Building a new community

I ran a FireBrick course today and met some very nice chaps from a small ISP who deal with the likes of BT and Talk Talk just like we do. (Freudian slip, I initially typed "lies" not "likes" in that!).

They were very interested in our efforts to start a Wholesale Broadband Buyers Forum  (wbbf.uk). They have even considered doing the same themselves. I was really pleased at how much common ground we had.

There are a lot of small (and even larger) ISPs that face issues with carriers on a daily basis, and we all want to work together and with carriers to solve these problems so that we can provide the best possible service to our customers. We really do not want to be fighting BT or anyone - but we need a way of working together that means we do not have to fight, and that is not how it is right now.

One of the biggest issues which has plagued all ISPs for over a decade now is BT SFI (Special Faults Investigation). It is a problem in the first place, as it was created on broken foundations so solve a problem that should not have existed. BT, in classic big company style, do not fix the underlying issues but pile on top new layers of bureaucracy and incompetence. The latest is that SFI disputes are such an issue with ISPs that they have a new process for management of disputes (rather than fix the reason for the disputes in the first place).

I had a meeting many years ago face to face with people that dreamed up SFIs, and in that meeting they said something along the lines that an engineer visit that did not find a fault was chargeable anyway so making an SFI added extra value by being able to check end user wiring and stuff as well for that fee. I pointed out that this was in fact untrue - there was never any charge for such a failed engineer visit - I even challenged him to show me where in the contract and price list (and indeed any bill we had ever had) where there was such a charge. He failed to do so - so the fundamental premise of SFI as adding benefit to customers who already paid for the failed visit was actually flawed. Even so, they did not change their ways. This is typical and key here - they created SFI because not checking user equipment was a waste of the customers money on a pointless engineer visit, when in fact the customer was not charged for such visits - the whole logic for creating the concept of an SFI visit was totally flawed and they refused to admit it.

Over the years they have changed the "service" of SFI one step at a time as we find ways to counter it. First it was "charge for work beyond NTE" and we said "do no work beyond NTE". Then it was "work beyond NTE includes visual inspection of end user equipment" so we had people hide their kit when engineers visited - they still tried to charge for a visual inspection, but of what?! Then it was "Fault not in BT so has to be end user equipment at fault" so we sent end users BT branded modem/routers purchased from BT and within warranty as the only equipment on site, making them liable for the line being faulty or liable for saying their supplied equipment was faulty.

The battle goes on to this day with each new stage being a new battle.

The latest two crazy steps are: (1) That a series of engineer visits where the last one finds and fixes a fault in BT requires payment for the previous (presumably incompetent) engineer visits that did not find or fix the fault, and (2) that all visits where the line tests to an analogue phone spec on arrival shall be chargeable even where they expect us to book an engineer for a PPP fault, or a BRAS fault or a BT modem profile fault all of which happen on a line that is good at a copper test level.

With Talk Talk they create another level of indirection and similar stupidity. At least TT are only at the stage of "SFI" being a product and us saying no thanks, we do not wish to buy that product, now fix the broadband we have already bought. This is a stage we had with BT many years ago.

Somehow we need to find a way to solve this - to get BT and carriers to actually work with us to fix faults, and not create a machine for creating and disputing charges. We also need to work on other things like the cost of backhaul bandwidth, the stupidity of wires only FTTC (leading to yet more SFI issues), issues over backhaul congestion within the network, and stupidities of ordering and faults systems not working properly. We even have to consider OFCOM related stupidities like the latest broadband line migration systems.

So the idea of WBBF is to allow this - to bring together ISPs using BT and TT, and solve these issues. We are gathering momentum now.

The next step may have to be a physical event - in London (pretty much has to be, sorry), and with as much chance to talk as possible (so no pub/restaurant with music or entertainers). Something with food and drink - beers with peers sort of thing - and a couple of presentations and question sessions. Mostly about drinking, eating, and talking.

The trick is very much not to be another LINX, LONAP, UKNOF type of thing (all of which are good, but not what we are aiming for) - but a place to allow as much networking and talking as possible. I did wonder if there is any venue with a big anechoic chamber (with booze and food) - I'd pay to use that!


  1. After six SFI visits over the space of a year including time taken off work, my TT A&A broadband is still hideously unreliable. I've given up trying to get it fixed, it's too much stress and disruption. I'm just hoping I get the predicted FTTC this year, because I'm really fed up with how crap my line is. If six SFI visits can't fix it what hope is there for anyone?

    1. Sorry to hear that, and we are happy to keep trying - in fact "giving up" is a problem, as we don't ever get the final "found and fixed" SFI visit for us to justify not paying for all the other visits.

    2. After six SFI visits that have achieved either nothing or made things worse, I am extremely reluctant to have another one. What is needed is a real engineer that will check the entire line and all joints in it between my house and the DSLAM in the exchange.

    3. I went through much the same process last year, with six visits before BT eventually discovered it was a bad 10Gbps fibre port on their own backbone, an hour's drive from here. (Even then, it was only a happy coincidence and some clever diagnostics from RevK that identified the fault!)

      In that particular case, where a routine configuration change on A&A's end provided a workaround (changing which IP address BT delivered the traffic to on A&A's end happened to change the route within BT's core network, bypassing the faulty segment) perhaps it's something that could be tried routinely, just in case?

      I have wondered if a bit of unbundling might help, too: instead of renting line+port+backhaul from BT Wholesale (who in turn rent bits of that from Openreach), change it to buying the Openreach bit directly: same price, but a bit more control over the Openreach end (where SFIs happen) and perhaps reducing finger-pointing, if you could test and report faults on the Wholesale bit separately.


Comments are moderated purely to filter out obvious spam, but it means they may not show immediately.

ISO8601 is wasted

Why did we even bother? Why create ISO8601? A new API, new this year, as an industry standard, has JSON fields like this "nextAccessTim...