Someone tell Thorpe Lane in Suffolk their internet sucks – they're still loading the page A street in Suffolk has been revealed as the slowest area in Blighty for broadband, where residents have to wait a painful eight hours to download just 45 minutes of telly. Average speeds on Thorpe Lane in Trimley St Martin are 0.68Mbps, 260 times slower than the fastest street, Benford Avenue in Motherwell, Scotland, which …
"Reasons for such sluggish speeds can vary and can include a user's distance from the nearest exchange or difficulties with the property itself – thickness of walls, for example, can affect wireless connections, he said."
True. But mostly it is because of a shit WAN connection. I'm sure someone from an ISP will be along in a while though to explain why it's not their fault, despite being given a LOT of public cash to resolve it.
If people logged on to their router they will see the connection speed to the exchange.
Anything they experience below this is down to their local setup or their ISP back-haul.
I know most people on here will already know this, just thought I'd mention in case there are any developers reading ;)
@Sir Runcible Spoon
just thought I'd mention in case there are any developers reading ;)
If it were not that you appear to be a knight of the realm, I'd challenge you to a joust.
However, I do have some sympathy with the statement - I have come across many developers who would not recognise a computer motherboard if one fell on their head.
That's… awkward to read because of the fact that your browser window was too wide.
Regardless, once you've ruled out local cabling etc., that's TalkBork's problem to fix. They're currently in the process of testing a new network, so (assuming that your problem isn't entirely between you and the cab or exchange) you should see some improvements when they roll that out for everybody.
Looking at your numbers, they break down as:
Attenuation: 51dB
Approx. distance form exchange: 3.7Km
Expected max speed: 4.839 Mbps
Attainable downstream rate: 4.536 Mbps
Attainable upstream rate: 1.047 Mbps (typically the max for ADSL2+ without Annex M support once the ISP has added a little margin for stability)
These numbers remain fairly consistent throughout the screen shots.
Next we have SNR which varies between 21dB (best case speeds) and 25dB (worst case speeds). I'm unsure why Talk Talk are trying to run with such high SNR, as although it will be very stable, it's giving up almost all of your speed (given your distance from the exchange) in the process. Typically you would see this between 3-9 dB.
My recommendation would be:
a) connect your DSL modem directly to the master socket to take any strange, local issues out of the picture (i.e. probably a cabling issue within the house, but can also be caused by a faulty phone or a security alarm) and see if it alters the attenuation. This will identify if you really are ~3.7Km's from the exchange
b) with the DSL modem connected to the master socket, ask Talk Talk to get 2nd line to look at the issue in case yo have been placed on the wrong profile. I have seen this with Talk Talk before and it takes around a week to get to the right team who can fix the issue.
c) I have no experience with the Netgear DSL router you are using but it may be worth trying another unit if you have one. Given the numbers shown, it maybe a router issue stopping the DSLAM in the exchange training correctly.
I would expect you to be getting 3Mbps+ downstream and 1Mbps upstream - not great, but a step up from what you have...
I cant say I've ever noticed a BT logo on it before (then again, I've never really looked that closely at it), but yeah, that's what it looks like.
But someone did live in the house from it's building (which was the late 80s) to when we moved in in 1997, so unless they had something done...
*shrugs*
Hmm that should be able to have an i-plate fitted.
If the lower bit comes off revealing a phone socket it should be compatible, these are supposed to reduce interference which lower the speed (Our line at home went from 2mb to 2.5mb on adsl2). Obviously be aware for anyone saying it made their connection 10x faster, because its only going to help a bit.
For over three years BT have been unable or unwilling to fix the 256/512Kbps (yes, that is the correct way around!) broadband at Seething observatory in Norfolk. We had download speeds of single digit Mbps until BT had a black node issue which 'was fixed', since then all they can tell us is 'the broadband is working". So I hold my head above the parapet with confidence and state that *we* are the slowest "broadband" ... unless someone knows better ...
"So I hold my head above the parapet with confidence and state that *we* are the slowest "broadband" ... unless someone knows better ..."
A relative's rural exchange runs at 512kbit max - but delivers somewhat less than 64kbit/sec (contention). When using webmail the pages frequently fail to load (timing out due to the weight of spamvertising) - POP3 for the win... Interestingly BT did get the gov cash to upgrade the exchange (closing the door to alternative providers), but no cable has been replaced and the exchange remains as it was.
No mobile reception either - welcome to "It's grim up North" Cumbria. :)
BT can't fix it, it's got nothing to do with them. It's Openreach who should fix it.
You could try another provider that uses talktalk's backhaul. It *might* help. Talktalk's DSLAMs are more flexible when it comes to choosing target SNRs etc. Also some routers allow you to specify desired target SNR values, and that might help, even if you continue to use BT.
