Highway Data Discussion (CHM/TravelMapping)

[yakra]

I like the idea Si had for UK A Roads -- starting an issue on GitHub & saying "post your issues & comments here."
I don't think we need anything more formal. (Unsure what that'd even be.)
I like the idea of a GitHub issue, keeping most of the nuts-n-bolts discussion over there, and not going too spamola on AARoads.
At the same time, I think a topic here could bring us a lot more pairs of eyes, and site enthusiasts who don't necessarily want to set up a GitHub account. Comments here could be copied over to a GitHub issue, as I've seen you do.
Are you thinking of keeping this geared more toward CHM/TM veterans who are well versed in The Manual, Peer Reviewing, etc?

[english si]

I like the idea Si had for UK A Roads -- starting an issue on GitHub & saying "post your issues & comments here."

Only problem is how long do I leave it for, before activating stuff?

I checked the in-development SRBA system and I think it's currently complete.

I've made some minor tweaks since then of things I saw, but thanks for taking initiative!

I would add a waypoint at the end of srb.a001nov:

Done.

What's required to get the system online?

I'd want a CHM collaborator to look over the files, but that's really it, as far as I can see. Note - it's 5 files, all freeways and all pretty short.

[oscar]

I like the idea Si had for UK A Roads -- starting an issue on GitHub & saying "post your issues & comments here."
I don't think we need anything more formal. (Unsure what that'd even be.)
I like the idea of a GitHub issue, keeping most of the nuts-n-bolts discussion over there, and not going too spamola on AARoads.
At the same time, I think a topic here could bring us a lot more pairs of eyes, and site enthusiasts who don't necessarily want to set up a GitHub account. Comments here could be copied over to a GitHub issue, as I've seen you do.

I think a separate topic here often would work best. That's what I plan to do shortly for Alaska State Highways, which will entail some general issues about what routes to include in the system, not just nuts-and-bolts stuff.

[michih]

I checked the in-development SRBA system and I think it's currently complete.

I've made some minor tweaks since then of things I saw, but thanks for taking initiative!

Thanks.

What's required to get the system online?

I'd want a CHM collaborator to look over the files, but that's really it, as far as I can see. Note - it's 5 files, all freeways and all pretty short.

Is it possible that I could be a collaborator?

I thought all in-development systems are in this list (all systems at the end of the list with "Yes" notification.) and the region managers are here.

Some new Finnish routes recently went online but I don't know where it's announced which new systems are in progress. Is there a list?

[froggie]

I think a separate topic here often would work best.

I agree with Oscar, especially since I'm not on GitHub.

[sipes23]

I'm not quite sure where to put this one.

US87Bus (Buffalo, WY) or US87BusBuf is goofed up. I was in Buffalo the other day (again, because once really isn't enough) and I'm pretty sure that Business 87 doesn't follow the north bypass. The signs weren't super clear on the ground (though I wasn't paying super close attention). If I'm in the area again (and I will be), I can field check this if it's still unclear.

Feel free to split this off into its own topic as necessary.

Highway here: http://tm.teresco.org/devel/hb.php?r=wy.us087busbuf

[Jim]

US87Bus (Buffalo, WY) or US87BusBuf is goofed up. I was in Buffalo the other day (again, because once really isn't enough) and I'm pretty sure that Business 87 doesn't follow the north bypass. The signs weren't super clear on the ground (though I wasn't paying super close attention). If I'm in the area again (and I will be), I can field check this if it's still unclear.

I've created Issue #156 https://github.com/TravelMapping/HighwayData/issues/156 in the HighwayData repository for this one.

[rickmastfan67]

I'm working on getting I-269 ready to go into the site for TN.  I'll also help out Froggie and do the short MS section since we don't have a state highway system yet for Mississippi.  That way, the border points will be synced.

