PA: I-276 'exit 334'

[rickmastfan67]

Seems that it's still considered as 'Exit 20', and didn't gain a mileage based exit number back in 2000.  The number is only posted going WB on I-276, as it's unnumbered going EB.

So, I see doing two things here:

1) 334 -> I-476 +334
2) 334 -> 20 +334

I'm in favor of #2 since it is posted on the mainline (though lacking the gore sign).  Sure, the '20' would look out of place in I-276's file, but it's what's posted there. :/  But I would also have no problems with #1 too.

[vdeane]

I'd go with option 1... the "exit 20" is actually specific to I-476 southbound, with northbound having no signed number, hence why it's WB-only.

[yakra]

#2 > #1.
Exit numbers [perfectly good transitive verb that I detest using now because it's the same as the name of a prominent American politician that I rather dislike] route numbers.

I SO want to walk away whistling casually, but...
Stupid PTC!

[vdeane]

Even when the "exit number" is incorrect for a majority of movements in the interchagne, which are unnumbered?  Note that exit 20 is not really for I-476, but for the Mid-County toll barrier.  That's why the numbering is weird here in the first place: because it's based off what is convenient for printing toll tickets.  The movements going to/from the barrier to the rest of the PTC are "20".  All others don't have a number.

We can only hope that this will be fixed when the Turnpike goes AET, but it's the PTC, so I'm not holding my breath.

The next-closest situation I can think of is the MassPike, where the situation was "solved" by pretending that the one point per interchange rule does not exist.

[yakra]

The next-closest situation I can think of is the MassPike, where the situation was "solved" by pretending that the one point per interchange rule does not exist.

Where specifically? I'd be interested in having a look at it.

[froggie]

I'm with Val on this one.  Given what the Exit 20 is intended for and given the lack of an exit number eastbound, I'd label it as I-476 and not with an out-of-place exit number.

[yakra]

Using a cross-road label when there's an exit number to be used gives me indigestion; the only way I can really justify this is if we only consider "20" to apply to one part of the interchange, after the exit leaves the mainline. Like so.

How about this?...
Think of this and Exit 333 as "double half interchanges". Consider their functional overlap:

Missing movements:
E->S via Exit 333, cross over Germantown Pike, bear right & loop around
N->W via Exit 20 to Germantown Pike N, loop ramp to Exit 333 interchange

Redundant movements:
W->S; N->E

The rub here is that it's not a perfect fit for the "Two separate interchanges numbered 6" criterion... maybe more of "Two separate interchanges, one numbered and one not".

[vdeane]

The next-closest situation I can think of is the MassPike, where the situation was "solved" by pretending that the one point per interchange rule does not exist.

Where specifically? I'd be interested in having a look at it.

14/15 (although the ramp distance and odd interchange makes me think that this might actually be warranted), and 18/20.  Basically the locations where the numbering was messed up by putting a toll barrier inside an interchange.

[yakra]

Ah yes. Thought it might have to do with these ones.

the situation was "solved" by pretending that the one point per interchange rule does not exist

Less pretending the rule didn't exist, and more that it hadn't yet come into being -- the Interstates were done in the very early days (2006? 2005?) before 1PPI had really started to solidify as a rule. Tim's early drafts thus went a bit more by "exit number" than by 'interchange".

(IMO, the canonical example was in the original version of NY I-278, which had separate points for exits 8 & 9.

The 18/20 example, if it were being drafted today, would get a single point, numbered 18. ("Exits 4 & 5 in one interchange. For a single interchange with ramps given different exit numbers, use the lower number.")

14/15 (although the ramp distance and odd interchange makes me think that this might actually be warranted)

Yeah, it's an odd one. Points "14" and "15" are placed 0.61 mi apart, which meets the "double half interchange" threshold.
You could even consider it one big spread-out interchange too -- I considered the idea of collapsing it into a single point, with the actual connections to/from I-90 being made farther west, near 14, but then noticed that 15 approximates the gores of the ramps to/from Park Rd. It can stay; why not. Plus, we get a graph connection & Intersecting/Concurrent Routes.

As for 14, it has Historical Neato Factor. Just on the off-chance we have any users who haven't been on the Mass Pike since 1957.

[Markkos1992]

Well I believe I am going to go-ahead and go with option 1 for the following reasons:
1. It seems to be best to treat this as if the "Exit 20" tab on I-276 does not exist at all.  Based on this thread on I-80 near Delaware Water Gap, I determine that we are not in favor on the fudging of exit numbers at this point.   Using an out-of-place "20" only because of abnormal circumstances as Valerie mentioned just feels like fudging an exit number because one happens to be here.
2.  The majority consensus seems to be with option 1 especially with me maintaining PA now.  (3>2 from what I see)
3.  Going from 333 to 20 to 339 randomly just looks odd and confusing to me normally. 

https://github.com/TravelMapping/HighwayData/pull/2167

[yakra]

FWIW my opinion has also flipped to agree here.

[dave1693]

I am also in agreement with Mark's action here.