatest variations here is your Internet connection, not the optimization flags...
I do and did all my testing using the same dialup connection which does not have the wider fluctuations as having a new connection for each test. It`s about as good as I can do on dialup.

Yes but what about the server? How about the routers on the way? How about your ISP?

realgeorge · Post by **realgeorge** » March 29th, 2004, 10:34 pm

Thanks for your comments. Any suggestions on how one could run reliable speed comparison tests using dialup.

mtigas · Post by **mtigas** » March 29th, 2004, 11:16 pm

Download stuff to your hard drive and then just view it from your hard drive in each browser build. That'd elminate the randomosity of using an intenret connection.

Only do it if your computer isn't so fast that you can't time it.

Hm, but doing the download thing, cache issues arise. ;p Bah, it's good enough. Better than a network-based test anywhoo.

sasquatch · Post by **sasquatch** » March 30th, 2004, 7:29 am

randomosity?

;-)

sasquatch · Post by **sasquatch** » March 30th, 2004, 9:02 am

currently using 20040325 Firefox/0.8.0+ (djeter)

This one actually makes things useable on the ol' PII PC's running WinNT 4.0 with 128mb RAM.
Bring on the PII builds!

What`s On My PII and How fast.

What`s On My PII and How fast.

Line Speed Variations.

Re: Line Speed Variations.

Fast, Faster & Fastest