EDIT: Now done. https://github.com/TravelMapping/HighwayData/pull/158

[mapcat]

Since dfilpus isn't an active collaborator anymore, could someone else make a few updates in Ohio? In particular, there are two new routes to add, at least one to delete, and two to adjust. Details are in the CHM forum, but the basics are:

Delete OH 238
Reroute OH38 to follow old 238; former 38 between 238 and Washington C.H. is no longer an ODOT route
Extend OH762 east of US23 along Duvall Rd then north along Ashville Pike, terminating at Airbase Rd
Add OH607 along Monastery Rd between OH60 and OH78 east of McConnellsville
Add OH684 along its former alignment and the northern leg of the former alignment of OH692 between OH143 and OH681 (the route was supposed to be decommissioned, but ODOT added it back after local opposition. Dave deleted it but never got around to restoring it)

[bejacob]

Since dfilpus isn't an active collaborator anymore, could someone else make a few updates in Ohio? In particular, there are two new routes to add, at least one to delete, and two to adjust. Details are in the CHM forum, but the basics are:

Delete OH 238
Reroute OH38 to follow old 238; former 38 between 238 and Washington C.H. is no longer an ODOT route
Extend OH762 east of US23 along Duvall Rd then north along Ashville Pike, terminating at Airbase Rd
Add OH607 along Monastery Rd between OH60 and OH78 east of McConnellsville
Add OH684 along its former alignment and the northern leg of the former alignment of OH692 between OH143 and OH681 (the route was supposed to be decommissioned, but ODOT added it back after local opposition. Dave deleted it but never got around to restoring it)

There is one more.

Delete OH 794. The route is no longer an ODOT route and is now maintained by Clark County. I drove the route in early October and can verify that no state route shield remain.

[rickmastfan67]

Doing some 'work' on getting the extension of {401} done for the site.

Only thing is I can't decide on a label for one of the 'interchanges'.  It's just an interchange with on-ramps from {3} to {401} in both directions.  {401} doesn't have any offramps there.
http://www.openstreetmap.org/?lat=42.238300&lon=-83.003517

I'm tentatively calling it '9'.  I would have given consideration to labeling it as 'ON3', however, I need to keep 'ON3' as a hidden point (as well as '0') to prevent breaking people's files at the old end of the {401}, which I'm labeling as '10A'.  '10' will be at the overpass for {3}.  While it will only be .2 miles apart, I feel it's justified because of the historical end of the {401} being there, plus, it allows anybody who cliched the {401} when it temporarily ended at the recently built {3} roundabout to have a 'proper' point.  I even had to add 3 'extra' shaping points that normally wouldn't be justified because of the close proximity of {3} there to prevent any false multiplexes.  I'll be doing the same with {3}'s file.

Here's the part I'm adding if anybody wants to check it out in the editor and make any comments before I submit to GitHub sometime tomorrow.

Code Select Expand

5 http://www.openstreetmap.org/?lat=42.264479&lon=-83.044217
6A http://www.openstreetmap.org/?lat=42.256237&lon=-83.038847
6 http://www.openstreetmap.org/?lat=42.251052&lon=-83.035929
+X002(ON401) http://www.openstreetmap.org/?lat=42.248867&lon=-83.033097
7 http://www.openstreetmap.org/?lat=42.246048&lon=-83.026117
+X003(ON401) http://www.openstreetmap.org/?lat=42.241918&lon=-83.013934
9 http://www.openstreetmap.org/?lat=42.238300&lon=-83.003517
+X004(ON401) http://www.openstreetmap.org/?lat=42.234416&lon=-82.996838
10 http://www.openstreetmap.org/?lat=42.233300&lon=-82.993512
10A +0 +ON3 http://www.openstreetmap.org/?lat=42.233371&lon=-82.988211
13 http://www.openstreetmap.org/?lat=42.244754&lon=-82.968943
14 http://www.openstreetmap.org/?lat=42.247598&lon=-82.958088
21 http://www.openstreetmap.org/?lat=42.240740&lon=-82.873574


[bejacob]

I found what may be a concurrency issue near Wilmington, NC. Just west of the city, four route (US17, US74, US76, and NC133) all cross the Brunswick River. If my memory is correct (I drove this route over a year ago) all 4 use the same bridge. Looking at my map for the state http://tm.teresco.org/hbtest/mapview.php?rg=NC&u=bejacob and zooming into the area in question, I see two lines, one red (US highways) and one brown (state highway). 

It appears that NC133 has two hidden shaping waypoints that the other routes on this concurrency are missing. It would seem that easiest fix is to remove those two hidden waypoints on NC133. The other option is to add them to the other three routes that share the concurrency.

