best sim only contract

Maybe your 'fidonet' isn't the best server to test Three on, there might be a slow peering link in the way or something. You should always test with a few to get a good idea.

You don't ever need to do that if you understand how speedtests, the internet, and 3G works.

When you can explicitly monitor the saturation of your 3G link while doing a speed test - indicating your local hop is maxed out, regardless of any other bottlenecks that may exist towards the internet, it really doesn't matter what server you use.

Incidentally the route from here to Fidnonet over 3's network is the same as that of a lot of UK traffic, if 3 can't deliver 2Mbps to major parts of the internet because of a bad transit link, that's big a problem in itself.

jabbamk1 · 08-12-2013, 00:11

Quote:

Maybe your 'fidonet' isn't the best server to test Three on, there might be a slow peering link in the way or something. You should always test with a few to get a good idea.

Come on mate, fidonet is just fine. Just because i switch server doesn't mean it's going to get a better result.

I mean look, let me try it on the newbury server and.... HOLY S***!

http://i481.photobucket.com/albums/r...psb783d27c.jpg

qasdfdsaq · 08-12-2013, 00:15

You're obviously on Sodafone LTE since the Newbury server is run by their partner-in-crime, Vodafone.

Obviously his is a photoshop because it was 3 whole minutes between screenshots and his battery didn't go down one single percent.

Three · 08-12-2013, 01:21

Generally I find the Newbury server works best. Hosted by Vodafone..

natbike · 08-12-2013, 02:30

I tend to go with UNO or Namesco for fast connections. As qasdfdsaq said, for phones it doesn't matter as much and you should want to know that connections to London are good regardless of whether you can get a great speed from a local server.

Just look at the number of peering point and datacenters in and around London, most traffic will be passing there at some point.

qasdfdsaq - What part of the country are you in and I take it o2 are working best for you in general?

Thine Wonk · 08-12-2013, 09:36

There are people that really don't understand how the internet works here. Nothing about routing to "London" there are plenty of different network routes to different providers, transit and peers. Actually, there's a hell of a lot of stuff not hosted in London too.

My two speedtests 3 minutes apart showing 0.1Mbps from Fidonet and more than 10Mbps from a different speedtest site shows you that taking one site in isolation is not a good idea. We're not just testing the RAN to core, we're testing all the way to the endpoint and not always via transit, that's why you need to test a few.

The other thing is your tests are when tethered by the looks of it, I say this because you've given the desktop list to the speedtest result not a mobile and the operator name is "Three" (which is comes up with when doing desktop tests) as opposed to "Hutchison 3G" which is comes up with when mobile speedtests are performed. Due to the way Three traffic prioritises doing tethering tests could give a false result.

C:\Windows\system32>tracert 93.184.221.133

Tracing route to 93.184.221.133 over a maximum of 30 hops

1 13 ms 4 ms 2 ms 192.168.43.1
2 502 ms 411 ms 427 ms 172.31.33.161
3 474 ms 412 ms 409 ms 172.31.37.26
4 142 ms 70 ms 65 ms 172.23.162.6
5 72 ms 67 ms 34 ms 188.31.255.130.threembb.co.uk [188.31.255.130]
6 52 ms 53 ms 43 ms 188.31.255.21.threembb.co.uk [188.31.255.21]
7 99 ms 220 ms 217 ms te3-4.ccr01.lon11.atlas.cogentco.com [149.11.78.
5]
8 * 1205 ms * te0-0-0-2.ccr22.lon01.atlas.cogentco.com [130.11
7.48.181]
9 85 ms 75 ms 103 ms be2398.ccr21.lon01.atlas.cogentco.com [130.117.5
1.209]
10 88 ms 77 ms 41 ms ldn-b4-link.telia.net [213.248.70.237]
11 99 ms 45 ms 47 ms ldn-bb2-link.telia.net [213.155.132.196]
12 93 ms 53 ms 43 ms ldn-b3-link.telia.net [80.91.249.176]
13 57 ms 49 ms 46 ms edgecast-ic-147960-ldn-b3.255.12.195.in-addr.arp
a [195.12.255.50]
14 103 ms 69 ms 79 ms 93.184.221.133

Trace complete.

