OpenStreetMap logo OpenStreetMap

Changeset When Comment
55787946 almost 8 years ago

This changeset has been reverted fully or in part by changeset/55820191 where the changeset comment is: user requested revert of aligning ways to poor Bing aerials, rather than aligning background image to ways

55819724 almost 8 years ago

This changeset has been reverted fully or in part by changeset/55819893 where the changeset comment is: revert unauthorised use of Pokemon in changeset description

55795574 almost 8 years ago

Hello Katie -
This changeset description caught my eye when you mention moving roads.
And I confirmed, you are using bad Bing imagery which shows very poor alignment and correction.
Maybe you would be interested in an almost-identical discussion I am now having with changeset/55787946 ?
I can see there is far better aerial imagery for your area both from Mapbox and DigitalGlobe, as one area where you have had to severely move the roads looks to be rather hilly, where Mapbox is photographed from almost straight overhead.
Just an idea to try to keep you from doing too much unneeded or incorrect work...
Thanks!

55787946 almost 8 years ago

Hi, sorry for not replying sooner...
The easiest way to have a revert happen will probably be if someone takes pity and offers to do it Just This Once...
.
Before you do any more mapping, and while awaiting a revert, can I ask you to become familiar with the different background imagery -- including that which was used for the original tracing?
If I understand, there should be either the command `b' in editor iD or an icon like layered papers referring to Background imagery layers.
The default is Bing; others to note are ESRI, two DigitalGlobe offerings, Mapbox, and USGS Large Scale imagery to cover the US.
Don't actually move any sidewalks as that could make a revert difficult -- look, don't touch; play, don't save.
.
But, I'll point you to this forum discussion:
https://forum.openstreetmap.org/viewtopic.php?id=60988
In it is a URL that you can use as a learning tool, which gives you access to the old imagery (ESRI Clarity) that is clearer, even if well older, than anything for your area.
I've added it to my Potlatch editor to verify it brings up what I expect.
And it does! Yay!

Don't use it for tracing until the legal permission is clarified; but then, almost everything near you was already traced from it.
Note how it perfectly matches those sidewalks you have not yet moved. I think I measured over 3 metres offset between this and the default imagery.
Also, get a feel for how bad each of the other imageries are compared to this, even if they are newer.
.
And if you can discover how you can adjust the background images in your editor to line up, such as on the dual-Y diamond shape on the wavy walk, that would be great -- for me with Potlatch, it is holding down the Space bar and clicking and dragging the background in the editor, but I'm sure it's not that for anything else.
Notice how perfectly the waterbody outlines match this background too.
I'll try to explain details about each imagery and zoomlevel for your area, if it's not too much writing and you want to learn ;-)
However, I think now lack of sleep is taking its toll, as I was unable to carry out a conversation a short while ago, despite thinking I was awake, so I'll try to use this kezboard as a pillow for some time.
More later... (words of impending DOOM)

55792906 almost 8 years ago

Hi Glassman,
I believe I can see two aerials, at least one guyed, to the southwest of the GNIS import node.
This is in the Bing z20 aerial imagery that is getting somewhat long in the tooth now: one is prettu much 39,9m southwest with a small fence surrouning; the primary is 60,5m west-south-west away from the GNIS node and just outside the now-road-loop outline with what looks like a grey C-Band parabol and a black shed.
This can also be poorly made out in even older z19 ESRI Clarity old-Bing images, but in Mapbox the whole area has been cleared out, the others unclear.
With mediumwave aerial arrays falling around me like those trees the start of the year, It's probably no longer a station in service.
Ooops, Auntie Google disputes. Well, I know what my OCOSMD is going to have in store for a while...