Actually, scrub the bit about talktalk's anything.... I just checked and Seething Observatory can't even get one of BT's 21CN connections, let alone another provider. The only thing available is the old 20CN connection, which the status checker I used says will get 0.8-2.5mbps download and 0.1 upload, which is clearly not true, and clearly very poor. The annoying thing about 20CN tech (which was BT), is that even if it gives you a sync speed of 0.5mbps download, it'll packet-limit that to 0.25mbps. BT used to call it a bRAS profile, where the actual speed profile is slightly lower than the actual synced data rate. At low speeds it matters a lot.
I was amused to see there is an entry at your postcode "dark hole, toad lane".
For the non-techies it's actually worse than you portray. Most wi-fi speeds dwarf broadband speeds, but contention, cloud density and general lossyness will make it seem like molassus - even though a ping to the router might come back looking fairly healthy. (i.e. latency will be low, and if you only send 5 packets and no-one else is doing much you won't get any packet loss either - best to do a 1000 pings)
Sir Runcible:
I load pings. And frame test by setting packet size.
linux:
ping -c 500 -i .5 -s 1200 <target>
windows:
ping -n 500 -s 1200 <target>
(sadly - can't set the packet interval on winders 7 in the default ping. There is a non standard tool that does)
I've found that latency tends to ramp up as the packet size does. *especially* on wifi, in neighbourhoods where *everyone* has wifi and microwaves, bluetooth this and that, and oddly, electric radiant patio heaters.
Yes, local buffers will do that. Nothing to do with the connection, of course, but you knew that because you're a proper techie using proper techie tools...
Half your latency is probably happening before the packet hits the wire while it sits in a fifo queue.
Would be interested to see...
how many of the measured speeds they're using are via wireless rather than wired
If you want a proper measure of the true capabilities of a line you need dedicated equipment. Oddly enough, OFCOM with the SamKnows people have been doing this for years. I have a "whitebox" so old that it's actually black (a Linksys router).
M.
"degrading aluminium"
If you're lucky, you'll have "degrading aluminium" cable and it will get replaced. Aluminium has higher attenuation making it perform at a similar level to a copper connection that is twice the distance.
In addition, most of the aluminium cabling has been installed by developers rather than BT/OpenReach, so BT/OpenReach expects the housing developer to fix it. The developer promises to fix it when voice quality is impacted...
When I first got on the internet, I had a 9600 baud modem, so that 0.68Mbps speed on the slowest street there is still more than seventy times faster than that, if it makes them feel any better. I have 100Mbps now though, and it's pretty sad to think that their 0.68Mbps speed is actually closer to that old 9600 baud modem than it is to my current cable connection.
"What was the screen used for? "
It was to set all the bazillions of options , using some up / down / enter buttons and a hell of a menu system. There were many things to twiddle with - i had to draw a map
Cant remember the model but it was beige , about the size of a C64 1571 disk drive with a panel on the front that opened to show screen & buttons. think i used it with an Amiga at first.
edit: found it
The company I used to work for used Motorola Codex modems. There were even bigger ones than the 326x series - about twice the size - which we phased out in 1993 or so. They were very impressive looking, with their lighty-up display and buttons. They were still using them in 2001 when I left. We used to connect them to the serial port of postscript printers so we could send our clients proof documents.
A Motorola - I might have guessed, they were all effing huge and made Couriers look small. The manual was only slightly less weighty, but did describe a number of extremely obscure features. I seem to remember they basically worked fine. Certainly should have done for the cost..
Not so a pair of specialist encrypting ISDN terminal adapters I had to configure once. The customer wanted to be sure no-one was snooping their comms. The documentation was a pain, but eventually I managed to get a reliable encrypted connection up and running in the test rig (using actual ISDN lines, didn't have an ISDN exchange simulator)
Packaged it up, included the documentation, and off it went to the customer. They couldn't get it working at their end, and never bothered implementing it, as fan as I'm aware..
We had a normal sized external modem back in the dialup days about 20 x 15 x 7cm which my parents thought was revolutionary. Then one day a relative came to stay with their laptop and a PCMCIA modem that they used to get their emails etc. My mum never having been a shrinking violet asked why we had something the size of a house brick when things like that were available. So I fitted an internal modem to the computer in order that we could ditch the "house brick". Cue questions such as where has the modem gone? Why is the phone line going into the computer? Will I still be able to hear the noises? etc. You can't win.
