H2 and Rule 170 ... who is in the right here

1) that's a shame, to be clear the reason oncoming traffic has t wait is because the driver did not make correct observations before starting

2) yes, drivers need to adjust their thinking

3) this is the crux - I would not apply rule 170 to a roundabout, since all exits are equal and none is therefore the side road. While I agree that in general a car continuing round will be looking at other things that an exit they are not taking, in this case the cylcist is in a lane for that exit, so a driver following would be in the same lane.

4) Infrastructure is at fault, traffic light roundabouts should have pedestrian lights for the crossings 

https://www.google.com/maps/@51.6309769,-0.7549117,3a,75y,50.84h,71.8t/data=!3m6!1e1!3m4!1sBKE4M7Ehr6-n10q3IUmdwg!2e0!7i16384!8i8192

or a safe pedestrian crossing (zebra) sufficiently back from the exit to not block the roundabout

https://www.google.com/maps/@51.6309595,-0.7542669,3a,75y,70.77h,83.91t/data=!3m6!1e1!3m4!1ssEEeT7_3rjfLnVghFSP9ww!2e0!7i16384!8i8192

pedestrians should not need to wait for the roundabout to be empty to cross or have to guess which exit all the drivers will be taking, assuming 100% acurate signalling.

You assume that there is good visibility, the pedestrian could be hovering behind a car, halfway across the road - though to be fair this would always have been a stop scenario. More typically, a driver will be intent on traffic gaps and in chancing a gap will not really be in control. It is more recognising that a suitable gap has to include considering pedestrian (and cycle) activity and the reality is that is beyond many drivers and their patience.

But mainly,as you say, it is about adjusting thinking from cars having absolute right of passage where all others are expected to clear their path. It definitely does change driving style to actively seek out pedestrians to prioritise.

Exactly. Otherwise pedestrians would have to treat marked crossing points differently depending on if it's reasonably close to a roundabout exit or not. Imagine teaching children how to cross and having to explain this distinction.

Of course, as a pedestrian I'm still not going to trust my life to drivists stopping, nor teach children (mine are grown now, fortunately) to do so.

Simon E wrote:

 

Rroundabouts, rule 185: "In all cases watch out for and give plenty of room to pedestrians who may be crossing the approach and exit roads"

The impatient bell-end behind you can shove his aggressive use of the horn up his arse.

https://www.highwaycodeuk.co.uk/roundabouts.html

 rule 187, and in the case the pedestrian was not crossing but waiting to cross, this is different guidance to the rule 170

"watch out for and give plenty of room to

pedestrians who may be crossing the approach and exit roads" (187)

vs

"give way to pedestrians crossing or waiting to cross a road into which or from which you are turning" (170)

If the rules for roundabouts were intended to be the same (as some are arguing) then the text would be the same. It is not, because they do not want drivers stopping on roundabouts where following drivers are splitting their attention between the vehicle they are following, any traffic lights, their exit, and most importantly traffic coming round the roundabout from their right

However your last sentance is still correct, in all cases aggresive use of the horn due to impatience is wrong

I'd agree. Roundabouts "work" for the same reasons motorways work, because they ignore all other road users besides motor vehicles, for which they work very well (mostly). But it all falls apart once you have traffic stopping unexpectedly. There was a good piece in the Guardian...
https://www.theguardian.com/cities/2015/oct/19/traffic-lights-roundabout...
“Roundabouts are brilliant at moving car traffic, but not a safe space for people who cycle or are crossing on foot,” says Grant, who, mindful of the emotional nature of this issue, is careful to talk of “people who cycle” and “people who drive” rather than “cyclists” and “motorists”.

The whole point of roundabouts is a) high capacity for motor vehicles with b) good safety for that capacity.  So I agree.  Too much traffic - even just at peaks - the roundabout is not suitable for any vulnerable road users and a grade-separated alternative route for them should be provided.  This situation seems common in the UK.

Roundabouts can work with vulnerable users at the same level if they're designed right but there are then fairly low limits on their capacity.

Doesn't this have to do with the road markings? A side road will have the dashed white lines for both lanes. A roundabout will only have the dashed white lines indicating traffic joining the roundabout must give way. The lack of dashed white lines at the exit of a roundabout suggests a continuation of the carriageway.

Probably going to need Hoarseman to comment but a lot of these technicalities are buried in the regs or highway markings detailed regs.

I refer you to page 119 of the 'Manual for Streets'*; a junction is still a junction even if there are no road markings...

*(it's redacted, but it explicitly states that there is no statutory requirement to mark a road junction)

ooh, and page 87! Which describes the carriageway of a roundabout as circulatory and exiting as 'leaving the junction'...

So it's a junction then. I suppose the old phrase "approach with caution and prepare to stop" is the best policy anyway. 

Yes, it's considered a junction.

I did think the scenario where a very slow cyclist was using the outside lane of the roundabout to turn right is quite similar to a pedestrian crossing an entry/exit. Cars are supposed to slow/stop on the roundabout and allow a cyclist to cross their path.

The highway code has always said to 'watch out' for pedestrians crossing the entry/exits. Just like it says drivers joining the roundabout should look forwards and 'watch out' for vehicles already ahead of them on the roundabout. But it makes it sound like you're 'watching out' for a rule breaker!

Interesting discussion and we have the definitive answer here - it is a junction. But as mentioned above in the comments, whatever the Highway Code says, you can't assume drivers remember what they had to study to pass their driving test however many years ago, nor - due to the woeful lack of communication by the government - that they will be aware of the changes.

And then you have the issue that drivers entering a roundabout are focused on their exit, meaning that many simply will not be looking out for cyclists, or if they do, will often put their foot down to make the exit ahead of the rider ... see here for some examples 

https://road.cc/show/tags/nmotd-roundabout/151017

The roundabout I encounter most often while riding in London is Shepherds Bush, and even though it is traffic light controlled, there is no way am I cycling across that - eastbound, heading towards Holland Park from Shepherds Bush Green, I'd be cutting across the main motor vehicle flow, which heads onto the A3320, effectively an urban motorway.

So what I (and many other cyclists) do is take the bus station slip road, ride across it to the ramp up to the pedestrian crossing on the A3320, or take the underpass, then join Holland Park Avenue the other side.

Similar going westbound, where I need to be on right-hand-side of carriageway to get into the cycle lane. 

Thanks for the link.

It has been suggested that police are more likely to take action for careless driving at roundabouts so I looked through them. One was reported but had no feedback, one in Birmingham had no further action and Thames valley police sent a letter. Those were the only three reported.

Any plans for a similar analysis for other types of near misses and different forces?