given your recent change in username but most importantly, the changeset comments in the past hours you have made, i will cautiously welcome you to the club of productive and correct mappers.

of course, i will be reviewing these, but i hope to be able to offer constructive input to improve the map, rathr than getting involved in meaninggless edit wars.

then maybe like with so many others, i can lose my interest in stalking you personally, and return to stalking amazon logistics.

map well, and be well.

This changeset has been reverted fully or in part by changeset/73003141 where the changeset comment is: revert persistent pokemon vandalism. this time it's personal. dwg has been notified.

This changeset has been reverted fully or in part by changeset/73003129 where the changeset comment is: revert persistent pokemon vandalism. this time it's personal. dwg has been notified.

This changeset has been reverted fully or in part by changeset/73003100 where the changeset comment is: revert persistent pokemon vandalism. this time it's personal. dwg has been notified.

This changeset has been reverted fully or in part by changeset/73003064 where the changeset comment is: revert persistent pokemon vandalism. this time it's personal. dwg has been notified.

This changeset has been reverted fully or in part by changeset/73003034 where the changeset comment is: revert persistent pokemon vandalism. this time it's personal. dwg has been notified.

This changeset has been reverted fully or in part by changeset/73002968 where the changeset comment is: revert persistent pokemon vandalism. this time it's personal. dwg has been notified.

This changeset has been reverted fully or in part by changeset/73002905 where the changeset comment is: revert persistent pokemon vandalism. this time it's personal. dwg has been notified.

This changeset has been reverted fully or in part by changeset/72999658 where the changeset comment is: restore highway=footpath tagging

This changeset has been reverted fully or in part by changeset/72999607 where the changeset comment is: restore carpark service drives turned into walking paths

This changeset has been reverted fully or in part by changeset/72999363 where the changeset comment is: these are service drives. not paths. stop trying to turn everything on the map that was correctly tagged, into paths. osm is a serious project used for navigation and delivery, and false data like this makes that impossible and harms osm.

This changeset has been reverted fully or in part by changeset/72998987 where the changeset comment is: what is this rubbish? this is vandalism pure and simple and has no place in osm. https://nrenner.github.io/achavi/?changeset=72856322

yeah, that looks good enough to be almost real that i'll let it pass. the esri aerials seem ideal in this area for tracing - oh, connecticut, of course - so i resisted the temptation to repeat my earlier mistakes adding copious details and striving for improvement.

thanks

while one node was included looping back over itself twice. now renders.

you have listed google maps as a source.

you may not copy any information from google maps for osm or source from google.

by the way, don't you invoke your editor with a middle-button click in a new tab? i'm into that habit because of how painfully slow reloading and redrawing any page is so by now i have something like a hundred tabs to sort through with enough works-in-progress that i almost hope for a crash or disconnect to give me a clean slate. then remember to enable the notes layer and login before setting off again.

in general i think i've cleaned most of the dupe nodes outside california, to get a better handle on their frequency. same with self-intersections that are starting to recur in ways i've not really been able to reproduce without effort in potlatch2, that seems to better safeguard against these things in a lightweight manner that would not surprise me if it were more efficient than the bulk of editing i see in many amazon changesets. or i could be jaded by the beginner-friendliness, animations, and pop-ups iD offers when i can't refresh my screen in under a third of a second at best earlier, now actually worse.

determining the source of the bugs is probably not something i can do the way i was 40 or more years ago with FORTRAN i've long forgotten. i'd probably cynically grep the source for rnd() or rand() or however the language in use i'd have to learn describes once-in-a-blue-edit functions. or maybe search the code for
/* now let's annoy those snivelling Potlatch potheads by dumping bogus data for them to fix */
but i'd abandon my criticism upon finding such an obvious and impossible smoking gun.

the best i can do is point out operator error so it doesn't flood the map at the rate amazon are contributing, and comment on the unusual cases like where six nodes get piled up in the hope someone who already has a grasp of the code, where to find it, and what the design intention is, will question that intention. i've too often seen little documented code divert in reality from the intended flow, and i'm no longer maintaining any k0dez i know intimately where reproducible bugs can be tracked by me. plus i can't even start iD now, not that i could reproduce these errors.

i suspect that amazon simply applies a few cosmetic patches (host is obvious) but doesn't do much to the innards, although i could see stripping out a bunch of unused functionality like brand name links and such. luckily they do not seem to be doing showstopper edits like other organised teams (über/lyft/grab/can't-remember) making mincemeat of relations on the west coast in ways like i've cleaned after you somewhere, but worse, as it was regularly damaging map data every handful of edits.

it's just a matter of getting the fix(es) into iD in the first place.

hi,
looks like in this edit you added way/703230884 , joining it at several locations to other features, and then moving it out of position, dragging the nearby paths and a pond out of shape.

could you look at this? thanks

hi amazon,

same issue, stepping through the nodes starting from the highway shows the terminating node is not part of the sequence heading in.

revealed by https://tools.geofabrik.de/osmi/?view=geometry&lon=-98.27035&lat=29.41548&zoom=18&opacity=0.66&overlays=self_intersection_ways,self_intersection_points,single_node_in_way,duplicate_node_in_way

hallo amazon,

please check the termination of the loop of way/709883992 for it is reported as a self-intersecting way.

thanks

This changeset has been reverted fully or in part by changeset/72980610 where the changeset comment is: this is a cemetery, not a park. these are not paths, these are clearly roads and were correctly mapped earlier.