Hi there, what is the purpose of relation/19294540 ?

Do you think it will be better to keep the old nodes, only changing them to entrances, both to "keep the history" and to explicitly state that there are two entrances?

Are the old paths really demolished? Afaik it's only the direction of oneway that has been changed. Would it be better to simply replace the geometry to "keep the history"?

Would also suggest cleaning up https://overpass-turbo.eu/s/2crR

Is this actually a name, instead of, say, description?

Is the peak really there? Out of copyright maps seem to say otherwise.

Afaik, `name` is the current commonly used name. Being commonly used has nothing to do with using standard romanisation. As long as "Dit Sei Kau" is the popular English name, even if it is caused by an error or a typo, it should still be used.

Languages (including vocabularies and proper nouns) can (and do) change, and in this case your edit in 2019 was influential enough to modify the actual name. If you wanted to revert it to "Tit", you should have done so before "Dit" became the popular name. But now it's too late, and the change should be respected.

"I believe many online reference simply copied the name from OSM without reasoning." Then it still doesn't explain why no one used "Tit Sei Kau".

Since all current uses are "Dit Sei Kau", one of four things is true:

1. This peak has no English name, and "Dit Sei Kau" is not widely accepted. In that case, `name:en` is not necessary.
2. This peak has an English name of "Tit Sei Kau". If this is the case, please give some evidence of its uses (from someone other than you), since I can't find any on the internet.
3. This peak has an English name of "Dit Sei Kau", both before and after your edit in 2019. Then, this changeset is wrong.
4. This peak has an English name of "Tit Sei Kau" before 2019, but your edit changed the general public's consensus and made people use "Dit Sei Kau" instead. Then, use the new name, as `name` is meant for the commonly used name, not the "correct" or "original" name. Remember that words are an artificial construction, and all names were made up.

Yes, I know that the government romanisation uses "Tit" for 跌. The question is whether the commonly used English name is the same as the government romanisation. Based on internet searches, almost all instances use "Dit" and not "Tit". The only use of "Tit Sei Kau" I can find is https://www.flickr.com/photos/minghong/albums/72157622473602534/. (and that post was made by "minghong"... is that you?)

Are you sure? Names don't necessarily follow the government romanisation, and based on online information, Dit Sei Kau seems more common than Tit Sei Kau.

Please be careful when panning; there are multiple dragged points. Also, why is way/296724850 deleted?

It appears that this changeset re-adds vandalism?

Pretty sure it's "TKO tunnel portal" as in "the tunnel portal on the TKO side"? Hence "tunnel portal" is not capitalised.

Isn't it at least a local name or a nickname? It's in common use afaik.

Is it really a general store?

Are you sure it's still under construction, when a note (note/2327347) says it started almost 5 years ago? Aerial imagery doesn't seem to show construction, though it's not very clear.

Thank you for your reply. Apologies if my intention to write a comment was not clear - I didn't mean to say I somehow know better and could confidently improve this changeset, nor did I mean to discredit your intention and effort in attempting to improve geometries, both for this changeset and other calibration works you have done - I simply wanted to point out a few things that I assume you overlooked while working on this. Sorry if that was not clear in my initial comment.

Regarding aerial imagery, I am aware that ESRI imagery is preferred. When I used the iD editor to inspect this area, only the ESRI imagery seems to supports the new geometries, while Bing, "ESRI Clarity Beta", and Mapbox all show a much straighter tunnel approach. This led me to think that the new geometries were due to ESRI's or the altitudes's distortion, hence points 1 and 2.

Regarding barriers for editing, I honestly did not have a clear "other tools" in mind when I said "I would suggest using other tools to verify the alignment". I just thought "mappers more experienced than me might have better ideas of which mapping techniques to use in a specific situation, so maybe they can try alternative tools here to clear things up!", which is why I wrote that point.

Regarding absolute accuracy and balancing, while I understand that it's impossible to have pixel-perfect alignment with the real world and that compromises have to be made, the problems I pointed out in point 4 are just too obvious imo; it goes beyond what I personally consider to be "reasonable" curves. It seems like the west-bound tube is bent south suddenly at the end of it, just so that it fits the aerial imagery without needing to modify the other parts of the tunnel? No insult intended, but it feels like a very lazy and minimal workaround to avoid dealing with the bigger question of the accuracy of the geometries of the tubes and the eastern portal. At the very least, the transition from the existing data that was not modified to the new data should be somewhat believable, unlike that current mapping, which is blatant enough that an unexperienced user like me noticed something's wrong just by taking a glance.

Now again, I'm not saying that I can do better or improve on this; I just want to point out things you probably didn't consider while working on this so that you can make a better judgement on what to do with this changeset.

Are you sure the tunnel geometry changes are correct?

1. Is aerial imagery the only source? This is on a hillside, so the accuracy wouldn't be great.
2. Other aerial imageries seem to disagree with the new geometry.
3. From my east-bound bus-taking experiences for TKO tunnel, exiting the bus interchange and into the tunnel is almost straight, while other vehicles have to bend slightly left to align with the tunnel approach, before turning slightly right again immediately after entering the tunnel, which seems to correlate with the old geometry better.
4. Under the new geometry, the west-bound tunnel turns slight left before exiting, then turns right again. I find this road design rather unbelievable, and while the former western tunnel approach has seen a lot of works after the demolition of the toll plaza, it can't possibly change the tunnel alignment, right? (Is it even possible to move a tunnel? If not, then the tunnel geometry should be the exact same as the original plan in the 1980s, and I don't think a sensible planner would make such curves.)

As such, I highly doubt the accuracy of the "calibration" and "beautification" of the path shapes here, and I would suggest using other tools to verify the alignment. See also another reverted dubious hilly calibration: changeset/168519808

Are you sure the sources you cited can be used in terms of copyright?

Is `contact:name=` supposed to be `contract:name=`? Typo?