As it has been a while since I drove this route, it might be helpful for someone to verify the concurrency of the US routes with NC133 before making any changes.

Brian

[dfilpus]

North Carolina highway changes that need to be updated:

I-74 High Point: Extended west to I-40. Delete I-74 Future High Point.

US 421: Moved to Sanford Bypass. Create US 421 Business Sanford along old route.

US 401: Moved to Rolesville Bypass. Create US 401 Business Rolesville along old route.

NC 24/87: Moved from Bragg Boulevard to Murchinson Road (NC 210) and NC 295.

[yakra]

Only thing is I can't decide on a label for one of the 'interchanges'.  It's just an interchange with on-ramps from {3} to {401} in both directions.  {401} doesn't have any offramps there.
http://www.openstreetmap.org/?lat=42.238300&lon=-83.003517

I'm tentatively calling it '9'.

I agree; I'd probably do the same there.

I would have given consideration to labeling it as 'ON3', however, I need to keep 'ON3' as a hidden point (as well as '0') to prevent breaking people's files at the old end of the {401}, which I'm labeling as '10A'.  '10' will be at the overpass for {3}.  While it will only be .2 miles apart, I feel it's justified because of the historical end of the {401} being there, plus, it allows anybody who cliched the {401} when it temporarily ended at the recently built {3} roundabout to have a 'proper' point.

Agree here too.

I even had to add 3 'extra' shaping points that normally wouldn't be justified because of the close proximity of {3} there to prevent any false multiplexes.  I'll be doing the same with {3}'s file.

While +X002(ON401) may not be strictly necessary to prevent a false multiplex (centerlines are different at Todd Ln), I would probably have put this point here too. Cuts down on visual confusion on closely zoomed-in maps. ON3 could get away without a shaper here IMO (my 2nd choice would be a visible point at Huron Church Line Rd).

[oscar]

I found what may be a concurrency issue near Wilmington, NC. Just west of the city, four route (US17, US74, US76, and NC133) all cross the Brunswick River. If my memory is correct (I drove this route over a year ago) all 4 use the same bridge. Looking at my map for the state http://tm.teresco.org/hbtest/mapview.php?rg=NC&u=bejacob and zooming into the area in question, I see two lines, one red (US highways) and one brown (state highway). 

It appears that NC133 has two hidden shaping waypoints that the other routes on this concurrency are missing. It would seem that easiest fix is to remove those two hidden waypoints on NC133. The other option is to add them to the other three routes that share the concurrency.

Good catch! I edited the US route files as part of adding NC 140 (among other NC and VA major changes), but forgot that NC 133 (which is indeed still concurrent with the US routes, I drove it last week) would also be affected. I'll fix the NC 133 route file accordingly, so TM will once more detect the concurrency.

EDIT: NC 133 fix is now in the TM database.

[dfilpus]

North Carolina highway changes that need to be updated:

I-74 High Point: Extended west to I-40. Delete I-74 Future High Point.

US 421: Moved to Sanford Bypass. Create US 421 Business Sanford along old route.

US 401: Moved to Rolesville Bypass. Create US 401 Business Rolesville along old route.

NC 24/87: Moved from Bragg Boulevard to Murchinson Road (NC 210) and NC 295.

In addition:
NC 87 Bypass Sanford: Created along the Sanford Bypass from NC 87 to US 1.
NC 44: Extended west from I-795 to US-70.

[oscar]

North Carolina highway changes that need to be updated:

I-74 High Point: Extended west to I-40. Delete I-74 Future High Point.

US 421: Moved to Sanford Bypass. Create US 421 Business Sanford along old route.

US 401: Moved to Rolesville Bypass. Create US 401 Business Rolesville along old route.

NC 24/87: Moved from Bragg Boulevard to Murchinson Road (NC 210) and NC 295.

In addition:
NC 87 Bypass Sanford: Created along the Sanford Bypass from NC 87 to US 1.
NC 44: Extended west from I-795 to US-70.

Thanks (and good to see you back here). Do you have route files for any of these changes, that you could e-mail to Jim or me to get them through GitHub into the TM database?

