I also asked them to move me off a CGNAT network, which they did with no issues, to their credit.

Very true, my apologies

Very interesting indeed, thank you! My biggest concern has always been gaming pings but I also WFH every day so this is definitely something to test for.

Also had a browse of the other thread linked further down and the comments aren't very promising to say the least...

Not to pile on the negativity but worth mentioning - As I've left my TalkTalk connection as the primary today for work, I have noticed that video on Zoom calls is much clearer. Zoom has always worked fine on HeyBroadband but I've had some compression on video. I'd always assumed it was just normal for Zoom because many of my calls have well over 50 participants. But having joined some calls today with the other connection, the incoming video is much clearer. Network stats in Zoom seem pretty comparable between the two but there's definitely a difference in video clarity.

I realise this sounds somewhat unlikely and maybe a little petty but it has genuinely surprised me that the two connections can be so different in performance on a like for like comparison. If I could share screenshots of the difference between them I would but obviously can't do that with work Zoom calls.

It just goes to show that you can't trust a speed test as an indicator of true real world performance. Don't get me wrong, the HeyBroadband connection has been very stable and has top notch bandwidth for big downloads but does seem to be let down by some routing & real world usage. I expect it's quite easy for an ISP to prioritise speed testing sites to make their connectivity appear superior.

For gaming, HeyBroadband has very good latency almost all the time. I suspect their routing to bigger providers like AWS & Azure where many gaming services are hosted is set up better than general traffic. For me, Rocket League is easiest one to use as an example, as you hit multiple different servers in a short period of time. I sometimes get a ping as low as 10ms but also as high as 40-45ms depending on the server (I'm guessing but I think they have some in the UK & others elsewhere in Europe). I've never really had any issues with high latency, except for a couple of periods of streetwide issues mentioned in the other thread. However, my TalkTalk connection does still get around 10% lower latency across the board, going as low as 8ms in Rocket League.

Somewhat relieved to hear that gaming latency is OK, though I mostly play Deep Rock Galactic, which is P2P rather than server based.

Meanwhile I've spent the morning setting up the BQM that I had no idea was a thing, and I've opened a route between my office where they have a 10Gb dedicated line and the OpenSpeedtest server on my LAN so that I can compare things that way rather than relying on the easily prioritized Ookla test.

Tomorrow will be very interesting!

Ugh well the install didn't happen in the end. The pole is condemned/defective and the guy insisted the whole thing needs to be replaced before he can do the install. I mentioned that the Swish installer used a cherry picker but he wasn't having it, so it's now re-booked for next week. Of course the pole won't have been changed in that time so I'm not sure what'll happen then...

On the one hand I'm mad at myself because I knew it was condemned but didn't think to mention it to them in advance, and I'd chosen today specifically so that it lined up a convenient end-of-the-month payment date. On the other hand it's giving me extra time to gather comparison stats from Swish.

When my most recent fibre service was activated in Suffolk, my nearest pole was also "condemned" (actually just Cat D) and after probably 6 weeks of to-n-fro around the houses at Openreach, it was magically "un-condemmned"

...in other words they sent out a picker, rather than climb it.

To be fair that pole is probably older than me 🤣

A D pole , is defective, it doesn’t necessarily mean dangerous or decayed, a condemned pole is not the same as a D pole ….a pole can have a D classification for several reasons, not just decay , but can’t be climbed by using a ladder ….if the D is for something other than decay , access by alternative means ( like a hoist or scaffolding ) is allowed.

If the Alt Net doesn’t have the wherewithal to arrange a hoist on the date of the installation should a D pole be encountered, or don’t survey in advance, that’s up to them…poles are periodically checked , and if decay is detected they enter a program of replacement, if this isn’t as quick as an Alt Net requires, they are at liberty to install their own infrastructure

In our case Openreach were insistent (initially) that the pole definitely needed replacement - that was the directive from their surveyor. No other furniture could could installed on the pole.

We were told the pole was going to be replaced. It was in the queue waiting for the pole replacement team. It was 6-8 weeks later after much prodding and chasing that Openreach came back out again and decided that the first surveyor was wrong and it was fine and just a picker was needed.

No alty’s were involved here, it was pure Openreach infra (Ethernet rather than FTTP) They changed their mind and kept the pole. The pain of it was it took them 8 weeks of fumbling to decide. 😂

In reply to a post by Pheasant:

The pain of it was it took them 8 weeks of fumbling to decide. 😂

To be fair, that's pretty much "next day" if you measure time using an Openreach calendar...

Well, it's installed now. The guy had no record of me wanting to use my own router so there was a long delay and multiple phone calls to actually get connected, and then once I was connected they hadn't assigned the static IP that I ordered. Another 20 minutes on the phone to get that escalated and they're promising a call back within two days.

Not the best start so far but I am at least seeing the speedtest speeds as advertised, and it's 2 fewer hops to work's VPN gateway, and only 8 hops to thinkbroadband.com. No weird trips via Amsterdam or Spain...

Edit: Just got forcibly reconnected and I'm now on a static IP. Hops look much the same as before.

Edited by Noiz (Wed 06-Nov-24 15:43:45)