Your use of 'primary' in the changeset description is confusing and different to others' use of the word. A roads are primary (or trunk), do you mean the road is NON-trunk - and in the convention used by the OSM community in the UK. And you have removed junction=roundabout.

Has the mini-roundabout that was previously mapped and is shown on imagery been replaced by a roundabout with a central island?

Your actions and changeset comment do not seem to match. The road was primary (see history, was changed from trunk in 2018). You also deleted some junction=roundabout tags, modified the geometry of some junctions and removed some dual carriageway sections.

I see you've not added names for any of these roads or removed the construction=residential tags, or changed the limits of the construction or residential areas - did you survey the site to see what is actually in occupation, or just change them all arbitrarily?

I assume it was an unintended error to delete way/77156922 (residential tower Blok 7 in Sint-Maartensdal, Leuven) in this changeset.
I have undeleted it.

Are you sure the main road route should be _link and the linking routes primary? And can't you go from the B road to the primary without going via a _link?

D'oh! thanks. Looks like I dropped a node to remove kink in the old roman road line and managed to shorten the 2 halves of B road, fixed now.

Hi Andy,
Looks like that construction=rail got set way back at changeset/50120352 when the construction area got its name and operator set. Guess as it's a rail freight interchange the assumption was it would be landuse=rail after construction, but that's not right, there'll just be a strip down the E side for the sidings. I'll get rid of the construction=rail on the main multipolygon.

The so-called Hilton Hotel Lane is an interesting one... I changed it from trunk (A50) to unclassified in changeset/60337918 2 Jul 2018. Then the corporate editor 'Hilton Hotels' set it primary with that name in ch. 64003196 30 Oct 2018, and un-usefully added oneway=no at changeset/64003240. You'll see that changeset has a discussion from me challenging the type + name, and I deleted the name and set tertiary in ch 64059574. Then Hilton Hotels shared the document you'll see in that discussion and reset the name. Hard to argue against, although as expected the hotel hasn't changed its address and there are no road signs. So the name is still descriptive eg mistaken writing of 'Lane' for 'lane' for me. Actually the tertiary is useful as it's pretty wide, white lined and crash barriered to the speed limit change point further north, where it physically steps to a narrower size (actually still with white line) but that's a useful place to switch to unclassified.
cheers,
Clive

The pattern of the editing is clear. This user signed up as an OSMAnd OSM Live recipient (supplying a Bitcoin address to receive small payments in return for editing). Every region the user edits has significant OSM Live subscriptions (and therefore editor payouts) - or at least, had, when the edits were made. The user is making enough changes (OSM Live counts number of changesets) to be 'rank 1' - in *every* single region touched, currently as follows:
(numbers after region abbrev are users' numbers of changesets, r1=rank 1, x of y is current mBTC 'earnings' for user out of total avail for region):

WW 4743 (was 4653 at last recalc on http://www.hdyc.neis-one.org) r1 .67 of 141 total for WW
Russia (tot 808)
Russ-CFDistrict 116 r1 .05 of .40
Nizh.Nov. Obl 637 r1 .20 of .20
Novosib. Obl 55 r1 .07 of .13
UK (total 293)
Eng 202 r1 .11 of 1.00
Eng-EM 202 r1 0 of 0 (was 3 or 4 contribs showing earlier in month when edits were done)
Wales 91 .11 of .20
Qatar (total 273) 273 r1 .07 of .13
Denmark (456)
Denm 456 r1 .12 of .33
Denm-RoS 456 r1 0 of 0 (now but presumably non-zero when the edits were done)
Iceland (187) 187 r1 .08 of .13
Malaysia (195) 195 r1 .13 of .53
Cyprus (156) 156 r1 .07 of .13
Thailand (290) 290 r1 .06 of .13
Poland (294)
Pol 361 [increased] r1 .20 of 2.86
Pol-LowSil 126 r1 .04 of .20
Pol-Sil 235 r1 .11 of .33
Germany (total 247)
Ger-Rh-Pal 180 r1 .15 of .47
Ger-Saxony 126 r1 .11 of .47
Ger-Schl-Hol [incr] 456 r1 .11 of .33
Ger-Thur ... not yet ...
West Bank (59) [Palestinian Territories on OSM Live] 59 r1 .08 of .13
Slovakia (291) 291 r1 .14 of .60
Ireland (177) 177 r1 .11 of .33
Hungary (350) 350 r1 .16 of .93
Romania (130) 130 r1 .14 of .46
Ecuador (72) 72 r1 .13 of .27
US (95)
US-Ca 95 r1 .10 of .60

total of the above approx 3.32 mBTC, or about euro 10. It's really not worth the effort.

