ON: cantch_con.csv issues

[yakra]

I remember when Tim first implemented the Connected Routes CSVs, he wrote a script to automate the process, then asked us to review the results and look for any errors or anything goofy. Looks like this issue flew under the radar for a few years:

This primarily affects rickmastfan67's turf.

From the first line of cantch_con.csv:
cantch;TCH;;(Main);bc.tchmai,ab.tchmai,sk.tchmai,mb.tchmai,on.tchlak,on.tchnor,on.tchott,qc.tchmai,nb.tchmai,ns.tchmai

Look at the 3 bolded routes in the HB and you'll see that they all come together in a three-pronged "route" in North Bay.
At the very least, on.tchnor should be broken out into its own route and CSV line.

Philosophical stuff:
The different TCH branches in Ontario were considered separate enough to be broken into different chopped routes, even if they intersect and form continuous routes. Should they thus be similarly broken up in the connected routes CSV? A drawback to this approach would be that we'd no longer have a single "Main" coast-to-coast TCH route. Just floating the idea; I am OK with them staying prefer they stay connected.

If keeping on.tchgeo & on.tchcen connected, I suggest changing the groupname to "Georgian Bay - Central Ontario" to more clearly acknowledge both chopped routes.

on.tchkir & qc.tchkir should remain connected, of course.

[si404]

That's a big mess philisophically.

Is there any reason why the Georgian Bay-Central Ontario route is split in two?

The main route seems to follow the Lake Superior-Ottowa route. And is there any reason why that isn't one chopped route? At North Bay, a split is deemed to exist, but not when the Northern Ontario route returns at Nipigon, nor when other loops come off it (the ON11 loop seems a more important one than the other loops, but that the 'main' route is ON17/417 all the way suggests that it ought to be considered the main route). I've just read the CHM thread, and it seems to be taking hobbyist names with no official support.

---

Obviously the North Ontario route needs its own connected route, but I'm not sure there needs to be a split across the country.

[rickmastfan67]

Is there any reason why the Georgian Bay-Central Ontario route is split in two?

You would have to ask Tim this question, as he made the routes before I made the main Ontario files. 

[si404]

OK, next question - any reason why it has to still be split in two?

[yakra]

I'll have to reread the CHM thread myself. I've long since forgotten what went on there, and am curious to see the reasoning behind it all. I *thought* there may have been something beyond hobbyist names, but it sounds like I'm off base on that.

[si404]

http://clinched.s2.bizhat.com/viewtopic.php?t=108

I've not spent that long, but I can't find anything official in there using those route names. Having (I think) found the hobbyist sites, the names are barely there on one of them.

[yakra]

Ugh. How deep does this rabbit hole go? For the sake of maintaining our sanity, and not having to reinvent the wheel, I'm fine with keeping the status quo, save for:
* breaking out on.tchnor into its own route and CSV line
* optionally changing the groupname of on.tchgeo & on.tchcen to "Georgian Bay - Central Ontario" or somesuch

[yakra]

Some discussion on GitHub leads me to bump this.

[Jim]

At the least, it makes no sense the way we have it now.  on.tchlak doesn't "continue as a connected route" eastward as on.tchnor.

To me, this looks like the cases where we split I-35 in Minneapolis-St. Paul and in Dallas-Fort Worth.  There is no longer a single implied I-35 in TM.  Is there a reason we need a single implied TCH Main more than we need an implied I-35?

[yakra]

Not a 1:1 comparison with I-35 IMO. There, there's a numerical designation that is discontinuous. Here, there's a signed TCH, which exists in multiple places. Numbers are secondary, and our segment names appear to be... invented?

I think it would be a Good Thing to be able to track our travels on the coast-to-coast mainline TCH.
...If there is such a thing -- Does our current implied "(Main)" route have more of a legitimate claim on being the Main route than any of the other potential routes?

[Jim]

I'd argue we need to pick one as a "Main".  If none is the "offically" the Main (is that an actual concept or a CHM invention?) maybe pick the shortest or most traveled.

In my own travels in Ontario, some segments do have signed names that we might want to use where they exist.

[yakra]

In my own travels in Ontario, some segments do have signed names that we might want to use where they exist.

Good to know there's some official support for our segment names, and that they're not just hobbyist inventions. This gives me some confidence they're legitimate names.

I'd argue we need to pick one as a "Main".

Agreed.

If none is the "offically" the Main (is that an actual concept or a CHM invention?)

I believe this is the first instance of the "Main" concept coming into use on CHM. Don't know about whether it's official or not, and thus whether it'd be a CHM invention.

maybe pick the shortest or most traveled.

The simplest most sensible adaptation of the status quo would be to break on.tchnor out of the (Main) connected route, and make it its own connected route. TCHNor can't be part of (Main) as things are, as at its west end, TCHLak continues in both directions. This would create another 3-pronged route.
This would leave the (Main) segment via TCHLak & TCHOtt.
Is it the shortest?
Longer than TCHKir by 64 km, however, QC TCHMai is already established as the route connecting to TCHOtt.
Shorter than TCHKen by 4.4 km
Longer than TCHNor by 35 km, however, TCHNor cannot be part of the (Main) connected route for reasons cited above.
Shorter then TCHVoy by 66 km. TCHVoy also cannot be part of (Main) for the same reason; other chopped routes do not end at its endpoints.
Shorter than TCHGeo+TCHCen by 170 km. TCHGeo+TCHCen also can't be used.
Continuing thru movements at intersections:
The (Main) route gets the main thru traffic movement at all its intersections with other TCH routes except for both ends of TCHKen. the ON17 vs ON17A designations give a pretty good indication of what's considered the main & auxiliary route here, and in any case, I wouldn't want to make major changes here that mess with people's list files anyway.
Most traveled: Determining this means a bit more getting into the data than I think necessary.

This all being said, I only see one workable solution for the (Main) TCH route:
cantch;TCH;;(Main);bc.tchmai,ab.tchmai,sk.tchmai,mb.tchmai,on.tchlak,on.tchott,qc.tchmai,nb.tchmai,ns.tchmai
It all comes down to a simple matter of where the chopped routes end, and what can be concatenated together to make a single (Main) connected route.

[7/8]

Here's two more TCH photos:

Georgian Bay Route on the 400 / Highway 12 concurrency


Central Ontario Route on Highway 7 near Perth


I would personally consider the Lake Superior and Ottawa Valley routes as the "main" TCH route across Ontario.

[rickmastfan67]

I'm game with whatever you guys agree with.

At least people could fall back on the multiplex detector if they input the Ontario base routes into their list files.  So, splitting up any TCH file would hurt a little less.

[yakra]

I'm going to go ahead and break out the Northern Ontario segment from (Main), and make it its own Connected Route.
It fixes the 3-pronged issue, easy & simple. There's not even any reason to disrupt anything.
https://github.com/TravelMapping/HighwayData/pull/1504