Is this something that could be pushed to maproulette? Not as reversions, but tasks to validate or update OSM entries that match a pattern - eg edited by this user and now has bicycle=no, highway=footway etc?
On Wed, 13 Oct 2021 at 16:56, Andrew Harvey <andrew.harv...@gmail.com> wrote: > I guess there would be nearly 0% chance that you would be able to cleanly > revert without dealing with conflicts. It can get complicated when > conflicts are detected by the JOSM reverter, you need to both know about > the OSM data model well (nodes, ways, relations, tags), know about the data > you're reverting and an understanding of the area you're working in so you > can decide how to handle the conflict and what final state you like. > > Tony since you know the data and area well, did you want to work on this > together with me? I can help out with any technical roadblocks, maybe on a > screen share? > > On Wed, 13 Oct 2021 at 16:32, <osm.talk...@thorsten.engler.id.au> wrote: > >> While I'm normally all for "you made the mess, you clean it up", this >> might >> be something better tackled by someone with extensive experience in >> reverting multiple changesets? >> >> Have we got any experts in that? >> >> -----Original Message----- >> From: stevea <stevea...@softworkers.com> >> Sent: Wednesday, 13 October 2021 14:13 >> To: fors...@ozonline.com.au >> Cc: OpenStreetMap-AU Mailing List <talk-au@openstreetmap.org> >> Subject: Re: [talk-au] Cycling on Victorian paths >> >> I did my best to help Sebastian, but near the point where we got the first >> launch of JOSM (he DID install Java, he DID have to move the .jar file to >> his Applications folder, he apparently was NOT using a capital A in >> Applications...) he suddenly went "radio silent" on me and didn't answer >> any >> more email ping-pongs. >> >> I had all primed my next email how to install a reverter, but didn't send >> that because it seems he remained in a low gear, and running a JOSM >> reverter >> is for those who are, um, "in a higher gear." >> >> Good luck getting your data in shape, there, mates. >> >> SteveA >> (where it is getting to be bedtime Tuesday night) >> >> > On Oct 12, 2021, at 9:06 PM, fors...@ozonline.com.au wrote: >> > >> > Adam >> > >> >> Spotting these >> >> and knowing how far back to revert to might be tricky I guess? >> >> eg https://www.openstreetmap.org/way/47771844/history >> > >> > Yes. I have never been involved in a reversion so complex and it worries >> me too. I presume they should be reverted in reverse date order, ie most >> recent first. And acting in a timely manner is important, before others do >> edits on the same objects. >> > >> > Taking your example, the first reversion is important and the following >> two swapping between path and footway make little difference. >> > >> > Tony >> > >> > >> > >> > _______________________________________________ >> > Talk-au mailing list >> > Talk-au@openstreetmap.org >> > https://lists.openstreetmap.org/listinfo/talk-au >> >> >> _______________________________________________ >> Talk-au mailing list >> Talk-au@openstreetmap.org >> https://lists.openstreetmap.org/listinfo/talk-au >> >> >> >> _______________________________________________ >> Talk-au mailing list >> Talk-au@openstreetmap.org >> https://lists.openstreetmap.org/listinfo/talk-au >> > _______________________________________________ > Talk-au mailing list > Talk-au@openstreetmap.org > https://lists.openstreetmap.org/listinfo/talk-au >
_______________________________________________ Talk-au mailing list Talk-au@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-au