This changeset accidentally displaced 18 nodes by a few metres. I reverted that accidental change in 77174289

I think this is also an issue in changesets 76617069 and 76685993; I haven't checked to see if you've made similar edits in any other changesets.

I've just noticed this changeset, and I disagree with the tagging. You may be correct that these aren't ' normal highways', but they aren't bus guideways either - highway=bus_guideway should be reserved for sections that are physically guided (including perhaps brief road crossings where the guidance resumes immediately after the crossing). The wiki suggests (and I agree) that highway=service would be more appropriate, perhaps in combination with access=no and bus=designated (or something else consistent with local usage).

highway=bus_guideway

I've just noticed this changeset, and it looks like you've changed a lot of ways to highway=bus_guideway where these sections are not guided (and in some cases are used by unguided buses as well). I think this is incorrect tagging - highway=bus_guideway should be reserved for sections that are physically guided (including perhaps brief road crossings where the guidance resumes immediately after the crossing). The wiki suggests (and I agree) that highway=service would be more appropriate, perhaps in combination with access=no and bu=designated (or something else consistent with local usage).

highway=bus_guideway

Oops, I forgot that one - thanks!

This change is wrong - the old A14 exit slip is not a motorway (let along a motorway_link), never has been, and never will be. Access=no is unnecessary because it is no longer tagged as an extant highway (it might still physically exist for a few days, but leaving it tagged as a highway would be more confusing and difficult to keep up-to-date). I have reverted this change in changeset/76936447

Oops, I forgot to use _link on the turning lanes.
osm.wiki/Highway_link says "The _link tags are used to identify slip roads/ramps and "channelised" (physically separated by an obstruction or painted island) at-grade turning lanes" - in this instance there are physical islands forming part of a larger painted island, and knowing about these would be useful for someone trying to walk.cycle round here (as a cyclist trying to get towards Milton I miss my intended turning at this junction more often than not). There was also the anomaly of a pedestrian/cycle crossing where there was no highways drawn through it.
I'll change the turning lanes the turning lanes to links; let me know if that's better.

I've now reverted this changeset, and added the no U-turn restriction, in changeset/76881246

Hi Stu,

Firstly, here are a couple of initial links that you might find relevant:
osm.wiki/Highway_link
osm.wiki/Junctions

In general, I reckon there are two good ways of finding out how to map things - either look for similar situations elsewhere on the map to see how they are mapped, or search on the internet to see what other people have written about mapping that sort of feature. In particular, browsing the wiki is probably a good way of finding some of the latter information. However, you should bear in mind that there is often more than one to map things (e.g. sometimes there's an accuracy/time trade-off), and there's sometimes disagreement between different contributors. Many proposals that are made never become widely accepted.

As for your last paragraph: Usually the the angle at which ways are drawn is some compromise between representing the point at which two ways join, and representing the centrelines near the join in a consistent way. So a single carriageway becoming a dual carriageway cannot avoid including a wiggle in the line mapped for traffic in each direction, but in the case of a slip road this can often be avoid by extending the sliproad's way along its centre line to meet the mainline before actual diverge point. Routing often doesn't care about the exact geometry at the intersection - any constraints are represented by oneway roads or explicit turn restrictions. And I've no idea what you mean by "road type progression", but I think you can probably find some helpful information by searching the wiki or the wider internet.

As for this changeset/junction, I think the best way to proceed would be to revert this changeset (since the previous version probably had appropriate geometry) and add the turn restriction - if you're not sure how to do that then I can easily do it for you later.

I think this edit is incorrect - Esri World Imagery (Clarity) Beta shows that there is (or at least was) a traffic island here. The carriageway on either side of the island should therefore be mapped as separate ways, and this is how it was previously mapped. I therefore think this changeset should be reverted, unless the traffic island no longer exists.

Also, I suspect there is a missing turn restriction in the mapping here - do you know (from a source we are permitted to use here, so e.g. Google StreetView is not permitted) whether there are signs indicating that a U-turn from the A1 northbound to southbound is banned? The geometry I see in the imagery suggests that is likely to be the case.

(There is on exception I'm aware of to the general rule of using separate ways around a traffic island, which is for a small traffic island away from any junctions and which can be represented instead by a labelled node. However, that approach isn't possible here.)

In this changeset you've managed to move the way previously representing the farm access track to the removed highway alignment, and created a new way for the farm access track (I'd guess this happened by extending the way and then splitting it, with the history being assigned to the wrong half). Please try to avoid doing this, as it makes the history more confusing and misleading.

I've done something slightly different - changeset/76563347 fixes assorted issues, some of which were introduced here, and some of which were already present. Have a look at it and see if you agree and/or think I've missed anything.

I think you have misunderstood the meaning of many of the construction tags you have removed. They are being used to indicate what tags will be appropriate when all the construction at the junction is complete and the new roads are fully opened. Your edit removes all of these indications. I would agree that a few of the existing tags were incorrect, but a blanket removal of constuction: tags on all the roads that are open in some form (not necessarily their final form) seems incorrect.

I've reverted this in changeset/76562296; if it turns out that the roundabouts are fully open (which would surprise me), then this can be reinstated as a new changeset using appropriate refs for the temporary layout.

That GPS trace drawn in the image corresponds to the layout as it was mapped before your changes, and hence provides no evidence at all to support your changeset. If that is the only 'evidence', then this changeset should be reverted.

Do you have a source for this? I haven't yet seen any evidence that the roundabouts are fully open (though I am unable to check in person), and wasn't expecting that to happen while all A14 traffic was still passing through them.
Also, I don't think it's correct to label bits of this junction as the A141 yet - I wouldn't consider any of this as the A141 until the A14 is diverted onto the new bypass.

I don't see any reason for including the tag "oneway=no" - it's the default for this type of road, and there's no reason why any other mapper might think it weren't twoway.

I think the wiki page, along with existing usage, makes it fairly clear that "motorway_junction" is appropriate here.

You made various errors in this changeset, which I think I fixed in 76255908. Firstly, you used "trunk" instead of "trunk_link" in many places. Secondly there were a number of gaps where you left the old motorway status or refs. And thirdly, you changed the future status of a number of slip roads to incorrectly show them as forming part of the final layout.
You may like to check that all my changes are accurate and have fixed all the problems with your changeset.

The county council website suggests that the name changes at the corner of the parish boundary near Crome Lea Business Park. I've no idea if this is an admissible source anyway.
https://www.cambridgeshire.gov.uk/residents/travel-roads-and-parking/roads-and-pathways/highway-records/