C:\Windows\system32>tracert 93.188.183.17

Tracing route to 93.188.183.17 over a maximum of 30 hops

1 3 ms 3 ms 1 ms 192.168.43.1
2 46 ms 42 ms 45 ms 172.31.33.97
3 134 ms 79 ms 54 ms 172.31.37.26
4 47 ms 40 ms 39 ms 172.23.162.6
5 93 ms 34 ms 38 ms 188.31.255.130.threembb.co.uk [188.31.255.130]
6 43 ms 43 ms 84 ms 188.31.255.21.threembb.co.uk [188.31.255.21]
7 103 ms 52 ms 50 ms te3-4.ccr01.lon11.atlas.cogentco.com [149.11.78.
5]
8 65 ms 64 ms * te0-0-0-2.ccr22.lon01.atlas.cogentco.com [130.11
7.48.181]
9 60 ms 52 ms 46 ms level3.lon01.atlas.cogentco.com [130.117.15.246]

10 115 ms 58 ms 58 ms vl-3604-ve-228.csw2.London1.Level3.net [4.69.166
.157]
11 64 ms 83 ms 906 ms ae-25-52.car5.London1.Level3.net [4.69.139.102]

12 42 ms 50 ms 47 ms 195.50.118.226
13 384 ms 73 ms 67 ms core-3-b1.access.fido.net [80.252.124.30]
14 61 ms 46 ms 48 ms 93.188.183.17

As you can quite clearly see, one route goes via Level3 to Fido, where as Telia is used for Vodafone. If you tried this from a different internet connection like a Virgin Media fixed broadband you see THN before Fido, so it can be a peering link in the middle that is slower depending on the BGP routing.

.1.244.73]
3 15 ms 17 ms 17 ms brnt-bb-1c-ae11-0.network.virginmedia.net [80.
240.173]
4 18 ms 24 ms 62 ms brhm-bb-1c-et-410-0.network.virginmedia.net [6
253.175.210]
5 * 23 ms 23 ms tcl5-ic-2-ae0-0.network.virginmedia.net [212.2
.15.210]
6 24 ms 24 ms 21 ms xe-4-1-1-core0.thnor.uk.as6908.net [195.66.225
2]
7 22 ms 22 ms 23 ms te4-2-cr0.thn.uk.as6908.net [62.149.52.6]
8 23 ms 24 ms 21 ms l1-c1.access.fido.net [84.246.196.249]
9 21 ms 22 ms 22 ms core-3-b1.access.fido.net [80.252.124.30]
10 20 ms 20 ms 23 ms 93.188.183.17

That's why I say always try a few, and I have evidenced this with a 0.1Mb/s speedtest to Fido and then 3 mins later a 10Mb/s speedtest to the Vodafone server.

Thine Wonk · 08-12-2013, 10:21

Apologies I got the wrong IP for Vodafone, here's the correct one.

From Virgin Media

3 14 ms 14 ms 15 ms brnt-bb-1c-ae11-0.network.virginmedia.net [80.1.
240.173]
4 18 ms 17 ms 18 ms brhm-bb-1c-et-700-0.network.virginmedia.net [62.
253.175.206]
5 23 ms * 23 ms tcl5-ic-2-ae0-0.network.virginmedia.net [212.250
.15.210]
6 45 ms 40 ms 43 ms lndgw2.arcor-ip.net [195.66.224.124]
7 40 ms 38 ms 40 ms 85.205.116.14

