What existed before the Routeing Guide?

[crablab]

As I understand it, the Routeing Guide was created at privatisation as part of the whole TSA and Peterborough Agreement, to find some sensible way to issue tickets across TOCs/TOUs, agree settlement etc.

I believe that the Routeing Guide itself wasn't always available from ATOC/RDG until they were made to publish it, but perhaps I'm wrong on that?

Presumably the Routeing Guide, as created, was based on some internal BR equivalent? Was that accessible to passengers?

I'm aware of the existence of the NFMs and of course these have direct equivalents today.

I'm also interested in whether the 'anomalous routes' exist purely as a result of privatisation codifying bad data.

 

[P Binnersley]

Before the Routing Guide, fares were "any reasonable route" (unless they were "via/not via" a named station). This was no formal description of what was or wasn't reasonable. If you could give a good reason (the "Man on the Clapham omnibus test") why you were travelling that route then you were OK. In practice most routes would be fine as long as you were heading in the the right direction and not going miles out of your way.

 

[crablab]

The status quo at the moment seems a considerable downgrade then, with passengers confused about whether tickets are or are not valid on different routes, especially in times of disruption (depending on ticket acceptance being agreed). Combined with TOC specific tickets and many more Advance fares...

That said, I don't think the open ended "any reasonable route" would have survived indefinitely, but certainly has some advantages.

 

[Bletchleyite]

If I recall it wasn't wholly open ended but rather in some ways more restrictive than now. For instance I recall if part of the route you wanted to take was more expensive you couldn't go that way. I don't have a copy of the rules but someone else might have?

 

[Merle Haggard]

Don't forget that in B.R. days all tickets were paper or card and carried routing usually in the form of 'via ...', an intermediate station on direct line of route.

For instance, a day return from Cardiff General to Barry had the words 'via Cogan or Lavernock', i.e., via either route of through trains. Tickets between Northampton (Castle) and Peterborough (East) were endorsed 'via Barnwell' to preclude travel via Rugby. With these endorsements it would be difficult to justify travelling by a more 'imaginative' route

 

[Haywain]

Don't forget that in B.R. days all tickets were paper or card and carried routing usually in the form of 'via ...', an intermediate station on direct line of route.

For instance, a day return from Cardiff General to Barry had the words 'via Cogan or Lavernock', i.e., via either route of through trains. Tickets between Northampton (Castle) and Peterborough (East) were endorsed 'via Barnwell' to preclude travel via Rugby. With these endorsements it would be difficult to justify travelling by a more 'imaginative' route

I think when people asked about BR days they mean the 1980s and 1990s, rather than the 1950s and 1960s!

 

[akm]

I'm also interested in whether the 'anomalous routes' exist purely as a result of privatisation codifying bad data.

I'd say it's more "bad codifying of what 'reasonable route' means". I'm not saying it's easy! But the first iterations of the Routeing Guide had massive holes (see Clive Feather's adventures!) and the current one certainly isn't perfect...

 

[Doctor Fegg]

I vaguely recall “any reasonable route” included all direct origin-destination services, as now, but also “offers a user the chance to make the journey at the time they want to make it”.

In other words, if you had a ticket from London Terminals to Manchester, and the next itinerary via Sheffield would get you there faster than a train from Euston, then you could use that itinerary.

 

[Merle Haggard]

I think when people asked about BR days they mean the 1980s and 1990s, rather than the 1950s and 1960s!

Yes, sorry, it's a fair cop! But there were a lot more alternative routes back in those days, and it didn't seem to be a problem. The travelling ticket collector (guards had no revenue duties) would just give a little sigh, clip the ticket and move on...

 

[Bletchleyite]

I'd say it's more "bad codifying of what 'reasonable route' means". I'm not saying it's easy! But the first iterations of the Routeing Guide had massive holes (see Clive Feather's adventures!) and the current one certainly isn't perfect...

I think the error was in trying to codify everything rather than making it recursive. Much simpler if, say, a London to Manchester ticket was just listed as valid via Chesterfield or Rugby, and then that'd recurse down to routes from London-Rugby + Rugby-Manchester etc.

 

[crablab]

Much simpler if, say, a London to Manchester ticket was just listed as valid via Chesterfield or Rugby, and then that'd recurse down to routes from London-Rugby + Rugby-Manchester etc.