When I first got on the internet, I had a 9600 baud modem
Obligatory riposte:
"Eee by gum, thou were lucky, I used to dream of 9600, when I was a lad we used semaphore flags, took 3 weeks to load a single web page etc etc"
Actually my first modem was a 1200 down, 75 up - that's bits per second none of this fancy K, M or G malarky.
'minds me, when I first on got on the internet it wasn't even called the internet (we only had dial-up bulletin boards in those days).
I used used a fancy split-rate 1200/300 baud modem. Instead of 'ADSL', think 'AAPL' (asynchronous analogue phone line). The toll calls (bulletin boards were always hosted in cities other than where I lived) were flipping expensive.
"The toll calls [...] were flipping expensive."
Calling them toll calls, are you leftpondian? If so, you will have been saved the insane pricing tariffs of British Telecom when it took over from GPO. A set of tariffs that appeared to be based upon miles of copper travelled, a dice throw, and the phase of the moon. When I first got into BBSs and a bill for several hundred quid arrived, well let's just say that I believe that technically I should still be grounded. I seem to remember the phrase "until the day you bloody die" being used. Yeah, they were pissed. That was a pile of cash back then.
Kind of ironic I can connect to anywhere for "free" now. I think I might have been born prematurely...
my dads never forgot the time I reversed charges
Made a reverse-charge call in mid Wales last year; no mobile coverage at all, the two phone boxes I found didn't accept coins and the card readers were broken (or at least, I couldn't get them to work). A minute or so on the phone to my dad, asking him to ring someone for me, cost him about five quid.
In case anyone wants to avoid the same situation, this is the place on OpenStreetMap. This is also the place, on Streetmap.co.uk which has an OS map showing one of the phones (to the left, just above the words "Ty'n y Celyn"). The other phone was right in the middle of the village and presumably has since been removed.
M.
Thorpe Lane isn't far from me - I live in Felixstowe. If you check the satellite view on Google Maps (or map provider of your choice) you'll see that there are very few houses, so I'm wondering how many broadband-connected premises made up the sample from which the average was derived - there might have been only one, with a misconfigured router or some other anomaly. The Trimleys (there are two - St. Martin and St Mary) also aren't a million miles from BT's Adastral Park, so there's plenty of broadband expertise in the area :-) Disclaimer: between 1997 and 2014 I worked for BT, but not involved in broadband provision, and not interested in defending any ISP's reputation.
>> no-one gives a fuck about the Trimleys.
Worked at Schlumberger 1994-96. We had Mosaic and we were all in shock and aaawwwwwe. And yea, I can attest to the lack of f'ck about the Trimleys, even then. But because of that, I always imagined Suffolk had great Internet. Wrong it seems, unless you're a dirty energy monster.
I now have unreliable Internet, rather than low speed, caused by audible crap on the landline. The fault is 'intermittently persistent' and very annoying when the DSL connection drops just as the F1 lights go out.
The ISP tells me "If Outreach call, and they don't find the problem, we'll charge you." And I now say "I just don't give a flying f'ck any more....." and arrange Freesat.
Everyone's got a story.
Yeah but there are surely plenty of other places in the UK where the same is true, this terrible performance indicator isn't for Trimley(s) as a whole, its for one mostly unpopulated lane, leading to a farm - which is why I think it is the result of a particularly small sample.
Three little words - East Coast Cable. Since borged into Virgin.
So there are places on the Felixstowe peninsula where you are 5 minutes walk from the countryside, 10 minutes walk from the sea, and can get 200 Mb/sec Internet over co-ax, Although apparently not in the less fashionable parts of the Trimleys.
"Why should only yokels get the best of both?"
I live in a metropolitan area and have 100Mb/s BB (could have up to 300 but don't need it). In 10 minutes I can be in Newcastle city centre getting rat-arsed with the crowd or, in the same 10 minutes travel, be in the countryside.
Sucks to be you :-)
8hrs = 45mins of video => 22hrs < 2h4mins of video... About normal for a non-kiddie film
Amusingly, I reckon that means you could fly to New Zealand, go for a tour of The Shire, pick up a copy of the LotR boxset from the gift shop and fly back quicker than downloading the films...
I wonder what the standard size, in MB is for a film they used for this stat? Is it the full directors cut of Dune in 1080p in MPeG2 or a more normal 2hr film efficiently encoded with x265? I find that the best encodings in X265 are about half the size of the best X264 encodings for the same perceptual quality (and much, much smaller file sizes than DVD-like MPeG2). All IMHO of course.