[rickmastfan67]

ON3 could get away without a shaper here IMO (my 2nd choice would be a visible point at Huron Church Line Rd).

I did add a point there.  It was mostly a relocation of the old 'HurChuRd' point that was just to the West.

[Jim]

Heads up for highway data maintainers: I've expanded the updates.csv date field entries to YYYY-MM-DD format. All needed changes to the file itself have been made and other parts of the project that process updates data are updated, and the site is live with the new format in place.

[Jim]

I have an idea for comment and, I hope, implementation in the near future.  I propose expanding our current "active" vs. "in development" options for highway systems to be three categories:

1) Active systems: systems that we believe are stable and up-to-date, and for which we intend to do all we can to void breaking .list entries unnecessarily, and will publish updates entries when user-affecting changes are made.  This is exactly what we do now for active systems.

2) Systems in review: systems we believe are complete but still might need some fixes before becoming active.  For these systems, users are encouraged to start creating .list entries, but no promise is made that changes before activation would not break .lists.  I expect there are about a dozen systems that would slide nicely into this category given their current status, such as usaut, usanm, usavt, etc.

3) Systems in development: systems that are incomplete or have other mistakes, omissions, or other issues to be resolved before anyone other than the primary developers of the system should be looking at them.  There are a number of systems that would fall into this category for now, like the Asian systems, usanp, usaca, usafl.

The systems.csv file's last column would have entries "active", "review", or "development".  All would continue to be processed and would generate datacheck and other errors, but those datachecks would be broken into three tables instead of two.  Only errors in the third table, the in-development systems, should be ignored.  It would be continue to be the case that there should be no errors listed in the active systems, and one requirement for activation would continue to be that no datacheck errors exist in the system.

The motivation for this further breakdown of inactive systems is to speed the transition from review to activation.  I would modify the data processing program and the web tools to be able to generate two instances of the database.  One would be the active-only systems that we have now.  The second would include all "review" systems.  Not only would this be fun for travelers who could then see their maps and stats for more systems even though we haven't activated them yet, but would help the developers of those systems find errors by seeing the maps.  One would switch from the standard view to the view with the review systems included by adding a parameter to the URLs.  This would be automated once we have a better front end to the whole site.

I doubt that I will do much if anything with this before Thanksgiving or maybe even Christmas, but I don't think it's a huge effort, and I'd be interested in any thoughts on it.

[michih]

Not only would this be fun for travelers who could then see their maps and stats for more systems even though we haven't activated them yet, but would help the developers of those systems find errors by seeing the maps.

I like your idea  . I think the system developers would get more feedback from users who are able to see their maps of review systems.

[oscar]

2) Systems in review: systems we believe are complete but still might need some fixes before becoming active.  For these systems, users are encouraged to start creating .list entries, but no promise is made that changes before activation would not break .lists.  I expect there are about a dozen systems that would slide nicely into this category given their current status, such as usaut, usanm, usavt, etc.

3) Systems in development: systems that are incomplete or have other mistakes, omissions, or other issues to be resolved before anyone other than the primary developers of the system should be looking at them.  There are a number of systems that would fall into this category for now, like the Asian systems, usanp, usaca, usafl.

I agree with where you would place usaca for now (where it will stand once we've implemented the three-category system, can be decided later).

What about usaak? I think "in review" is the right category for that system, for which review comments have already started coming in. If so, I'd like to change the title of that thread accordingly.

I like the idea of being able to tentatively map my travels pre-activation. I usually find, when a new system is activated, gaps in my mapping of the new system because I left out a list file entry or two. I could more easily spot and fix in advance those gaps with a tentative map.

[english si]

I'm go for the three tier system.

[mapcat]

I like your idea  . I think the system developers would get more feedback from users who are able to see their maps of review systems.

I was thinking the same thing. Glad to know you're considering it.

[Duke87]

I like this idea as well, for two reasons:
1) For users actively paying attention, it makes us able to claim travels in new systems sooner than we otherwise could
2) Sometimes the best way to QC something is to just start using it and see what problems you encounter in real world use. Having users start logging travels in "review" systems would also help spread the workload out by effectively crowdsourcing much of the process of locating mistakes.