Much smaller changeset areas, please: osm.wiki/Changeset#Geographical_size_of_changesets

“FT5DR”

What makes for osm.wiki/Good_changeset_comments

source = ?

It's unclear what you've changed. Seems to be slightly moving a single node.

Much smaller changeset areas, please: osm.wiki/Changeset#Geographical_size_of_changesets

“Farewell”

Pity.

I'm sure that other mappers will pick up the slack, though, without the drama & butthurt.

“I'm trying to be helpful and contribute my time to populate OSM data”

This is a community project, with many volunteers, who are also trying to just help by donating their time. It's not merely a data project.

Enormous bounding boxes aren't helpful; they pose many problems for others. It wastes their time and distracts from (or hinders) productive work.

“if I'm going to be chased by the OSM police from here on because of a few bad StreetComplete-generated changesets, I'm probably just going to stop contributing altogether.”

If you regard other mappers asking you to contribute constructively & non-disruptively as some kind of OSM police …

Back in reality, you're spamming the changeset lists of French & British mappers with your changesets that are irrelevant to their local area.

So, if you don't want people objecting, then stop pestering them.

More generally, if you're unwilling to participate in a community, and be considerate of other mappers, then if you feel that ceasing is the best resolution, go for it.

Compare driving an unroadworthy car on public roads, and mechanical failure causing a collision. What do you think the investigating police officers would say if you claimed that it wasn't your fault because you didn't steer the car into the victims, the car did so because of said failure from neglecting maintenance?

From someone older & wiser: http://linuxmafia.com/~rick/faq/crybaby.html#allowed “I'm not responsible for the damage: If I wasn't supposed to do that, I shouldn't have been allowed.”

“People who think this way, Come The Day, will be strait-jacketed and deposited in rubber-walled rooms, the environment best suited to their leadership potential. Permanently. Meanwhile, by raising this traditional excuse of losers and screw-ups, you've demonstrated that whatever small trust (to exercise reasonable caution) was recently bestowed on you was a grave error that shouldn't be repeated.”

A few other entries from the same section also seem relevant & likely applicable.

Replied, there.

“This time I'm innocent”

It's your account, your choice & use of editor, your changes.

Focus on fixing the problem, rather than shifting blame.

“I was using StreetComplete the way god (?) intended, but it decided to upload edits from days ago, though I had connectivity all the while and opened the app a few times in between these changes. Please take the matter with StreetComplete developers.”

If you've really discovered a bug, then you're the one best able to describe it to said developers.

In the meantime, then refrain from using an editor in such a way as to exhibit the buggy behaviour (which affects other mappers; this is a community project, after all). Resume after the problem is fixed.

Have you checked that the configuration is suitable? The problem will likely recur if you have auto-upload enabled at all. Disable it and tell it to manually upload between areas and/or sessions.

Questions about properly using editors should be directed to their respective user-groups (or developers, if misbehaviour is suspected).

Take some responsibility for helping to improve the OSM software environment.

Being blasé toward other mappers (again; community project) is anti-social. Persistence is disruptive.

Smaller changeset bounding boxes, please: osm.wiki/Changeset#Geographical_size_of_changesets

Smaller changeset bounding boxes, please: osm.wiki/Changeset#Geographical_size_of_changesets

▪︎osm.wiki/Changeset#Geographical_size_of_changesets
▪︎what makes for osm.wiki/Good_changeset_comments

Smaller changeset areas, please: osm.wiki/Changeset#Geographical_size_of_changesets

Smaller changeset areas, please: osm.wiki/Changeset#Geographical_size_of_changesets

Much smaller changeset areas, please: osm.wiki/Changeset#Geographical_size_of_changesets

What makes for osm.wiki/Good_changeset_comments

That's an enormous bus stop.

osm.wiki/Changeset#Geographical_size_of_changesets

What makes for osm.wiki/Good_changeset_comments

“There's really 10 nodes.”

way/992953204 says 11 nodes.

Unless the outline overlaps, and reuses 2 nodes twice.

🙂👍

“The "issue" with the current way of work of iD is that anybody can write anything in the box and it will pull the closest match from taginfo. If there is no "tires" in taginfo, it won't suggest it. I'm "taking the firehose out of the house" by removing the service:vehicle:tires tag.”

You're still thinking about it narrowly in terms of data, not process or user-experience.

You say that with iD anyone can type any value (rather than proper presets with pre-existing (checkboxes, radio buttons, drop-down list) suggested values). So, what's to stop someone typing tires and assuming that it was simply missing from taginfo (if they even know that's where the matching values come from)?

How do they know what to begin typing without any (pre-typing, which is essentially guesswork) hints?

Ultimately, what's to stop a repeat? That may well be how it started.

So, no, the firehose hasn't been removed, and while you may have done only one mopping, that's only so far (and by you alone). I don't see that the (root) causal mechanism has been remedied.

“As you can see in taginfo it's not being used anymore and there were no other new uses. Note that I didn't have to do any additional clean up after the first "mopping".”

So far, perhaps; otherwise no, I see no such thing (with confidence, long-term). Your earlier description of iD's behaviour remains, and at least one vector hasn't been mitigated. Which was my original point when mentioning presets.

You want the conclusion to be true, but saying that it's true doesn't make it so.

Not really any different to a politician saying that there'll be zero murder in future because he's prevented it by making it illegal. Right; nice intention, but how does that actually prevent murder (and an exasperated ‘because I changed some legislation’) doesn't really convince or add anything.
The faulty assumption is that illegal==impossible.

“Developers usually don't have to do that sort of pattern matching since they could just query for all features with some tags. In case they only request one of the tags, they will notice that the data suddenly disappeared and they will (or at least they should) take action to fix their software.”

This misses the point, makes many assumptions, and ignores one of the conditionals I specified.

Well-implemented software will still need pattern-matching to query tags (for any reason, be it to find features, or determine detailed properties of each). You're assuming that this somehow won't break, yet querying other tags will. That's not internally-consistent. Again with the presuppositions and pole-vaulting to conclusions.

You assume that developers will be monitoring each instance of software they release. Maybe if a home-brew script running on their own hardware, but not something released for others to use.

I preempted the scenario which you describe: savvy developers will want to avoid & prevent their parser being fragile in the first place. Thus, they'll use pattern-matching (defensive design, not trusting (sanitising) input, being tolerant of errors, wanting robust software). Why wait until it breaks, out in the wild, before accounting for an obvious failure-mode? Just use pattern-matching from the start.

Imagine if that attitude was applied to security and/or safety systems.

“Assuming US English is what you meant by "Yankish".”

I meant what I said. English is spoken in England. What's spoken in the collection of states between Mexico & Canada is something else (for so many reasons).

An unambiguous term for the Americans between Canada & Mexico is Yanks. Thus Yanks speak Yankish.
You're independent, after all.

😄

Sigh. Fine. Smaller BOUNDING BOXES, please.

Your responses don't really address my original point.

Edit the buildings, whatever; just use separate changesets when spanning such huge distances so that each bounding box is small, rather than a single one which spans half the northern hemisphere.

How did a building with 10 corners end up with 11 nodes?

Unless the building really is to irregular shape, you should square it (press Q in that editor).