I've a neighbour on the same set of connections, but different ISP who *cannot* stop futzing with modem settings (DSL). I've been connected to the "net" since DARPA. I keep telling him that the crap he's reading on t'interforuwebs is meant for folks in (some other country on another continent, with different standards, with different provisioning). His networking knowledge is based on (dedicated enterprise two hop network on fibre) one round of experience. His household network speeds are utter crap. I go through -- tidy things up, move his wifi to a less loaded channel (there are just *no* free channels around here) and clean up his modem settings and two weeks later he's read some stupid post from the back woods of nigeria about tweaking wifi to get 5 times the throughput and its all crap again... including trashing his neighbours wifi to boot.
I've told his offspring to beat him with a bedpost if he goes near the modems or the wifi routers.
Fed up hearing of the amount of money being thrown at BT and then Openreach to get the internet out to “rural” areas. Less than ten miles from three towns and less than two miles from the local village exchange and the best we get is 0.5 Mbps. I have done speed tests several times and it has been as low as 0.33 Mbps. In the immediate area there are about 50 houses and the majority are on line, this is not including the village the exchange services, so it is not a lack of use that could be the excuse. Frequent requests for updates just elicit the answer of “don’t know”, when or even if we will ever get decent service, let alone cable. Keep meaning to research cost of satellite connection.
One of my local hostelries has two telephone lines coming in - one for punter broadband, one for the tenants. The lines come into opposite sides of the building, about 30m from each other. The punters have to fight over a total bandwidth of 2MB on a sketchy network with somewhat, er, elastic latency, whilst the tenant doesn't know what to do with the rock solid 70+ which pings in single digit milliseconds.
People do miss one big factor in the usefulness of Internet connection speeds. My connection would be classed as horribly slow, but I am the only one using it so it's OK. Yes, I'd like more, but shouldn't we at least have some idea of the number of people we assume are in a household when we're looking at these figures?
Let's say three people. It's enough to need 10Mbs for adequate performance.
"Let's say three people. It's enough to need 10Mbs for adequate performance."
No, it isn't. If you're viewing complex interactive websites, and all three of your working at once, sharing 10mbps would cause annoying delays while waiting for content to load.
And if you're trying to stream video, it's totally inadequate. You need about 30Mbps minimum for good 4K HDR video for a start.
We need to stop fannying around saying that '10Mbps is OK', '30Mbps is superfast' and so on. It might just pass today, but we need to build the networks out for the next several years. I have 30Mbps and it's only just about adequate. Next time I move, the speed of the broadband, and the technology used, is going to form a very important part of the decision-making process.
"Let's say three people. It's enough to need 10Mbs for adequate performance."
No, it isn't. If you're viewing complex interactive websites, and all three of your working at once, sharing 10mbps would cause annoying delays while waiting for content to load.
Horses for courses, innit? It all depends on what you expect to do with it.
I live in a household of two adults and four school-age children. We certainly don't stream 4k Netflix (and I acknowledge that Netflix recommend a minimum 25Mbps for that) or do online gaming but we do do many the other things you might expect. Our 6Mbps (down sync) connection rarely limits things and quite frankly nobody here is bothered if a website loads in 2 seconds, 5 seconds or 10 seconds. Websites in particular are "bursty" and it's amazing how little difference there is between one person browsing and three people browsing.
If we used Netflix, if two of the children were into online gaming, if we used IP phones or video calling, then I could see the 6Mbps being limiting occasionally. We have FTTC available locally, but (depending on supplier) as it would basically double the "broadband" part of our "broadband and calls" package (c£7/m to c£13/m), it's not something I'm rushing to order.
M.
In about 2008, I was talking to an American guy who worked for a company that wrote remote support software, and he explained that it does actually work with dialup, even poor dialup (but obviously not very well). I remarked that I can't imagine too many people on dialup and he told me a story of someone he knew who lived on a farm deep in the American midwest, and not only could they not get any type of broadband, but the dialup was pretty flaky too. He went on to explain that they only had one wire coming in from the phone company to the house. Not one pair. One *wire*. I asked how that worked and he told me that the return path was a metal fence that ran from the house to the telephone pole, a few hundred yards away. Now that's pretty basic!
He went on to explain that they only had one [phone] wire coming in from the phone company to the house. Not one pair. One *wire*.
Single-wire electricity supply is (used to be?) quite common in some very remote areas, for example in Australia. Not quite as daft as it sounds, you just need a really good earth connection (return path is via the bulk of the earth) and it saves you what could be miles and miles and miles of cable.
M.
... not the measly 200ish Mbps mentioned in the article. B4RN gives you synchronous (yup - read that again to make sure - SYNCHRONOUS!) 1 Gbps speed. It's a measurable 950+ Mbps speed. Both directions.
Rural communities can do broadband better than Openreach, if only the gubbermint was prepared to allow them the chance. They are not making it easy, yet.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