Three:
1 10 ms 1 ms 1 ms 192.168.43.1
2 84 ms 78 ms 66 ms 172.31.33.129
3 62 ms 63 ms 51 ms 172.31.37.30
4 67 ms 50 ms 39 ms 172.23.162.6
5 98 ms 72 ms 78 ms 188.31.255.130.threembb.co.uk [188.31.255.130]
6 75 ms 69 ms 78 ms 188.31.255.21.threembb.co.uk [188.31.255.21]
7 92 ms 67 ms 55 ms te2-4.ccr01.lon11.atlas.cogentco.com [149.11.78.
13]
8 123 ms 86 ms 80 ms te0-0-0-34.ccr22.lon01.atlas.cogentco.com [154.5
4.61.53]
9 114 ms 55 ms 45 ms level3.lon01.atlas.cogentco.com [130.117.15.246]

10 55 ms 51 ms 45 ms vl-3602-ve-226.csw2.London1.Level3.net [4.69.166
.149]
11 62 ms 45 ms 52 ms ae-229-3605.edge4.London1.Level3.net [4.69.166.1
7]
12 56 ms 46 ms 50 ms 195.50.122.66 - This is Lynx
13 114 ms 77 ms 78 ms 85.205.116.10

As you can see, different providers in the way, including the Lynx internet exchange in one of the cases where there is peering involved. Things like this can make a difference, hence the need to try a few test sites and not tethered on Three as that can also affect the results.

dave1981 · 08-12-2013, 11:28

Had a thought. The landline at my house has a max upload of 0.38mbps. The 3G is about 10 times faster when I use it at the house! Read that 3 allow tethering on their unlimited £15/18 a month deal. Anyone here use tethering? Would be handy for uploading bigger files

Thine Wonk · 08-12-2013, 12:20

Quote:

Had a thought. The landline at my house has a max upload of 0.38mbps. The 3G is about 10 times faster when I use it at the house! Read that 3 allow tethering on their unlimited £15/18 a month deal. Anyone here use tethering? Would be handy for uploading bigger files

They do allow tethering on the One Plan, just remember that P2P and torrents will be throttled heavily and that between 3PM and Midnight where capacity is low all tethering is has a fixed pool of bandwidth, this can means that speeds on tethering slow down so that mobile broadband and phone users aren't impacted by those using Three on tethering all the time to get cheap mobile broadband.

qasdfdsaq · 08-12-2013, 12:44

Quote:

Due to the way Three traffic prioritises doing tethering tests could give a false result.

So if the speed test represents the speed I get when using my connection for what I use it for, it's a "false result"?

How is it any more accurate to return a speed someone else will get if they explicitly block me from attaining that speed?

Also you've seemed to have missed the point. As I've already said, when you can see whether a speedtest is saturating your link to the core, there is no need to try different servers because what happens after the core won't make any difference.

qasdfdsaq · 08-12-2013, 12:47

Quote:

Had a thought. The landline at my house has a max upload of 0.38mbps. The 3G is about 10 times faster when I use it at the house! Read that 3 allow tethering on their unlimited £15/18 a month deal. Anyone here use tethering? Would be handy for uploading bigger files

For uploading bigger files you're better off on a network that has 4G.

Thine Wonk · 08-12-2013, 14:21

Quote:

So if the speed test represents the speed I get when using my connection for what I use it for, it's a "false result"?

How is it any more accurate to return a speed someone else will get if they explicitly block me from attaining that speed?

Also you've seemed to have missed the point. As I've already said, when you can see whether a speedtest is saturating your link to the core, there is no need to try different servers because what happens after the core won't make any difference.

You are comparing tethering tests, whereas people weren't specifically talking about tethering. They were talking about general network speeds to that was misleading.

It does make a difference, I have proved 2 side by side tests, one showing 0.1Mb/s and one showing 10Mb/s with the only difference being the speedtest site. You cannot measure at the core as you don't have access, all you can do is measure from the consumer endpoint and comparing with a different provider as I have already explained uses different internet links.

I have given you the technical reason why and evidenced it. I can't help it if you don't understand.

