[Chrysalis]

if we were to look at the flawed samknows data, even tho I think its flawed that suggests around 1/3rd of VM's customers have some form of congestion that affects download speeds.

upstream congeston I think is widespread enough in that it is quite possibly on over half the access network. I rarely see a uncongested jitter graph for VM ,they are few and far between.

I agree most customers seem happy as they they either dont care enough or dont know enough to know whats happening, eg. they load a web site it loads hence the service works.

But I wont agree with you on the point that I cant say VM is oversubscribed simply because it doesnt apply to 100% of customers.

Protocol shaping I have always said is bad, seems you realising that now as well. As it takes no account of to how heavy the user is and as to how heavy utilised ports are. I raised this in another thread tho and ignition raised it on VM's forums. Not many took notice.

To be clear I have never told anyone VM's entire network is oversubscribed and do often state it doesnt affect everyone. But I think its wrong to pretend its some insignificant part of the customer base that is only affected.

--edit--

With all that I forgot to post update.

not sure if more work is been done, yesterday latency and jitter significantly dropped, did sharply rise around 4pm tho but then dropped again just before 6pm and then my connection dropped. So I got no data for peak. Its back on now after I renewed the dhcp on the router.