chng: China National Expressways

[michih]

 https://www.skyscrapercity.com/posts/193173862/

G2211 Changzhi-Yan'an National Expressway Qinyuan to Guxian North section Opened

[michih]

It seems that most was changed with https://github.com/TravelMapping/HighwayData/pull/8354 except of

G15:
Move JS/SH to the borderline, apply to G4221 and JS's G15
293 is off, centerline
332 is off, centerline
SH/ZJ is off, centerline, apply to ZJ's G15

G60:
64 is off, trumpet
SH/JS is off, apply to ZJ's G60, ZJ's G92, SH's G92, SH's AH3 and ZJ's AH3


G92:
Does it extend to S20 as indicated on OSM, concurrent with G60? GM does not indicate G92 at all.


G1503:
Add wp south of 132

G4221:
see G15

I think that another G1503 in the NE should be added if OSM + GM right that it is in service and signed as such.

You've merged it into main route but removed the short additional G1503 segment between G1503's exit 204 and S20. you should add it again.

[Nagamasa]

I would like to disagree on the freeway-freeway double-trumpet setup at G60 and G1503 where I believe you're asking to disconnect the two freeways' waypoints. This seems to be a not uncommon setup in China where you could (or they did) have a tollbooth between the freeways:

🔗 There are common exceptions to positioning at centerline crossings, such as interchanges where ramps connect nearby, non-intersecting highways, or where a short access road connects a road to another with a trumpet or similar interchange. In these cases, the waypoints for the same interchange on the separate highways cannot be at the same coordinates. Instead, the waypoints should be where the connecting ramps or access road interchange with each highway.

This reads to me the guidance is intended for a trumpet interchange with say a parallel, non-intersecting road, which wouldn't apply here:

• There is a clear centerline crossing between G60 and G1503.
• "The short access road" I think literally refers to a road, not just short merge section.
• The freeways literally go over each other, so they are intersecting, and have waypoints be at the same coordinates.

Perhaps my reading of the guidance is:

🔗 There are common exceptions to positioning at centerline crossings where the waypoints for the same interchange on the separate highways cannot be at the same coordinates, such as interchanges where ramps connect nearby, non-intersecting highways, or where a short access road connects a road to another with a trumpet or similar interchange. Instead, the waypoints should be where the connecting ramps or access road interchange with each highway.

(i.e., if they don't intersect, don't try to draw them as if they do?)

[Markkos1992]

The separate points for double trumpets is also done with many turnpikes in the US (PA Turnpike, NJ Turnpike, New York State Thruway, etc.) whether the roads cross or not.

[Nagamasa]

Ah I see! I don't travel much on the east coast, so I didn't notice that's how they're done everywhere. Perhaps I should fork this for a discussion to better understand why? I'm curious to understand why this was put in place, given the focus on networks and graphs, we have literally a whole class of interchanges and freeways that don't add to it!

[michih]

@nagamasa you've fixed some open issues for SH with your latest PR. Please let me know when all are fixed. Thanks.

[Nagamasa]

@nagamasa you've fixed some open issues for SH with your latest PR. Please let me know when all are fixed. Thanks.

I believe they should all be fixed at this point (for both chnsh and chng)

[michih]

What's the story behind your 'China Select Provincial Expressways' system you want to introduce today? Shouldn't we have one system per province like you already have in place for two provinces?

There were major changes in 2017 and 2022 that downloaded many freeways from national level to provincial level, and the ones added to chns were just the ones that already in chng in various parts of the country that were previously in chng, rather than leaving them in unprocessed wpts, given that there is already a substantial list of them already. They can be parked there for when we are ready to take a stab at those.

We've reviewed 0.5% of chng and chn**s--there's 180,000km of freeways in China, so chns makes these stragglers a little more documented. Basically usansf with Chinese characteristics.

Fine. Great move

[michih]

@nagamasa you've fixed some open issues for SH with your latest PR. Please let me know when all are fixed. Thanks.