Three · 08-12-2013, 14:35

Quote:

Had a thought. The landline at my house has a max upload of 0.38mbps. The 3G is about 10 times faster when I use it at the house! Read that 3 allow tethering on their unlimited £15/18 a month deal. Anyone here use tethering? Would be handy for uploading bigger files

Tethering speeds are restricted to a fixed pool of bandwidth during peak times. I've found generally though that's it only download speeds affected during peak times, upload speeds stay pretty consistent with off-peak speeds.

petef · 08-12-2013, 15:52

For price comparison I recommend that you look at http://www.paygdeal.co.uk/ It allows you to enter your expected calls, texts and data per month and searches through all the PAYG rates with their add-ons.

My site at http://petef.22web.org/payg.html is more comprehensive for listing the operators but only has the basic rates, not add-ons.

For SIM only contracts see http://kenstechtips.com/index.php/th...nternet-access though it is a few months old, so rates may have changed.

qasdfdsaq · 08-12-2013, 16:08

Quote:

You are comparing tethering tests

I am comparing the speed I pay for and I get.

Quote:

You cannot measure at the core as you don't have access, all you can do is measure from the consumer endpoint

You can measure the core because all your traffic goes over it. As I've told you three times a simple understanding of network topology and bottlenecks and this is easy.

Quote:

I have given you the technical reason why and evidenced it. I can't help it if you don't understand.

All you've told me so far is stuff I already know is irrelevant, and that you don't seem to get what is. All your traceroutes evidence is that you don't actually understand what they're doing. It's very easy to determine if your bottleneck is en route to the core and thus nothing to do with the server, internet or peering - you've done so yourself though apparently without realizing it.

If you fail to understand basic networking, and insist on being ignorant of the fact, I can't really help you either.

Thine Wonk · 08-12-2013, 16:20

Quote:

I am comparing the speed I pay for and I get.

No, you did tethering test when others were talking about pure mobile data speeds. Knowing that Three throttle teething and not saying they were tethering until it was found to be the case.

Quote:

You can measure the core because all your traffic goes over it. As I've told you three times a simple understanding of network topology and bottlenecks and this is easy.

Your traffic goes over the core, but you can't measure FROM it, FROM it. Traceroutes only show you ICMP latency, they cannot determine throughput via different routes.

Quote:

All you've done is deny what I've said three times already. All your traceroutes evidence is that you don't actually understand what they're doing. It's very easy to determine if your bottleneck is en route to the core and thus nothing to do with the server, internet or peering - you've done so yourself though apparently without realizing it.

If you fail at basic networking, and insist on being ignorant of the fact, I can't really help you either.

You simply don't have a clue, if you did you would be able to explain (using your funny logic) why your 'Fido' speedtest server gives a throughput of 0.1Mb/s and another test site gives 10Mb/s + back to back. Everyone who uses speedtest sites knows that they are variable depending on which network routes you go through from your provider and to the host of the speedtest server.

shaggy_x · 08-12-2013, 17:06

Quote:

thinking the tesco £15 a month 2000 minutes/3gb data is the best
I don't need 2000 minutes but would like as much data as poss

currently on O2 and from what I can gather I'll get the exact same coverage/service with tesco mobile

any others ?

For the same price have a look at 3 one plan sim with unlimited data and tethering

dave1981 · 08-12-2013, 17:12

Quote:

For uploading bigger files you're better off on a network that has 4G.

no 4G where I live unfortunately

it's not even that the files are that big, more that my landline internet has errible upload

Have read that Three network has no 2G (would explain the name) so that means unless there's 3G the phone won't work for calls/texts ?

qasdfdsaq · 08-12-2013, 17:16

Quote:

No, you did tethering test when others were talking about pure mobile data speeds. Knowing that Three throttle teething and not saying they were tethering until it was found to be the case.

I'm sorry I pay for mobile data with tethering on three different networks and only one deliberately throttles the speeds.

However I'll just say this. I'm not tethering, you simply made that up and assumed it to be true.

Quote:

Your traffic goes over the core, but you can't measure FROM it, FROM it. Traceroutes only show you ICMP latency, they cannot determine throughput via different routes.

Did I say throughput? I said you can determine if the bottleneck is NOT the server. You simply made that up and assumed it to be true.

