Waypoint splitting at interchanges

[SSOWorld]

I am confused here.  What is the protocol for interchanges and waypoint placement?  Examples include US-41 at I-69 off Evansville, IN and FL826.  They both have ramps that split before they intersect a route. 

[yakra]

I am confused here.  What is the protocol for interchanges and waypoint placement?

The manual on point placement is at http://cmap.m-plex.com/tools/manual_points.php
Some of the rules of thumb include:

• Usually position the waypoint at the point where the centerlines of the two highways cross. Often the same coordinates can be used for both highways.
• These centerlines are defined by through lanes, not by turning lanes at intersections nor the equivalent in interchange ramps.

Generally, the rule is "one point per interchange".
Of course, what exactly is "one interchange" and what's "more than one interchange" can get wibbly-wobbly at times, and is subject to interpretation.
Re splitting waypoints/interchanges, the manual says:

• Double half interchanges: Usually use one central point and treat both halves as a single, full interchange. Exceptions: a clear gap of at least 0.5 mi/0.8 km separates the two halves, or each half connects to a different highway that we are also mapping.

Examples include US-41 at I-69 off Evansville, IN

The US41 file looks fine. One point, in the center of the interchange.
Were you referring to IN I-69Whe itself, with its KenAve endpoint beyond the Exit 0 interchange?  That was the subject of sone discussion a while back; I guess it was decided to leave it in...

and FL826

US41 @ FL826 looks like a normal interchange to me.

FL826 itself for the most part looks pretty straightforward.
• FL916 ant I-75 both connect to a different highway that we are also mapping, so this looks legit.
• The segment from I-95(12A) to 2ndAve looked like it was pretty difficult to handle. Without looking over it in too gory detail I'm pretty OK with how this was done.

They both have ramps that split before they intersect a route.

I'm not really certain what you're getting at here. This is, by definition, what ramps do... 
I-69@US41, US41@I-69, & US41@FL826, I see ordinary ramps splitting from the mainline.
Southbound FL826@US41, I see I see a ramp that leaves the mainline and then splits in two, but this is very straightforward geometry easily handled by "one point per interchange".

[mapcat]

Re Evansville, there is more than one interchange there. If we consider Veterans Pkwy and I-69 to be the same road here, westbound there are separate ramps go to US 41 and to Kentucky Ave.

The short segment between Kentucky Ave and US 41 is something we're calling I-69 for the time being because it was once part of I-164, all of which was added to I-69's mileage a few years ago.

[yakra]

A good example of open-to-interpretation wibble-wobble.

Re Evansville, there is more than one interchange there.

And there's less than two.

If we consider Veterans Pkwy and I-69 to be the same road here, westbound there are separate ramps go to US 41 and to Kentucky Ave.

I neither agree nor disagree with mapcat's interpretation here... One thing I'm paying attention to is how Kentucky becomes one-way southbound, turning into onramps for US41 & I-69. The northbound ramp feeds directly into the thick of the cloverleaf loops. Travelers following historic US41 onto Kentucky must take the loop ramp in the NE quadrant, then exit again from the C/D road. This could work in favor of a one-interchange argument. Wibble! Wobble!

This reminds me a bit, on a less complicated scale, of the big I-96/275/696/MI5 mixer. No real way to do that one perfect, either.

The short segment between Kentucky Ave and US 41 is something we're calling I-69 for the time being because it was once part of I-164, all of which was added to I-69's mileage a few years ago.

Prior discussion is here, for those interested.

[SSOWorld]

and FL826

US41 @ FL826 looks like a normal interchange to me.

FL826 itself for the most part looks pretty straightforward.
• FL916 ant I-75 both connect to a different highway that we are also mapping, so this looks legit.
• The segment from I-95(12A) to 2ndAve looked like it was pretty difficult to handle. Without looking over it in too gory detail I'm pretty OK with how this was done.

They both have ramps that split before they intersect a route.

I'm not really certain what you're getting at here. This is, by definition, what ramps do... 
I-69@US41, US41@I-69, & US41@FL826, I see ordinary ramps splitting from the mainline.
Southbound FL826@US41, I see I see a ramp that leaves the mainline and then splits in two, but this is very straightforward geometry easily handled by "one point per interchange".

I was and forgot to mention specifying the interchange with the Turnpike. and just realized I meant FL836 (wrong #) and for that matter the turnpike itself at exit.  26 and 26A waypoints on the turnpike and FlTpk(26) & FlTpk(26A) on FL836.  The flyovers from the south oppose the loop ramps from the north.  It doesn't help that NW 12th street interchange (27) with the turnpike is really close to the 26A waypoint (Which is actually Exit 26).

[rickmastfan67]

FL826 itself for the most part looks pretty straightforward.
• The segment from I-95(12A) to 2ndAve looked like it was pretty difficult to handle. Without looking over it in too gory detail I'm pretty OK with how this was done.

Don't get me started on that one. lol.  Too many posted routes in that little clusterfuck.

I was and forgot to mention specifying the interchange with the Turnpike. and just realized I meant FL836 (wrong #) and for that matter the turnpike itself at exit.  26 and 26A waypoints on the turnpike and FlTpk(26) & FlTpk(26A) on FL836.  The flyovers from the south oppose the loop ramps from the north.  It doesn't help that NW 12th street interchange (27) with the turnpike is really close to the 26A waypoint (Which is actually Exit 26).

Ok, for the 'FLTpk(26)' & 'FLTpk(26A)' points on FL-836.  There's a simple reason for that.  The route used to end on the 'ramps' to "FLTpk(26A)".  Then, back in 2007, they built the new 'extension', "FLTpk(26)", which was also only 'SunPass' for the start.  So, both points are there to allow people who clinched the original route, or couldn't use the new 'SunPass' alignment (till it became all electronic).

[regoarrarr]

Perhaps only tangentially related to the topic at hand - but what about roads where it's one-way and the westbound and eastbound (or north/south) routes go on different roads?  I was thinking about this as I was crossing the Mississippi near Quincy, Illinois this weekend.  US24 westbound and eastbound have a different bridge

https://www.google.com/maps/place/Quincy,+IL/@39.9354155,-91.4223369,14.62z/data=!4m5!3m4!1s0x87ddf79cd32e3b37:0x5f65b820ccb464be!8m2!3d39.9356016!4d-91.4098726

(Of course to be sure I "got" both of them, I crossed 3 times just to be safe  )

[Jim]

We always (others can correct me if there are exceptions) consider separate roads for a one-way pair as a single route, plotting down the middle of the two when possible.  With some oddballs like US 202 in Holyoke, Mass., the TM route follows one of the directions, in that case, southbound.