55787946 almost 8 years ago

Howdy, PugnaciousPoultry...
I have again taken a much closer look at the achavi URL, and I *believe* this particular changeset is almost nothing more than shifting the geometry.
That is, my cast-off LED flatscreen doesn't clearly distinguish green from yellow, but I don't think I see any newly-added sidewalks.
The red->green is where you moved things; a green-only without your name on the data seems to mean it was connected to something you changed, but is not new.
.
That is, I believe it is easiest to revert every change made in this changeset to restore how things used to be, and start anew.
(There are probably sidewalk-mapping tips you should be getting as well, but I'm not looking closely at that yet.)
Then you can spend some time to familiarise yourself with the different (lousy) aerial imageries available in your area. And if you want to get really depressed, visit other areas with sharp and detailed ESRI Community World Imagery, such as the entire state of Connecticut, select parts of California, Florida, Basel (Switzerland), and the zoomlevel 20 Bing imagery showing its age around major USian metro regions.
If I do see anything go missing completely from a revert, I might be able to restore it. But if you've added buildings, they were not done here and will not be touched.
I think that would be easier than moving things back (assuming you agree about the positioning) with as much work you put into being sidetracked earlier.
.
That is, I think this is good news. The Bad News is that like you, I'm still a beginner and have not carried out a full changeset revert, preferring to keep my hands clean, so someone will have to volunteer to help you with that.
I'm glad my first reply did not quite trigger a tl;dr reaction, but I'm sure everything I write can be given that treatment by someone better skilled with fewer words.

55787946 almost 8 years ago

Hi,
I saw that your updates to sidewalks included shifting the position of several in https://overpass-api.de/achavi/?changeset=55787946 .
That is often the case of poorly-aligned background imagery. So I thought I would take a look, as I know the available aerial imageries in the US have gotten much worse.
This is what I saw, as you used the newer poorer Bing imagery to shift, most strikingly, the small circular path west of Crowley Parkway, but not the wavy path directly east of same, which should align to older aerial imagery.
This is in fact what I see; Bing and other imagery providers match well your changes, but not the original data.
This imagery is known to be often poorly adjusted with perhaps several metres offset.
For aligning things, if a sharper imagery is not available, I like to use the USGS Large Scale imagery, and indeed, here it matches well with the wavy line, but north of where you shifted part of it, and probably those ways you have not shifted.
It is a pity there is not better imagery in your area, and that you spent your time moving things from a more correct position to one that is less accurate, rather than repositioning the background instead.
If you align the imagery to the original wavy path, you will then see that the roads line up much better.
I hope you understand what I am trying to say.
Unfortunately I cannot tell you exactly how to compare and shift the background imageries in your particular editor, but maybe someone else can be a better help with this.
Sorry for all your extra work!

55789400 almost 8 years ago

Hallo,
In this changeset you have converted an imported detailed apartment building into a hospital, another into a school, and you have surrounded the entire area as school grounds.
Again your changeset comment is simply `tags' but it is clear from the aerial imagery that these changes are not a simple mistake, but fantasy, false information or vandalism.
Can you provide evidence to back up your changes?
OpenStreetMap is not a playground and incorrect information and vandalism is not tolerated here.

55789412 almost 8 years ago

Hallo,
In this changeset, you appear to have deleted five similar detail-mapped buildings, and modified the geometry of another (for worse).
Your changeset comment is simply `tags' which gives me no clue as to the intent of what you did here, if the recently-imported buildings are being demolished, or if this was a simple error.
I cannot see anything in the Texas Orthophoto aerial imageries to suggest these changes were needed.

55779460 almost 8 years ago

What a mess. Multiple overlapping university areas, the whole grounds covered by a mulotiple fake parks, un-quadrilateralised buildings across Naniloa Road, constantly changing land/water-uses.
OpenStreetMap is not a plaground, nor a store for fictional information and made-up parks.

55774569 almost 8 years ago

Mojn,
Be aware that there are two red flags you are waving, first, OSM is not the place to carry out tests, and also OSM is not a map for Pokemon Go.
Your tests are unwanted. You have deleted a footpath, that is difficult to make out in the poor available imagery, while the criteria for what is to be in OSM is what is factually on-the-ground, not whether your test is successful.
By carrying out what you feel is a test in the name of Pokémon Go, you are almost guaranteed of having your changes reverted promptly, well before the game dataset gets updated, when it is already using half-year-old data.

55691311 almost 8 years ago

Hi Glassman,
Thanks for reverting these SEO hotel deletions.
I see here that there appears to be a retail outlet, Mattress (aagh, a Monty Python sketch) Superstore.
Would this be a general retail/shared building so that the Best Western Plus name= I see here does not apply to the entire building?
I see cowdog has left the spammy/malformed node intact here as well, just thought I'd bring it up as I seek out this SEO spammer's damage.
thanks!

55639505 almost 8 years ago

Hi neuhausr,
Refer to the initial changeset, changeset/41861684 .
There you see the particular edit was questioned by Georg but no reply was given, in over a year.
If this were to have been a legitimate business playing by OSM rules, they would have been interested enough to actually participate by replying, and would have been given tips on how to improve their edit.
There are other signatures that map-abusers leave behind, that a Real User may or may not make the same mistakes, that I also see in this particular original edit.
Hope this allays your worries about trigger-happy trashing of true truth, at least a bit.

55709659 almost 8 years ago

not to mention misuse of #hashtags, that themselves are #abuse to my aged eyes

55327940 almost 8 years ago

mornin'
I think you cautioned another mapper about his suggestions a while back.
He appears to be limiting the notes now to annoying notifications of genuine issues, sometimes difficult to decipher, sometimes out of OSM control.
But there are still lingering older suggestions I'm challenging.
I'm curious what gmaps has here, and if any street-level imagery is legally available... (but that means work I'd have to do)
I won't revert, as I'm positive there are lots of other gmaps easter eggs or errors that are imported, and I have no evidence so far for a revert.
Plus I've been up through the night and I almost think I can see the sky today...
Interestingly, the notes abuser in Freiburg/Br. just got a suspension, and has reacted not dissimilar to how ex-kimbo reacted upon finding florinbadita_telenav's Kim question, so I wonder if it's something in the air pushing both over the edge.
me, I'm that way all the time.
have a happy day adding pokeponds, pokeparks, and whatever else the increasingly demanding pokemappers-by-notes ask for today ;-)

54652384 almost 8 years ago

ciao constable,
the removed driveway is only visible as an access in ESRI to the newer construction southwest.
it's a well-worn access path rather than a paved drive.
the building it accessed is only visible there and in leaf-on Bing == DG Std, the others are too old.
As I pulled up this, it defaulted to TIGER 2017 which includes an unmapped yet clearly visible Old Fair Grounds extending to this new building but not much further, which I had not seen while tracing.
The pickup in ESRI looks to be part of construction.
So it may be that removed drive was only temporarily in use, with later access under the cover of the leaves, the grass partly healed in the newer imagery.
I may add these and other things I see now, but first, sleep. sometime.

55327940 almost 8 years ago

ciao constable,
are you sure about this?
TIGER 2017 still has the possessive form.
not that i believe it, but i wonder what source the anonymous contributor had and how you verified it.
it's dangerously close to our returned ex-kimberly fan who has admitted to sourcing google maps in the past, and appears to still be doing so around the same areas I'm editing.
grazie!
(no, I've not checked any other sources meself)

54793229 about 8 years ago

Hallo,
I thought I would try to prove you wrong, but I did things in the wrong order.
I regenerated a new Hofer after accidentally deleting the original -- seems I can only revert ways in Potlatch that are deleted, not nodes.
Then I moved it close to its old position.
It was already attached to the building at the wrong position.
Now that I have done things the wrong way, wilmaed, would you care to revert my changesets, put things back to the wrong position, and see if I can try it again and do it correctly for a change?
The changesets are:
#54801308
(move new node and detach)
#54801142
(restore deleted Hofer as a new node)
#54801064
(revert building with attached Hofer to no Hofer)
I should have saved the reverted Hofer which distorted the building, then detached that afterwards, which is what I tried first but it looked wrong. (revert and save in Potlatch1; detach in P2)
danke!
or you can leave things as they are now, losing the history behind the Hofer node, depriving me of the chance to revert it properly...

54773976 about 8 years ago

did I explain this one yet?
way/357488209:
created 30.Jun 2015
last deleted v12
appears tied to several things like the road
full data
way/546860888:
14.Dec 2017 - 19.Dec v8, created by Royoriti
slightly larger, coarser outline
independent nodes
no details

54792081 about 8 years ago

sorry, way/548169812 was reverted in error (rubbish line=school deleted; intent of the changeset was not to touch this way, but to restore a previous amenity=school schoolgrounds extent, to which this user appears to have developed a severe life-threatening allergy, despite having mapped such in the not-too-distant past).
I'm still a n00b trying to figure out this new-fangled editor thingy, so expect many more such mistakes.
Why does osm.org/edit?editor=emacs&changeset=54754118#map=17/33.97552/-81.30413 return an error anyway? Discrimination! Racism! Furrfu.