Re: [Talk-us] TIGER-completeness visualizer?

2019-12-29 Thread stevea
A reminder on this topic:  https://wiki.osm.org/wiki/TIGER_Edited_Map has a 
list (in Basic use) and even a graphic (at the top) of the "key / rules" that 
were used in the ITOWorld TIGER-completion map.

There are links to "the best of what remains" to further the endeavor of TIGER 
cleanup, including some very good Overpass Turbo queries, what I myself have 
found a useful alternative since the demise of ITOWorld's excellent 
visualization product.  That page links to Minh's excellent 
https://wiki.osm.org/wiki/TIGER_fixup/Overpass_queries, which while it states 
these remain "unreviewed," offers at least four such queries which I hope are 
helpful to all who endeavor to clean up TIGER data.  I have used at least some 
of these queries, they do work and they are helpful.

While I would enthusiastically welcome a re-creation (or reasonable facsimile) 
of the ITOWorld visualizer (so pretty!  so effective!), we can (and must!) 
continue to "muddle through" with the current tools available to us for TIGER 
cleanup.  At least now we have a reference for how the ITOWorld visualizer was 
implemented, that (wiki page) gives me some hope.

Can we reduce "it's still gonna take a few decades to complete TIGER Review?"  
We can!

SteveA
___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us


Re: [Talk-us] TIGER-completeness visualizer?

2019-12-24 Thread stevea
From: "Shawn K. Quinn" 
> I agree that we needed and still need a way to clean up all the
> remaining untouched TIGER data. However, without the TIGER import, the
> same 11-12 years would have been spent surveying and naming the roads
> one by one and tracing them from (sometimes outdated) satellite/aerial
> photos and we would probably still be way behind where we are with the
> TIGER import. If it had been on me to trace my neighborhood from a blank
> slate when I first started mapping, I might well have given up on OSM at
> that point.

Please don't misunderstand me that I wish the TIGER import never happened:  I 
am very glad for these data being in the project.  TIGER has, as you say, 
provided us with over a decade of "something" and all the effort many of us 
have endeavored to "make it better, into the superior structure of roads and 
rail we have now."  I was merely positing that as a MAJOR import, TIGER as a 
whole can teach us much.  Perhaps that requiring dedicated tools for 
post-import cleanup could be a requirement all of us might agree upon should we 
do something like this again.

As long as I'm on the topic, does anybody recall how ITOworld's four colors in 
their map (dark blue, light blue, orange red) of TIGER roads were defined?  
(Yes, there were also grey/black for "not TIGER").  I think dark blue was "no 
tiger_reviewed=no tag" (at all), light blue and orange I can't quite remember, 
but they had that tag, but one or both were something like "touched / modified 
in the last three years" (and something else for the other one), red was "still 
has the tag, hasn't been touched in five(?) years or longer."  The middle two 
were well-chosen, but I can't remember their exact rules.

If we haven't the good tools (in the same form), I'm hoping at least we can 
rebuild them from pieces!

SteveA
___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us


Re: [Talk-us] TIGER-completeness visualizer?

2019-12-23 Thread Shawn K. Quinn
On 12/23/19 18:42, stevea wrote:
> One more thing we might potentially learn from our (OSM-US')
> experience (of the TIGER import of 2007-8), would be to insist upon a
> high bar for such large-scale imports in the future (this was our
> largest, without a doubt):  the proposers of the import must
> "pre-load for the back-end" a renderer that will both display and
> foster such goal-oriented tools to "finish the job with high-quality
> AFTER the import."  Import proposers would be required to author and
> maintain this renderer / server for as long as satisfactory
> QA-completion of the import takes.

I agree that we needed and still need a way to clean up all the
remaining untouched TIGER data. However, without the TIGER import, the
same 11-12 years would have been spent surveying and naming the roads
one by one and tracing them from (sometimes outdated) satellite/aerial
photos and we would probably still be way behind where we are with the
TIGER import. If it had been on me to trace my neighborhood from a blank
slate when I first started mapping, I might well have given up on OSM at
that point.

-- 
Shawn K. Quinn 
http://www.rantroulette.com
http://www.skqrecordquest.com

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us


Re: [Talk-us] TIGER-completeness visualizer?

2019-12-23 Thread stevea
On Dec 22, 2019, at 11:34 AM, Greg Morgan  wrote:
> ... I think the challenge for the U.S. project is that most of these maps 
> have been maintained outside the U.S.  If implemented, I am not sure how long 
> OSMOSE would provide the service.

One more thing we might potentially learn from our (OSM-US') experience (of the 
TIGER import of 2007-8), would be to insist upon a high bar for such 
large-scale imports in the future (this was our largest, without a doubt):  the 
proposers of the import must "pre-load for the back-end" a renderer that will 
both display and foster such goal-oriented tools to "finish the job with 
high-quality AFTER the import."  Import proposers would be required to author 
and maintain this renderer / server for as long as satisfactory QA-completion 
of the import takes.

True, we might not be sure "how long such services would be provided" (nobody's 
good intentions or crystal balls are absolutely perfect), but it certainly is 
worth considering insisting upon something like this for large-scale imports.  
(Should we ever again risk them).

Our Import Guidelines seem a happy medium right now, though for something huge, 
a similarly huge commitment to having the technical chops and providing 
(renderer server) resources to finish the job well seems fair.  We have good 
examples of tools that help us "how," we might agree to require these "if and 
when again."

Still seeking a good TIGER reviewer / visualizer, though I can get some things 
done with OT.

SteveA
California
___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us


Re: [Talk-us] TIGER-completeness visualizer?

2019-12-22 Thread Greg Morgan
On Sat, Dec 21, 2019 at 2:03 PM stevea  wrote:

> As the ITOworld TIGER-completeness visualizer at
>
>
> https://product.itoworld.com/map/162?lon=-121.88=37.04=12=true_sidebar=map_key
>
> is no longer supported, does anybody know of a similar "product" that we
> can use to visualize how well we have reviewed TIGER roads (and rail?) in a
> given area?
>
> I REALLY miss that visualizer!  It showed whether highway=* ways were
> "touched in the last three years," whether the tiger:reviewed=no tag was
> removed and so on.  It was very well done and quite informative.
>
> I suppose a dedicated renderer could be built, that's pretty ambitious,
> though it is a worthy project, imo.
>

History:
* First there was Andy's green/red Tiger fixup map.  That was a great
simple map. The map would go from red to green when you killed the reviewed
tag.
* Mapquest picked up this map for a bit but it died when the group was
decommissioned.
* Itoworld was a nice version that added many more colors and the date last
touched.  The map was damaged when certain tags were no longer saved with
each edit.  These were removed in the background.  I started removing all
the Tiger tags because most of the map started turning black.  You could
tell what was edited or not.  The map almost reverted to Andy's map.
*There was a Battle Grid map. The original was lost but there is a
replacement that you have to run by yourself as far as I can tell.
https://wiki.openstreetmap.org/wiki/TIGER_Battlegrid
https://github.com/iandees/tiger-battlegrid

My problem with Battle Grid map was that the the colors were too light when
you revised Tiger data.  I liked the idea because it reminded me of the Who
Did It Map. Something simple like a grid but use two colors like the
original fixup map but with no fading.  All each tile at zoom 16 needs to
do is change color and tell you how many more ways need to be touched.
Most of the tiles would be red until you touched all the roads.  Then the
tile would go green.  Again my approach is to removed all the Tiger tags
because the street names have been expanded.  I don't do this on major
connecting highways because I am not sure that all the routes have been
defined yet.  Tiger will have some of that route data in the list of tags.


> Bonus points for your best guess at when OSM will eventually complete a
> full TIGER data review.  I'll start:  at the rate we're going now, 2045?
>

April 2050  for nodes
June 2028 for ways
based on Ben Discoe's burn rate calculations.
https://www.openstreetmap.org/user/bdiscoe/diary/44192
https://docs.google.com/spreadsheets/d/1HiC1-ixx30tbwgI27RJt1SvvK80BBRkLOg98Qcb0SD0/edit#gid=1034182050


Ben's spreadsheet mentions all these mappers except for balrog-kun for the
western states name expansion bot.
woodpeck_fixbot
TIGERcnl
jumbanho
jremillard-massgis
bot-mode
DaveHansenTiger
balrog-kun

I often use this list of mappers as a way of checking Tiger progress.  Note
that woodpeck_fixbot removed Tiger tags on nodes.  Hence, you have an idea
if the street has been touched or not for most of the cases.

Finally, I am not sure that OSMOSE would be a way to track Tiger data.  I
like the positive idea for moving forward.  I suppose that a pin would
render on every Tiger way might work.  The cycle times on some areas a
really slow in OSMOSE.  You'd have to close all the pins after each edit to
see how you are doing.  That effort adds to the pain.  I think the
challenge for the U.S. project is that most of these maps have been
maintained outside the U.S.  If implemented, I am not sure how long OSMOSE
would provide the service.

Regards,
Greg
___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us


Re: [Talk-us] TIGER-completeness visualizer?

2019-12-22 Thread Rihards
On 21.12.19 23:02, stevea wrote:
> As the ITOworld TIGER-completeness visualizer at
> 
> https://product.itoworld.com/map/162?lon=-121.88=37.04=12=true_sidebar=map_key
> 
> is no longer supported, does anybody know of a similar "product" that we can 
> use to visualize how well we have reviewed TIGER roads (and rail?) in a given 
> area?

Maybe authors of http://osmose.openstreetmap.fr/en/map are interested in
adding that.

> I REALLY miss that visualizer!  It showed whether highway=* ways were 
> "touched in the last three years," whether the tiger:reviewed=no tag was 
> removed and so on.  It was very well done and quite informative.
> 
> I suppose a dedicated renderer could be built, that's pretty ambitious, 
> though it is a worthy project, imo.
> 
> Bonus points for your best guess at when OSM will eventually complete a full 
> TIGER data review.  I'll start:  at the rate we're going now, 2045?
> 
> SteveA
> California
-- 
 Rihards

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us


[Talk-us] TIGER-completeness visualizer?

2019-12-21 Thread stevea
As the ITOworld TIGER-completeness visualizer at

https://product.itoworld.com/map/162?lon=-121.88=37.04=12=true_sidebar=map_key

is no longer supported, does anybody know of a similar "product" that we can 
use to visualize how well we have reviewed TIGER roads (and rail?) in a given 
area?

I REALLY miss that visualizer!  It showed whether highway=* ways were "touched 
in the last three years," whether the tiger:reviewed=no tag was removed and so 
on.  It was very well done and quite informative.

I suppose a dedicated renderer could be built, that's pretty ambitious, though 
it is a worthy project, imo.

Bonus points for your best guess at when OSM will eventually complete a full 
TIGER data review.  I'll start:  at the rate we're going now, 2045?

SteveA
California
___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us