I believe they should all be fixed at this point (for both chnsh and chng)

No. Missing implementation (or comment on this thread):

G15:
Move JS/SH to the borderline, apply to G4221 and JS's G15

G92:
Does it extend to S20 as indicated on OSM, concurrent with G60? GM does not indicate G92 at all.

G1503:
Add wp south of 132

I think that another G1503 in the NE should be added if OSM + GM right that it is in service and signed as such.

You've merged it into main route but removed the short additional G1503 segment between G1503's exit 204 and S20. you should add it again.

[Nagamasa]

@nagamasa you've fixed some open issues for SH with your latest PR. Please let me know when all are fixed. Thanks.

I believe they should all be fixed at this point (for both chnsh and chng)

No. Missing implementation (or comment on this thread):

G15:
Move JS/SH to the borderline, apply to G4221 and JS's G15

G92:
Does it extend to S20 as indicated on OSM, concurrent with G60? GM does not indicate G92 at all.

G1503:
Add wp south of 132

I think that another G1503 in the NE should be added if OSM + GM right that it is in service and signed as such.

You've merged it into main route but removed the short additional G1503 segment between G1503's exit 204 and S20. you should add it again.

G92 only goes to G1503: https://en.wikipedia.org/wiki/G92_Hangzhou_Bay_Ring_Expressway.
The old G1503 segment is no longer signed as such. https://www.bilibili.com/video/BV1sK411e7gY/?spm_id_from=333.337.search-card.all.click&vd_source=45addfe14efe619bf0f24730196bf14b

[michih]

G92 only goes to G1503: https://en.wikipedia.org/wiki/G92_Hangzhou_Bay_Ring_Expressway.

OSM has it for the whole length, wikipedia halfway, GM not at all. What source is actually the most reliable out of the unreliable source?
In the end, G92 is fully concurrent to G60. If we'd fully remove SH's G92, it would still be clinchable as G60. I'd tend to remove G92. Anyone?

The old G1503 segment is no longer signed as such. https://www.bilibili.com/video/BV1sK411e7gY/?spm_id_from=333.337.search-card.all.click&vd_source=45addfe14efe619bf0f24730196bf14b

Fine.

[Nagamasa]

G92 only goes to G1503: https://en.wikipedia.org/wiki/G92_Hangzhou_Bay_Ring_Expressway.

OSM has it for the whole length, wikipedia halfway, GM not at all. What source is actually the most reliable out of the unreliable source?
In the end, G92 is fully concurrent to G60. If we'd fully remove SH's G92, it would still be clinchable as G60. I'd tend to remove G92. Anyone?

Looks like it is signed to G1503:
WB 0:04@ https://www.bilibili.com/video/BV1S94y1Q79V/?vd_source=45addfe14efe619bf0f24730196bf14b
EB 2:51:26@ https://www.bilibili.com/video/BV1RJmiYQEcQ/?spm_id_from=333.337.search-card.all.click&vd_source=45addfe14efe619bf0f24730196bf14b

[Nagamasa]

BTW, I still don't feel comfortable seeing this system all the way to active. Data quality outside cities drop significantly, leading to significant research time.

As well, I believe only about half of the system is only mapped (which is already the size of usai), and basically the entire system as checked in needs to be re-edited.

[michih]

BTW, I still don't feel comfortable seeing this system all the way to active.

Ok.

Data quality outside cities drop significantly, leading to significant research time.

What kind of data are we talking about? Routing, numbering, exit numbers, missing interchanges on maps,...?

As well, I believe only about half of the system is only mapped (which is already the size of usai), and basically the entire system as checked in needs to be re-edited.

Oh, I wasn't aware that only half is mapped. That might change my mind....

[Nagamasa]

Mostly missing exit numbers, there's a lot of in the middle of nowhere.

Naming all the exits G123_Que knowing there's a "final form" that would require scrubbing a route again is not fun :/