Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[aesmith]

Does anyone know how this is actually derived?   Blurb that I've read variously says it's derived from the percentiles of speeds from customers on the same exchange (which sounds impossible) or percentiles of speeds from customers with "similar line characteristics" (which sounds no different from the old BT "estimated speed".

Just asking because I know at least one ISPs is now treating this as the fault threshold, with an unwillingness to act unless the customer's speed is below that MGALS for their line.

[ejs]

Presumably the 2015 Voluntary Code of Practice for Broadband Speeds pdf linked to in that article is a fairly definitive source.

I would prefer to see effort directed at improving the worst lines, rather than towards whoever complains the most. Yet some people carry on expecting to have however many Openreach visits are needed until they are 100% satisfied with their DSL stats, and it's all turned into a stupid charade of knowing exactly what you need to say when complaining to your ISP to get what you want done.

I don't think there's actually any guarantee that your line will never get any worse, yet that's what people expect.

[aesmith]

Cheers, the PDF was the document that referred to "similar line characteristics".   I can't see anything that would stop an ISP simply plucking a figure from out of the air, citing that to the customer and effectively telling him to take a hike if his degraded speed is still over that figure.  I can't see any rigor behind how the figure is supposed to be defined.   I could see this being effective for new customers, where MGALS is quoted at the point of sale, and the customer has the option to cancel if he's not happy with the figure quoted.

[Weaver]

ejs wrote:
> I don't think there's actually any guarantee that your line will never get any worse, yet that's what people expect.

Good point. Leaving aside the ever-present second law of thermodynamics, there's also the question of noise levels and crosstalk in particular. Some noise sources can of course get better as well as worse, but increases in crosstalk are very much to be expected as more users get added into copper networks.