They are sticking to HLG though. This will exempt a some panels (how many?), however I would rather BBC concentrate on delivery of 4k either down the pipe or aerial before it gets fancy with any version of HDR. Football is not something I watch, but Rugby is superb in 4k on another provider. No HDR apart from a few tests at the moment, it is good.
Get the basics of content management and delivery sorted then add the bells and whistles later.
You have a TV that can't do HLG as manufacturers like adding incremental features so you have to buy a new one every year, some have included it already some are saving it for later.
The reason for making HLG non optional (though backward compatible if a manufacturer should choose to do the right thing for their customers) is that if it is not included now then it will take years to get it in all TVs resulting in more people complaining they don't have it. BBC is taking the critisism now to make it better for more people later.
[doublepost=1528042768][/doublepost]
There are ways around this. It might sound silly, but bittorrent was actually invented to address this exact issue. They could have a distributed system, they just choose not to.
bittorrent was invented to avoid paying for bandwidth by using someone else's (ignoring the easier piracy)
The BBC did use a bit torrent like system in the early days of iPlayer, it wasn't much use due to the asymmetric nature of ADSL. That asymmetry remains with BT "fibre" services. At the time ISPs were talking of charging for commercial use if it became widespread.
Consider that for this stream you need 40Mb/s. The BT FTTC upload speed is usually 10Mb/s. So for every stream someone plays it requires 4 or more people to already be playing it to feed them. I hope you can see that is a system that does not scale, you can never add enough viewers to meet demand (system gain 0.25, needs to be at least 1 to work). If the stream was 5Mb/s then each viewer could feed two more, that would scale better (system gain 2). In practice using the upload like this is likely to hit further bottlenecks upstream reducing efficiency.
They say the 10k stream limit is just for testing, it'll not be a limit as it's scaled up for production use later.