I would like to add something to Clay’s general comments, with a DWG hat on:
Had the DWG received complaints about this import - which to my knowledge we have not - then we would not judge it by “how many improvements has this brought” but by “how many things have deteriorated”.
A frequently heard phrase about imports - also in this discussion - is “well yeah it may have broken some things but on the whole, the map is now better because we have added vastly more POIs”.
But an import that breaks even a few existing things and wants to compensate with adding more new things is making the map better for users at the expense of mappers. And we tend to treat mappers as a valuable (and limited) resource.
We would always expect imports to take great care to not damage existing data, even if they bring in vast volumes of new data. An individual POI might not be worth much to OSM, but we don’t want to lose a good mapper over an import because they feel their contributions have been ridden roughshod over.
But, as I said, this is just a general note that would apply to all imports, we haven’t (to my knowledge) had a complaint about this import and I have not looked into @SimonPoole’s claims of deterioration.
Can we please split this discussion into “concerns about general imports” and keep “concerns about THIS import” going on this discussion? I’m all for addressing concerns brought about by this import, but general comments are starting to bury the specific ones.
I think you completely misunderstood Clay’s comments or are purposely hijacking them to make another point.
This is where DWG goes wrong. In DWGs non transparent process and acting as judge, jury and executioner they fail to see any benefits of improving data.
They need to understand the tags and their relative usefulness and compare it to what is there before.
Having a complaint from someone the other side of the world who has never edited there or edited these type of elements in OSM, should have more weight than the overall benefit that an import and automated update can bring.
Like I said earlier these were often before comparable to shop=yes
Compare this to the transparent discussion here. The questions marks and failing surrounding the import were openly discussed by people that support this import.
Without these transparent and open discussions about the imports what is the alternative? We had them all hidden away as private messages discussing it? the wider public and DWG would not know any better.
The shortcoming were highlighted here for all to see and then people with no skin in the game thought that had enough ammo to bitch and complain about it. It wasn’t through their own edits being altered or broken, it was perceive and imaginary harm to them.
You should consider the mappers who are a passionate about and map such things have a larger weight here, what they think overall of the import and if/how any steps are made to fix/improve before you break thing more by reverting them.
That is the shortcoming of DWG.
Hopefully you haven’t been stuck in the Judge, Jury and Executioner mindset and role too long to see how you can grow and improve.
How to improve DWG
talk to people that have knowledge of the type of import, people that under the tags in it, etc
talk to people that done good and successful imports/ mechanical bulk imports/etc before (ideally look to have a board member with this experience because to my knowledge noone there does)
Don’t presume all import bad
this claim is in contrary to “we don’t want to lose a good mapper over an import because they feel their contributions have been ridden roughshod over.”
(also, it is ironic to reply with non-transparency complaint to post from DWG explaining criteria - if you wanted more info you could ask for it rather than preemptively complain)
As with everything there are multiple sides and finding the right balance is not easy. As a mapper I certainly appreciate the DWG thinking about the importance of mappers when making a decision. But as a mapper I can be equally demotivated by stale data and good quality imports/conflations can help here.
Any balance needs to take both of these into account when assessing the impact on mappers. And it should reflect which mappers are affected as that is likely to vary based on what is being imported and in which country/regions.
Lastly, having just worked on a manual import of EV chargers in the UK, I can say that some mappers add these from a position of significant knowledge and do a fantastic job. Other mappers who are less informed on the specifics of EV chargers have added some lower quality data. Having someone look across a whole country worth of data can be beneficial. For something like EV charging stations that does require some forms of automation. Good checks and balances, yes, but some automation is required to make it manageable. That need not be a bad thing.
Ok, this is a bit awkward now because someone has moved my post into a new thread with a more generic headline, but you’re now going back and discussing the concrete case, so technically your question and this reply would again belong in the other thread…
Normally for the DWG to investigate an import, someone would have to formally complain about it. To my knowledge, nobody has formally complained about this import yet. I assume that if Simon had found the matter important enough to raise a complaint, he would have by now. So unless someone else feels this particular import warrants DWG’s attention and files a complaint, it would be unusual for DWG to occupy itself with it.
That is literally not a thing that is disputed, even Rovastar suggested that the import should be reverted.
Obviously it could have been redone once the glaring mistake had been fixed. But to have any mitigating effect that would have had to been done in the immediate aftermath of the import, it is far too late now, and (IMHO) just to do it for the educational effect long afterwards doesn’t make sense considering the audience.
Since I myself was thinking, “Gee, we should do an automated edit to update some of this EV charging infrastructure”, but then knowing the minefield I would be entering, I was exploring the possibilities when the 6/22 import to OSM from supercharge.info happened. Mostly it was good, but there were a few issues. How could those have been avoided? I think by waiting until more people knew about the plan and discussed the possible issues, like the ones I had. I’m not sure how the word would have gotten out, though, since I was doing my thing and others were doing their thing.
For the future, the way to handle this might be to spread the word to all areas and websites known to be involved in the data you’re proposing to bulk edit. It will take longer, but more issues will be foreseen, which then will make the bulk edit as painless as possible.