Presumably Germany-Thuringia is on the list as a region to be 'hit' soon, it has a nice high amount of support?

To achieve rank 1 in *every* region touched, and *only* to edit regions with OSM Live support makes the situation pretty clear to me - this looks like armchair mass editing for (slight) personal financial gain, not for the benefit of the map, and needs to stop. OSM Live will be contacted so they can review if these payments should be made.

It that a strange enough reaction maus?

對不起，也許是錯誤的語言：

在此編輯中，您移動了倫敦滑鐵盧車站附近的兩個節點。 別擔心，它是固定的。 在編輯時移動地圖時要小心。

享受添加到地圖！

(Google Translate)

In this editing you moved two nodes near Waterloo station in London. Do not worry, it was repaired. Be careful when moving the map while editing.
この編集では、ロンドンのワーテルロー駅の近くに2つのノードを移動しました。心配しないで、修理されました。編集中に地図を移動するときは注意してください。

Thanks for your reply. I had seen your website with the updated directions, but wasn't convinced it meant any more than that there is a new roundabout exit lane for the Hilton Hotel, ie. accidental writing of "Lane" not "lane".
I anticipated a name like Derby Road or London Road (which various parts of the old A6 were, pre-A50), or perhaps something like Lockington Lane. Seems surprising to call the Lockington-M1 J24 road after the hotel, although it is the one obvious feature along it!
I wouldn't set the name without other evidence (actual hotel address change, road sign), but if you're certain you can (re)set it - I've currently left the road as unnamed tertiary, with the north end temporarily oneway as it only serves the entrance until the connection to Main St opens.
regards Clive (Cebderby)

The road had already been marked open two-way, (with the northern part temporarily one way to the entrance). Unless this road has a reference (A- or B- number) it cannot be more than a tertiary road. Also is the name you have added official (part of hotel's new address?) or a description? Suggest the road be made tertiary, including the slip roads on/off the roundabout, with no name unless official.

Thanks for the info, done in changeset/63074143 - fixme remain for exact pylon positions, end station shapes etc

Hi Dave,
Heading E after the A4241 double river bridge / roundabout, arriving at the footbridge with the big cycle ramps, the route 4 seemed to be signed to continue along the A4241 Harbour Way - on the S side at this point. (I crossed to take the Wales Coast Path foot route / old? 4 route around toward the railway station). Looking at the Sustrans site, NCN 4 looks to stay on the A4241 until a (currently unmapped) connection to a railway underpass toward Central Road / end of Penrhyn Street where the (old) route is. Don't know where the best cycle lanes and crossings are for the A4241, maybe on latest imagery. Actually, the new station bridge at Port Talbot is probably fine to cross pushing a bike + using the lifts. Clive

There won't be any no-U-turn signage in a situation like this because there's really nothing to U-turn around - it's more a less a normal 2 lane road with a narrow divide to prevent all right turns and aid pedestrian crossing. The ends of the one ways can have ahead-only restrictions as that is the only safe and intended flow.

Why remove junction=roundabout from way/35955222
and from parts of roundabout (already fixed) way/95588402 and adjacent?

Thanks for the reply, seems that J12 is interchangeably referred to as Theale or Calcot, and J11 doesn't really have a name, so I'll drop the current names from the J11 exits.

What was the source for these junction names? I note particularly you have set M4 J11 as 'Calcot Interchange', when it is J12 that serves Theale and Calcot (another editor has since set 'Theale Interchange' on J12).

It's really annoying that man_made:bridge is not rendered on layers along with ways. But I understand the idea is that it should have the layer of the crossing way(s), rather than being part of the landuse, and ways underneath can pass uninterrupted. osm.wiki/Proposed_features/man_made%3Dbridge