Author Topic: ENG: M11 issues  (Read 164 times)

0 Members and 1 Guest are viewing this topic.

Offline rickmastfan67

  • TM Collaborator (A)
  • Hero Member
  • *****
  • Posts: 1386
  • Gender: Male
  • Last Login:September 19, 2022, 09:04:05 am
ENG: M11 issues
« on: July 17, 2022, 11:28:28 am »
Exit #7A needs to be re-centered.  Seems it was added when it was in the incorrect place on OSM.

Also, Exit #14 has been completely reconfigured and needs to be updated on the site, due to M11/A14 split being farther north now.  Plus the location for A1307 inside of that interchange has changed dramatically, and honestly, should be a completely separate point, due to 0 connection with A28.
« Last Edit: July 17, 2022, 02:24:28 pm by michih »

Offline si404

  • TM Collaborator
  • Hero Member
  • *****
  • Posts: 1682
  • Last Login:Today at 04:05:41 am
Re: UK: M11 issues
« Reply #1 on: July 17, 2022, 01:30:50 pm »
Exit #7A needs to be re-centered.  Seems it was added when it was in the incorrect place on OSM.
Fair enough - the siting of the point dates to when the interchange was marked as under-construction.
Quote
Also, Exit #14 has been completely reconfigured and needs to be updated on the site, due to M11/A14 split being farther north now.
Not true (look at Stamen layers, which show the old layout).

And, even if that was the case, the logical centre point of the interchange hasn't changed just because the East->North slip at that junction has changed from being a loop to being a directional connection.
Quote
Plus the location for A1307 inside of that interchange has changed dramatically, and honestly, should be a completely separate point, due to 0 connection with A28.
Other than having a new extension to the west, it's pretty much where it always was. It never connected to the A428 (though it did lose a connection to the M11 south).

There's a case for two or three points for two or three interchanges (or one interchange that stretches too far) - M11/A428 <-> A14_E, A1307 and M11 <-> A14_W (and I wouldn't be adverse to being persuaded that that is the case), as there was before - but the one interchange with one point still has a good case. Nothing has changed.

Offline rickmastfan67

  • TM Collaborator (A)
  • Hero Member
  • *****
  • Posts: 1386
  • Gender: Male
  • Last Login:September 19, 2022, 09:04:05 am
Re: UK: M11 issues
« Reply #2 on: July 17, 2022, 11:56:14 pm »
Quote
Plus the location for A1307 inside of that interchange has changed dramatically, and honestly, should be a completely separate point, due to 0 connection with A28.
Other than having a new extension to the west, it's pretty much where it always was. It never connected to the A428 (though it did lose a connection to the M11 south).

There's a case for two or three points for two or three interchanges (or one interchange that stretches too far) - M11/A428 <-> A14_E, A1307 and M11 <-> A14_W (and I wouldn't be adverse to being persuaded that that is the case), as there was before - but the one interchange with one point still has a good case. Nothing has changed.

Hmmm, still honestly think a separate point for A1307 on at least A14 is justified here.  Mainly because of this following rule:
Quote
🔗 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.

Pretty clear justification for a '31A' in A14, and removal of the unjustified graph connections for A1307 with the other routes there.
Code: [Select]
31A http://www.openstreetmap.org/?lat=52.233671&lon=0.067570
31 http://www.openstreetmap.org/?lat=52.230625&lon=0.070810

This is no different than how I-85/I-73/I-85 Business are dealt with in NC at the SW corner of the Greensboro Loop with all the separate routes there.  They are all mapped where the proper center points are.

Offline si404

  • TM Collaborator
  • Hero Member
  • *****
  • Posts: 1682
  • Last Login:Today at 04:05:41 am
Re: UK: M11 issues
« Reply #3 on: July 18, 2022, 06:33:35 am »
the unjustified graph connections for A1307 with the other routes there.
The A428 only has access to the A14 to the east, just as the A1307 only has access to the A14 to the north. Both roads only connect to the A14, and in one direction only. 'Unjustified graph connections' (and this seems to be the root of your objection) are not a justifiable reason for changing the graphing, as it would be inconsistent tweaks.

And general practise is not to care about limited access junctions mucking stuff up - eg I was asked by more than one person to merge M60 J14 and M61 J1 into one point off the A580 rather than have two very close points with access between A580 and respective motorway at each, but not having the two motorways meet to avoid the unjustified graph connection.

Quote
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.
The first word there is key: "usually".

And the final rule of thumb is "Some interchanges have unusual shapes or are stretched-out versions of normal interchanges. Use your best judgment.". Here I've used my best judgement to say that it's one interchange and that the centre of the interchange is where the centrelines of the E-W and N-S major roadways cross. That the unusual shape and stretched-out ness means that the other roadway's centreline doesn't cross that point is neither here nor there.

This is no different than how I-85/I-73/I-85 Business are dealt with in NC at the SW corner of the Greensboro Loop with all the separate routes there.  They are all mapped where the proper center points are.
But, that can very clearly be treated as 1PPI. In fact, there's a similar example used to explain 1PPI:
Quote
As an example, imagine a cloverleaf interchange where one loop ramp was replaced by a flyover ramp. Consider the I-97 & MD 3 & MD 32 interchange.
The point should be centered in the middle, and the same point should be used for I-97, MD 3, and MD 32.. The centerlines are between the NW-SE and the NE-SW through lanes. One might have expected the point for I-97's file to be on the apparent SE-NE mainline, or even two points where MD 32 connects at the SE and MD 3 at the NE. But it is one interchange, and the point for all three routes should go in the middle.

It's purely 'best judgement' on the 'stretched out' justification that means that it isn't blatantly wrong to have three points there. That the exits have different numbers, and it's rather spread out, means that Greensboro has good justification for multiple points. Girton, not so much - it's smaller and has the same exit number. Both 1PPI and multiple points can be justified at both, but my 'best judgement' (which matters for Girton, but not Greensboro as I don't maintain it), is that they sit on opposite sides of which is more justifiable.
  • ENG A14 31a (its 31, but the 31 label is taken. It's definitely not '31A', which would be a different junction) <-> 31 0.26 mi
  • NC I-73 97A <-> 97B 0.30 mi
  • NC I-85BLGre 33 <-> 33A 0.41 mi
  • NC I-85 120 <-> 121 0.58 mi

Girton is, in my best judgement, 1PPI. No change.
« Last Edit: July 18, 2022, 06:36:36 am by si404 »