PDA

View Full Version : why is ultrafxp so much faster??


x-dream
09-01-2004, 01:36 PM
UltraFXP v1.07 free edition:
[i] cs-shell.r07 47,7 Mbytes/18,73(s)/2_659,39Kbps

FlashFXP v3.0.1015:
Transferred: cs-shell.r09 47,68 MB in 42,31 seconds (1_154,0 KB/s)

Why isit so much faster is there anyway of increasing the download buffer in flashfxp? (i know several ppl that find ultrafxp faster)

MxxCon
09-01-2004, 01:53 PM
don't look at transfer speed results from 2 different clients since they measure speed slightly differently.
rather you should look at what your SERVER is reporting, since both times it'll be counting exactly the same way.

was this transfer done over the internet or local network?
did you eliminate possible traffic congestions during either of the tests?

bigstar
09-01-2004, 02:07 PM
There is no obvious answer to your question.

I don't think anyone can say for certain why product X is faster than product Y.

Everyone has different hardware and software setups and often you will find that some software performs better on certain setups.

That being said I have no problem maxing out my 768kbit DSL with FlashFXP, my 100mbit LAN, my 400mbit firewire pc to pc.

As far as increasing the download buffer, The buffer is 32KB

x-dream
09-01-2004, 02:14 PM
Originally posted by MxxCon
don't look at transfer speed results from 2 different clients since they measure speed slightly differently.
rather you should look at what your SERVER is reporting, since both times it'll be counting exactly the same way.

was this transfer done over the internet or local network?
did you eliminate possible traffic congestions during either of the tests?

This was done at the same time (also tested on my windows 98 computer same result) from the same server(internet) and was run several times. Well you can only count seconds in one way and 18 sec is far better than 42 sec

As far as increasing the download buffer, The buffer is 32KB

Maybe that is the problem or is that the best buffersize?

MxxCon
09-01-2004, 04:09 PM
Originally posted by x-dream
This was done at the same timenot an accurate way of measuring, becuase ufxp might be more 'network aggressive' so to say...
Well you can only count seconds in one way and 18 sec is far better than 42 secufxp and ffxp can start counting seconds from different points..one can start counting from the moment you pressed 'GO' button or start counting from the moment PORT command is issued or from the moment SORT command is issued or from the moment 1st byte is received. on a very slow server those differences can add up. it doesn't look like it's the case in your example but still something to keep in mind...
i'd say that ufxp is just more aggressive when it's fighting for it's bandwidth. try doing those tests one at a time and/or monitor connection speed with external application such as RAS Graph & Stats (http://forum.flashfxp.com/showthread.php?s=&threadid=2400) or any other similar application.

Mouton
09-01-2004, 05:30 PM
FFXP: [Speed: 11044.3kb/s]
UFXP: [Speed: 10498.6kb/s]
ftp.exe: [Speed: 13904.4kb/s]
Same file, from same source, to same dir, after a couple of test transfers to cache the file as much as possible on the server.

So like bigstar says, results vary.

x-dream
09-02-2004, 12:15 PM
Originally posted by Mouton
FFXP: [Speed: 11044.3kb/s]
UFXP: [Speed: 10498.6kb/s]
ftp.exe: [Speed: 13904.4kb/s]
Same file, from same source, to same dir, after a couple of test transfers to cache the file as much as possible on the server.

So like bigstar says, results vary.

On the lan i get 11mb/s on both also