[Talk-us] Excellent progress, u.s.

stevea steveaOSM at softworkers.com
Tue Apr 17 21:59:24 BST 2012

>>Sure, Alan.  I'll try to help by explaining a core of my process, 
>>and you and others can take it from there.
>Great job - thanks for this. I'm sure it helps.

Appreciate the kudos.  We might all share like this when asked.  This 
is called a "workflow" and workflows with specific numbered steps are 
valuable/crucial nuggets of knowledge when somebody needs one to 
solve a specific problem AND a workflow HAS BEEN SHARED.  So, ask, 
and, if you KNOW, share WHEN asked!

>>7)  For ways (I'm not going to explain points), here is what I do:
>So, we're basically duplicating the existing way and then "blessing" 
>it. Is this really sufficient - to verify the "tainted" geometry 
>instead of re-drawing it? If so, why is it not sufficient that, in 
>many, many cases, the original creator of the way has not accepted 
>CT, but many other accepting mappers have afterwards aligned (i.e. 
>moved nodes) and tagged (in my case, with sources of sat imagery, 
>local photo survey, county records, and/or even GPS survey) it? 
>Haven't I already "blessed" it? Can't the redaction bot look at the 
>source tag and see this?
>Another point, at least in SoCal, is that many of our "tainted" ways 
>are created by "blars", who has not accepted the CT. However, these 
>are TIGER-imported ways. They carry the TIGER tags. I'm sure they 
>could be verified as having come from the TIGER import. They were 
>no-doubt the result of having split an existing TIGER way. In this 
>case, why is it not sufficient to see the TIGER tags on the way to 
>consider it "blessed" along with all the other TIGER ways? 
>Especially when tagged afterwards by accepting mappers with sources 
>as above?
>>8)   Repeat step 7 for all bad ways (or points) in the list of 
>>"data loss" that License Problems displays.
>I'd like to suggest step 8.5: Run the OSM validator. It will find 
>all the intersections that were missed, and probably a bunch of 
>other problems that may or may not have pre-existed.
>>9)  Upload your re-mapping efforts.

I agree that a step 8.5 to run JOSM Validator is well-indicated here. 
Thank you for that excellent suggestion.

>So, can someone from the redaction squad comment on the logic being 
>used and the questions above?

I don't know that they reply to this list.  I believe they read it, 
but I have no proof of that.

Finally, it is true that my suggested workflow "duplicates an 
existing (tainted) way and then blesses it."  Yes.  The additional 
step of visually verifying with Bing Sat allows this to crystallize 
into a "solidly legal" footing:  "I can see it in Bing, therefore the 
duplication of something which was unlicensable is now licensable." 
In other words, "re-drawing" is not necessary (it is sufficient), but 
it is overkill (unless points are also tainted).  Duplication + 
visual inspection via Bing seems sufficient to me, but it would be 
really, really good to get the redaction squad to directly address 
that point.  Right here (talk-us) would be just fine.  OSM's wiki 
would, too.


More information about the Talk-us mailing list