Remove tiger:reviewed tag when user performs certain edits

[bhousel]

I feel bad that tiger:reviewed=no is used by the cycle.travel router to penalize roads, but iD (rightly) hides this tag from users. New users are doing substantial cleanup to the US map and they shouldn't need to know anything about what tiger:reviewed is or how to use the raw tag editor to remove it.

Are we at the point now where we can safely remove the tag for the user if they make a substantial edit to a way? I'm thinking it's probably safe to remove tiger:reviewed (and maybe certain other tiger: tags) after any change to:

• road classification (preset)
• visible fields: oneway, maxspeed, structure, access, surface, cycleway
• realignment (moving child nodes)

I don't think it's safe to remove the tiger:reviewed tag for every kind of edit (e.g. splits or merges or connecting ways to other ways), but only for the ones listed above..

Just putting this out there for discussion.. Thoughts?

[pnorman]

I wouldn't base the discussion around what a data consumer is doing. Most of the changes you mentioned would make tiger:reviewed irrelevant to cycle.travel anyways.

[darrell]

I think those are entirely reasonable situations to move the tag. All of those imply that the way has, in fact, been reviewed.

[simonpoole]

@bhousel I'm likely not untypical in that I've done two of the three tasks on 100s, more likely 1000s of roads in the US and would have never considered changing/removing tiger:reviewed .... simply because I was not there (I'm not actually sure if I've removed it on all roads that I surveyed in person, but that is a different story)..

[bhousel]

@simonpoole I thought about what you said some more, and I still think it's entirely reasonable to remove this tag even if someone is armchair mapping. The TIGER data is from 2005 and I believe all of the Bing imagery covering the United States (at least, at the zooms that iD works in) is much newer than that.

[brycenesbitt]

I think that either tiger:reviewed needs to be explained to the mapper (not hidden).
Or, removed when the node/way is edited.

Most of the time I find edited tiger ways, the mapper has not felt bold enough or been empowered enough to remove the tag tiger:reviewed. The problem in part is that who defines when a road has been edited "enough" to remove the tag?

That said I like knowing that data came from tiger into perpetuity. The tiger origin is often helpful, even after a node/way has received a lot of attention.

[jfirebaugh]

There are (at least) four common problems with imported TIGER data:

• Misalignment, both minor and gross
• Incorrect names, mostly commonly the presence of names from nearby roads on tracks or service roads that should be unnamed
• Incorrect classification
• Lack of surface tags

Technically tiger:reviewed should be removed only after all of these issues are addressed, and I'm not confident that iD can reliably detect that in an automated way.

[bhousel]

@jfirebaugh How about a different approach. We implement changeset hashtags (#2834), and if a magic hashtag like #tiger is present, we remove tiger:reviewed from everything the user touched.

(I've been doing a lot of these cleanups lately and really looking for ways to make this kind of editing easier. 🐯 )

[pnorman]

Technically tiger:reviewed should be removed only after all of these issues are addressed, and I'm not confident that iD can reliably detect that in an automated way.

This is not a universally held view - part of the problem with tiger:reviewed is that there's no clear conditions on when to remove it.

[jfirebaugh]

@pnorman Yes, that's also a problem.

I do a lot of TIGER cleanup as well. Personally, I'm happy just deleting tiger:reviewed in the raw tags editor after I fix all the issues with a given way. To me that seems like an interaction and level of effort commensurate with the experience and effort necessary to perform TIGER cleanup.

[bhousel]

Fair points, I'm dropping this issue for lack of interest.