IND-AH, IND-NH, IND-xx: India

[Bickendan]

Fixed locally, though the SH files are still very quick and dirty.

[michih]

I think that we face the same problem like we have in Japan. There are more than 500 NMP pairs right now because the route wps are not synced (unlike the French NMPs which are mostly FPs).

[si404]

I think that we face the same problem like we have in Japan. There are more than 500 NMP pairs right now because the route wps are not synced (unlike the French NMPs which are mostly FPs).

That isn't, AFAICS, an issue, unless we're talking about within the rework, rather than just between the rework and my initial drafts.

[Bickendan]

I think it's the initial AH and Select Named Expressways drafts coexisting with the revised IND-AH and overlayed IND-NH (and IND-WB overlaps) that are causing all of these.

[Bickendan]

I need the Union Territory abbreviation for Jammu And Kashmir as it's not in the Asia-Pacific Master Thread.

I assume it will be indjk?

[michih]

I need the Union Territory abbreviation for Jammu And Kashmir as it's not in the Asia-Pacific Master Thread.

I assume it will be indjk?

Yep: https://github.com/TravelMapping/HighwayData/blob/master/regions.csv#L240

[Bickendan]

Drafted out NH 44, which runs north-south from the southern tip of India to Srinigar in Jammu And Kashmir.

As it's unclear whether NHs are being routed through cities or onto bypasses, I'm doing split alignments ala I-35 DFW/MSP, as 'NH 44-E' and 'NH 44-W'.

Extended NH 19's draft from Agra to Delhi at NH 48. I'll look into its city/bypass routings and make 19-S and 19-N files as needed.

[Bickendan]

Drafted out UP 39 and 62 (as I've physically been on portions of them in Agra), as well as NH 21.
Split WB 1 into WB 1-E and -W in Kolkata.

Next, drafting out NH 48, which will complete the Golden Quadralateral, the Maa Flyover in Kolkata, and poke at a couple expressways in Delhi.

The good news is it actually appears that GMSV is appearing in India now. The bad news is coverage isn't comprehensive yet, and isn't helpful in the sense of confirming routing or street names (matching on the ground experience).

[Bickendan]

Checking the uploads...
Looks like that the split alignments of NH 44, 48, and WB 1 are throwing invalid parameter errors when attempting to directly view the routes; they do show if loaded as part of the connected route.
(e.g, NH 44-W and 44-E). I'm assuming the hyphen in either the filename or as the route name in the .csv is causing the error, whereas the _con.csv doesn't have much issue with it.
­·WB 1-E loads as part of the WB 1, 1-E, 1bar complex. Does not show on its own. WB 1-W is not accessible.
·Among the errors surrounding NH 44, both 44-W and 44-E in Nagpur (Maharashtra), Seoni and Dabra (Mahdya Pradesh), and Agra (Uttar Pradesh) load in as part of the Connected Route.
·NH 44's connected segments are being treated as two separate entities - Tamil Nadu through the Gwalior segment of Madhya Pradesh, with the southern Maharashtra segment missing;
and the Agra portion of Uttar Pradesh through Punjab.
Missing segments appear to be the southern Maharshtra segment, the Rajasthan segment, and the Jammu-and-Kashmir segment. I suspect .csv and _con.csv errors.

NH 48's abbreviation for Maharashtra is wrong. Used MA instead of MH. It loaded anyway (as MH NH48 -- if you fixed that, Si, thank you). All labels in the .wpt will need to be corrected.

---
It looks like GMSV is slowly rolling out for India. From what I've seen, it hasn't been helpful in figuring out the names for various points. However, my current plan is moreless to get each state to a presentable state and get them into preview status that way (ie, clean up West Bengal's AH, NE, NH, and SH and bring them up from Devel to Preview that way -- the NH system on its own is too massive to get it all up in one go).

[si404]

Missing segments appear to be the southern Maharshtra segment, the Rajasthan segment, and the Jammu-and-Kashmir segment. I suspect .csv and _con.csv errors.

The .csv errors here were that they were looking for files that weren't there - I actively removed those bits from the .csvs as it was easier than making .wpt files for the missing segments.

NH 48's abbreviation for Maharashtra is wrong. Used MA instead of MH. It loaded anyway (as MH NH48 -- if you fixed that, Si, thank you). All labels in the .wpt will need to be corrected.

I did fix it, as it threw up an error. I fixed the border points (including on the adjacent routes) but not the State Highway labels.

[Bickendan]

Found an error in my local .csv for NH 44 -- mistyped indmh.nh044 as indup.nh044
Fixed, hopefully that will unite the two disparate connected NH 44s into a single route.
It looks like I actually failed to actually create the .wpts for indjk.nh044 and indrj.nh044 and copy/paste the waypoints from the editor to my text file. sigh.

indma.nh044 -> indmh.nh044 and all relevant tags in the .wpt converted from MA to MH. Both .csvs also corrected.

[yakra]

Looks like that the split alignments of NH 44, 48, and WB 1 are throwing invalid parameter errors when attempting to directly view the routes; they do show if loaded as part of the connected route.
(e.g, NH 44-W and 44-E). I'm assuming the hyphen in either the filename or as the route name in the .csv is causing the error, whereas the _con.csv doesn't have much issue with it.
­·WB 1-E loads as part of the WB 1, 1-E, 1bar complex. Does not show on its own. WB 1-W is not accessible.
·Among the errors surrounding NH 44, both 44-W and 44-E in Nagpur (Maharashtra), Seoni and Dabra (Mahdya Pradesh), and Agra (Uttar Pradesh) load in as part of the Connected Route.

I assume the problem is showroute can't cope with the hyphen in the root name. There are similar problems in Brazil with braspa & braspi.
These look like I-35E/I-35W scenarios; I recommend just removing the hyphens from the roots.

Good catch on the routes being visible as part of a ConnectedRoute; I didn't think to try that.
Which brings us to another problem...
We can't have both sides of the E/W split alignments in a ConnectedRoute.
For a .list line like

Code: [Select]

IND-UP NH44jaj RJ/UP IND-UP NH44mat NH530Bthe system won't know whether to make the user's connection via the E or W branch.

In practice, with the way the ConnectedRoute is entered now, what will happen is:
• The first .listed chopped route (NH44jaj) gets marked as traveled from the .listed point (RJ/UP) to its end.
• The last .listed chopped route (NH44mat) gets marked as traveled from its start to the .listed point (NH530B).
• All intermediate chopped routes (NH44-E and NH44-W) get marked in their entirety.
These _con.csv entries will have to be split into multiple connected routes for the main route designation, with separate entries for the E & W branches.
Unfortunately, there will be quite a few.

This brings us to an even bigger problem...
I'd expect that this would be throwing up DISCONNECTED_ROUTE errors all over the place.
But it's not. There are only 2 outside the US.
This means there's a bug in the datacheck routines and/or a scenario in the underlying highway data that I didn't anticipate. The debug session for this may be a pretty deep dive, one I don't have the time to get into right away.

https://github.com/TravelMapping/DataProcessing/issues/610

[Bickendan]

I have a large update ready for upload, with all of the NH routes for Uttar Pradesh and Karnataka (and their segments in all relevant states/union territories), along with a handful of Karnataka's state highways. Who should I send the .zip to?

[si404]

I can do it