Quote:

You simply don't have a clue, if you did you would be able to explain (using your funny logic) why your 'Fido' speedtest server gives a throughput of 0.1Mb/s and another test site gives 10Mb/s + back to back.

Umm, you've not posted a single verifiable speedtest result for either of those two, and once again you've completely missed the point. Instead of beating around the bush with your irrelevant mumbojumbo why don't you just admit you're wrong? I never said you can't make up any random variability you want. I said it is possible to tell when a speedtest is maxing out your connection so you do not need to try a different server. I've said this five times. How hard can it be to get that through your thick skull?

Incidentally I've posted 10 different verified results all of which are consistent with Fidonet - explain that (without using doctored screenshots).

Thine Wonk · 08-12-2013, 17:57

Quote:

I'm sorry I pay for mobile data with tethering on three different networks and only one deliberately throttles the speeds.

However I'll just say this. I'm not tethering, you simply made that up and assumed it to be true.

You are posting speedtests using the desktop site which provides you with a sharing link which is different to the mobile app.

Quote:

Did I say throughput? I said you can determine if the bottleneck is NOT the server. You simply made that up and assumed it to be true.

What else do you think you're measuring in a speedtest? The bottleneck isn't the server, but it could be the hop before the server as I've explained countless times, but it's like trying to hammer a nail in with a banana.

Quote:

Umm, you've not posted a single verifiable speedtest result for either of those two

See post #25

qasdfdsaq · 08-12-2013, 18:05

Quote:

You are posting speedtests using the desktop site which provides you with a sharing link which is different to the mobile app.

Yes, but that does not mean I'm tethering.

I prefer using the desktop website on my phone because it actually provides a more reliable result, partly because it uses four simultaneous connections instead of one, as well as a host of other reasons.

Quote:

What else do you think you're measuring in a speedtest? The bottleneck isn't the server, but it could be the hop before the server as I've explained countless times, but it's like trying to hammer a nail in with a banana.

No, it can't. Because as I've explained countless times, there are simple ways to see where the bottleneck is. You just seem to be blind to that fact.

You have ONE connection to the network and the core. If EVERYTHING over that connection slows down, then the cause is NOT the path it takes at the other end.

Quote:

See post #25

Neither of those are completed speedtests nor are either verifiable. No more worth believing than jabbamk1 posted in #27. Though frankly I'd take his result over yours anyday.

Thine Wonk · 08-12-2013, 18:13

Quote:

Yes, but that does not mean I'm tethering.

I prefer using the desktop website on my phone because it actually provides a more reliable result, partly because it uses four simultaneous connections instead of one, as well as a host of other reasons.

No, it can't. Because as I've explained countless times, there are simple ways to see where the bottleneck is. You just seem to be blind to that fact.

Neither of those are completed speedtests nor are either verifiable. No more worth believing than jabbamk1 posted in #27. Though frankly I'd take his result over yours anyday.

I see you're moving the goalposts now, after saying there was nothing you're now saying they aren't completed (they had done the download part when I took the screenshot.

You say "there are simple ways to tell" but then don't elaborate. Quite simply just trying to talk yourself out of it.

First it wasn't about throughput, then it was. I'm not going to continue to debate with you as it's a waste of time, you'll either change tact, wriggle, twist or just post no descript things like "I can tell" when you really can't.

qasdfdsaq · 08-12-2013, 18:19

Quote:

I see you're moving the goalposts now, after saying there was nothing you're now saying they aren't completed (they had done the download part when I took the screenshot.

Umm, you're making complete rubbish yet again:

Quote:

Umm, you've not posted a single verifiable speedtest result for either of those two

Quote:

Neither of those are completed speedtests nor are either verifiable.

'nuff said.

Thine Wonk · 08-12-2013, 18:23

What do you think I've done, sat on photoshop to try and argue against your weird and wacky way you think internet routing works? I can assure you I haven't, but I suggest we leave it there as I cannot be bothered to argue any more.

jabbamk1 · 08-12-2013, 18:29

Quote: