Actually that's about 6.4 times less than the 3.6 because you've mixed up the units. 3g is far from a WLAN+DSL/cable. If you want performance from your internets, you shouldn't be depending on 3g. 100 kB/s downstream was a reasonable figure for me when I used 3g

did i mix my units up,

I read it as, 3.6mb to start with, speedtest said 0.6mb (600kb). which is 6 times worse, then the actual speed of a download is 70kb, which is nearly ten times slower than that, which is how i got my 51 times worse.

But, bits and bytes, f--k knows lol!

I know that 3g deffo isnt the way for good internet but even 8 times slower than advertised when in a 3g area (which i am) is still pretty shocking i thought, maybe i expect too much.

I am looking at getting a proper connection, but with BT wanting to charge me a one off connection charge of nearly 200 quid, and no other ISPs willing to give me an internet package without an existing line i'm pretty much screwed!

3.6 Mbps => 3.6 / 8 = 0.45 MB/s (450 kB/s)
450 kB/s / 70 kB/s = 6.42

Err, i see now i think, you need to divide the Mbps value by 8 to give you the MB/s value, i thought that Mbps and MB/s were two different ways of writing the same thing, oops! :hide:

That's a mistake many people do, so you should not feel ashamed

Mbps is "megabits per second" and can also be written as Mbit/s
MB/s is "megabytes per second"
kB/s is "kilobytes per second"

One byte is 8 bits, so divide Mbps by 8 to get MB/s and vice versa. 0.45 MB/s looks clumsy, so you can write it as 450 kB/s (MB/s * 1000).

Usually you'll find that the below is used, but people tend to slack and use the wrong things here and there. Like a small 'm' (milli) instead of 'M' (mega) and confuse bits and Bytes.
b = bit
B = byte
k = thousand
M = million

I use WiFi for my computer since the router is in my roommates room and I race almost exclusively in FOX servers and leagues. If you want I can provide replays of me racing (and very very close racing, if I were jumping about to everyone else they surely would have crashed many times over) to prove that (not all) wireless is the horrible spawn of satan that it was proclaimed to be (albeit 3 years ago).

If anyone expects me to ask every race "am I lagging" then they are out of their fool mind. If airio (the true spawn of satan) tells me I'm lagging I check then to see if other people see me lagging or if it was just one spike. I've had people tell me I'm lagging before and guess what, most of the time it was because I forgot to turn off utorrent or my roommate was downloading something.

I don't think the problem still exists with the new WiFi hardware, but with the old one I did always use the cable

Quote from danthebangerboy :

I know its not exactly the same, as it is 3g internet, but it is wireless so kind of related to this...

What really pisses me off is the fact that all the documentation that came with my huawei e156g dongle says that it can run at 3.6mbps on the fastest (3g) network, which quite frankly total bollocks to be fair, as i can always pick up a 3g signal, usually between 2 and 5 bars, with 5 being full signal, and i have never had the upload speed go any faster than about 600kbps and download of about 5kbps regardless of weather i have full signal or only 2 bars.

So why is the so called available speed allowed to be, well, seemingly made up by the ISP, as i know of a few other people on other forums who have the same dongle as me and the best speed any of them have seen via a speed test is about 790kbps and 6kbps respectively, and he lives in central london on the top floor of a tower block with flat out full 3g signal all the time, which is still nowhere even remotely close to the 3.6meg quoted speed.

Using MobilePartner on my friends e160G (which is very very similar to the e156g) I get 3.6 MBPS in Needham Market no less.

Yet to get a decent speed on O2 3G using my ZTE 627 in East Anglia - at home I get dialup which is ok as a backup stable connection but even in Ipswich etc I still ONLY get dialup

At work as we sell mobile broadband I have run coverage checkers for the big 3 networks for most places in Suffolk - I think i've only ever had one T-Mobile come back as having any coverage even though my phone on T-Mobile gets coverage in tons of places.

Saying that, however, I was on holiday in Aldeburgh recently and took both my phone (rooted and tethered) and my O2 SIM expecting to get internet on at least one, and not only was their no 3G signal, Orange, T-Mobile, O2 and sometimes Vodafone all had pretty much no coverage at all.

Wireless ROCKS. My phone picks up emails faster than 3G even when it's in my pocket or on my desk, I can turn my laptop on anywhere in the house and surf the web - it's so simple. But Once you need high bandwidth (ethernet = 12 mb/s, WiFi within 3 feet of router = 2 mb/s) or low latency / packet loss (i.e. gaming) all sorts of issues crop up.

I still think WiFi is the answer for laptop users, though.

Sure the hardware has a lot to say, but the problem is the wireless itself, due to the fact it's air. Sure, it can be very good, and superb for laptops around in the house when you canot be bottered to plug it with a cable just to check your netbank / forum or newspaper page. But it will allways, no matter what, give more latency/lag and similar things compared to cable.

So, as people have pointed out before, wireless is good enough for most things, and very userfriendly. But for games please try, as much as possible to run a wired connection.

And, even if you does not see the lag, the others will

Found some test reports ( All in Chinese, so won't link here. ) of the Huawei E156G. Some say they can enjoy full speed, but some others suffer from 30k.

Where you pick up the signal makes much difference.


Off topic: Why only 6Hz refresh rate ? Maybe it's not too bad for internet, but in LAN 20Hz+ is not difficult at all.
Under 6Hz you have to wait at least 167ms to get the next package, and under 4Hz it's 250ms. A 50ms ping is not big compared to this.

@ Dan - the 3G dongles go for nothing on the 'bay btw. I got my ZTE 627 for £5 from 3dongle4free.co.uk (they now charge £10), you can get the e220 or e160g for a similar amount from eBay.

Have you tried using Mobile Partner? 3 take their original software and rape it ruining both the signal and speed, in the same way networks **** up their phones OS's. If you download Mobile Partner for the right device (you can grab them from the DC Unlocker site) it should work if you put the right APN settings in.

You can get a similar bit of kit for ZTE devices - but I never got it working and am stuck using a modified version of Telstra's connection manager.

Quote from The Very End :

Sure the hardware has a lot to say, but the problem is the wireless itself, due to the fact it's air. Sure, it can be very good, and superb for laptops around in the house when you canot be bottered to plug it with a cable just to check your netbank / forum or newspaper page. But it will allways, no matter what, give more latency/lag and similar things compared to cable.

So, as people have pointed out before, wireless is good enough for most things, and very userfriendly. But for games please try, as much as possible to run a wired connection.

And, even if you does not see the lag, the others will

we are racing a lot with friends so doing some regular tests won't be a real problem, will check it, because it made me curious

I've switched from cable to WiFi after changing my laptop for a new one and none of my friends have complained about the lag since

but the old one was simply impossible to use, that's a fact

Be sure to enable the latency thing inside the game, I don't remember the command for that, something about ctrl+shift+8 if I guess right, canot check since I'm not at a computer with LFS installed.

Do various test, and make sure your friends can see your car at all times, see if there are drop-spots, times where it does lag every now and then.

And one more thing, the test should be done online with random people around the world, if all your friends are living very nearby the difference probally is slim, but then again then we're moving over to the distance between players and shouldn not be rellevant in this case.

Just try a lot, and make sure everyone has the latency thingy running

Also be sure that your computer runs LFS without no problems at all. If your computer has a hard time running LFS, overheats or similar, it will aswell lag and cause other problems.

Quote from danthebangerboy :

This is what i mean about mine, three (3) and huawei both say that this thing should run at 3.6mb per second, but according to speedtest i can only get a d/l speed of 0.61...



And then, if i actually try and download something the speed is nearly ten times worse than what speedtest says my upload speed is...



So, all the way from a 'possible speed' of 3.6meg, and it actually works at about 70k, 51 times slower FFS!!!

Barely better than dial-up, actually, scratch that, it's worse, dial-up would be more stable!

How can they get away with quoting these big numbers that are nothing but lies.

You're wrong. Your "3.6meg" that you say, is probably 3.6megabits/second, which works out to be about 450 kiloBYTES per second. Sure you're not upto your full connection, but you've been mislead by megabits vs megabytes.

Also, it's not faster than dialup as a 56k modem produces transfer rates of 7 kiloBYTES per second, as the "56k" part is 56 kilobits.

EDIT: I must not have refreshed recently, seeing when I first started this post, all the other ones weren't here...

im on wifi and i think im fine

Quote from The Very End :

Be sure to enable the latency thing inside the game, I don't remember the command for that, something about ctrl+shift+8 if I guess right, canot check since I'm not at a computer with LFS installed.

Its Shift+F8

Network debug mode

Very interesting thread this one, I have never liked or trusted wireless as so many things can interfere with the signal.

SD.

Quote from SparkyDave :

Its Shift+F8

Network debug mode

Very interesting thread this one, I have never liked or trusted wireless as so many things can interfere with the signal.

SD.

Thanks for the info

One thing that is the main problem, as you say, is the interference. Lets say your living in a wall-to-wall appartment. If your neightbour buys a baby caller (the one that let you hear what happens on the bedroom while your sitting watching TV in the living room and just don't care) , your basically ****ed. In my work, I have had so many cases where things like these nasty shitcallers have ruined the network for the ones that own it, aswell as their neightbour. The main reason to this is that most wireless things are running on the very same frequence noways, that means that sooner or later something will interfer with eachothers and all hell is lose. You can of course try to change the channel on the router, but at most times it will help little or nothing, tho little can be a difference it does not help if you get kicked out of the net each time the baby caller is turned on.
You can try buying these fancy new routers, they runs on a little different frequence, but that does not mean your safe, and it costs LOADS of money, and if you got a lot of interfering stuff around it still will have problems.

But will be interesting to see if there are any noticable differences when we're taking the tests with wired connection and wireless

I did a quick test to see how much of a point you anti-WiFi guys have. The IP is www.lfs.net's. I did a ping test from the router then did the same ping test from my computer, which is two rooms away from the router. As you can see the minimum, average and maximum are all pretty much identical. at the router and my computer. My WiFi isn't adding any extra latency.

If anyone isn't happy and would rather me do more than 5 rounds let me know. Keep it reasonable though, I'm not taking all day to prove a point.

I would also be interested to see the same sort of test from other WiFi users. Be sure to have the same number of rounds in each test (for statistical accuracy) and the same number of bytes. You can see in my screen shot how to control these.

I think that we'll need a longer test to simulate a propper situation, let it run for 30 min and let's compare to highest/lowest, avrage and such. I am not sure what the command is, ping blalba -t or something, or was it -l? Haha, I have no clue about that sorry, but there should be a command to let it just run without doing just 5 packages.

Yes I can make it run indefinitely, but the two tests should have the same number of runs (or very very close) to be statistically comparable. I'd rather keep it controlled.

Edit: I'm gonna do the test again with 100 times.

Do a 10 min run, please? Then we compared the results after 10 minutes with cable and 10 with wired connection.

Oh, and I found what I was looking for :

Quote :

By default, the XP ping command sends 4 groups of data packets at intervals and reports any successful replies. The '-t' switch sends a continuous stream, only halted by pressing CTRL+C or closing the command prompt window.

Apparently my router can only do 50 at a time. Whoops.

I've gotta go somewhere in a few minutes, I did a 50 test run from the router with the following results:

Round-Trip: 110.6 min, 113.4 avg, 128.7 max (ms)
Packets: 50 transmitted, 49 received, 2% lost

Doing 200 tries from the command prompt, I'll do a 10 minute test later for you TVE.

Oo That's strange..
Well, for now that have to work. Wish I was home at my equiptment, then I could do the same results. Because I have no doubts that wireless will work close to identically as cable for some people, but I think the mayority will struggle obtiaining the same results.

Come on people, are there anyone that can check this? Do 10 min runs with pinging www.lfs.net both with wired and wireless connection.

Go to CMD (by open program in windows), then do ping www.lfs.net -t .
When you have runned that test for 10 minutes, press ctrl+c, and it will stop. Compare the results.

Also, I am not sure if it will be an accurate test, since it's sending one and one package each second, instead of a lot each second.. but for now that will have to work and we'll see

Edit : Thanks a lot mate
Also, to make the results more easly, write what that is done with wired and what's done with wireless, thanks

July 29th, 2009: The day TVE caused LFSers to DDOS LFS.

Here's the longer results, I'll be back later to do more if it's needed.