Well the routing points are kind of recursive, if you're going via London or groups (eg. Birmingham Group).

I wonder if this was a deliberate technical choice to fit within the constraints they had at the time? Recursion can be complex to reason about and is memory expensive.

A simpler way of solving this might have been to just apply an upper bound on the distance you're permitted to travel to get between two points, which is kinda how the journey planners work I believe, and then apply any 'special restrictions' (eg. via/not via) on top of that if required. That would stop people doing stupidly indirect routes, but allow the longest 'reasonable route' and anything in between.

== Doublepost prevention - post automatically merged: 15 Dec 2022 ==

I think when people asked about BR days they mean the 1980s and 1990s, rather than the 1950s and 1960s!

Yes!

 

[JB_B]

I think the error was in trying to codify everything rather than making it recursive. Much simpler if, say, a London to Manchester ticket was just listed as valid via Chesterfield or Rugby, and then that'd recurse down to routes from London-Rugby + Rugby-Manchester etc.

Can you explain your stop condition for the recursion? How would you evaluate routeing for the next step - finding e.g permitted routes for the London-Rugby sub-leg?

If you applied your first step ( using the routeing guide as it is) you'd be massively reducing the available permitted routes - losing e.g. routes via the Thames Valley or Chiltern for London-Manchester.

 

[Bletchleyite]

Well the routing points are kind of recursive, if you're going via London or groups (eg. Birmingham Group).

I wonder if this was a deliberate technical choice to fit within the constraints they had at the time? Recursion can be complex to reason about and is memory expensive.

I suppose that was the issue - the need for it to be human readable. There is no longer such a need - almost nobody does that bar enthusiasts, you just ask a journey planner.

A simpler way of solving this might have been to just apply an upper bound on the distance you're permitted to travel to get between two points, which is kinda how the journey planners work I believe

Journey planners don't work like that, no. They apply the routeing guide, and it's NFM64 fares comparisons that prevent the use of silly routeing points.

 

[JB_B]

A simpler way of solving this might have been to just apply an upper bound on the distance you're permitted to travel to get between two points, which is kinda how the journey planners work I believe,

I've never heard of journey planners applying a limit based on distance but maybe they do - where did you hear that?

My guess would be that planners start by enumerating possible journeys just based on the timetable and then attempt to find optimum tickets (or combinations of tickets in the case of splitting sites) to match those journeys - starting with the quickest journey first. Matching tickets to journeys uses resources (particularly for splits and/or advances) so there's no reason for retail planners to explore extremely slow routes which will probably never result in a sale.

 

[Bletchleyite]

Can you explain your stop condition for the recursion? How would you evaluate routeing for the next step - finding e.g permitted routes for the London-Rugby sub-leg?

To follow one branch of the recursion (I'll follow the branch nearest Euston as I know that part of the network well), something like:

Manchester-London: via Rugby
Rugby-London: via Bletchley
Bletchley-London: via Watford Jn
Watford Jn-London: direct* [this is the stop condition - when the network provides only one option]

* Change/BoJ as you wish, I just mean there's only one route. Or potentially if you consider LO via Watford High St to be a separate one, you would have one more recursion, perhaps via Harrow. Same with Northampton.

Computers are quite good at recursion!

I have actually thought of trying to write this at some point just as a demonstrator of feasibility.

If you applied your first step ( using the routeing guide as it is) you'd be massively reducing the available permitted routes - losing e.g. routes via the Thames Valley or Chiltern for London-Manchester.

The example given was not intended to express the actual set of current Permitted Routes for that, it was just an example!

 

[crablab]

I've never heard of journey planners applying a limit based on distance but maybe they do - where did you hear that?

I have a feeling it's been the cause of some bugs in the past, where new stations have had 'zero distance' in the data and caused journey planners to do interesting things. I am probably misremembering.

It's certainly used in the TSA for calculating flows and who the fare-setter is, although that isn't directly relevant to Routeing.

(Edit: having thought harder, I believe the issues the mileage data causes are related to incorrect charging of fares, rather than Routeing persay)

Computers are quite good at recursion!

I have actually thought of trying to write this at some point just as a demonstrator of feasibility.

Yes! I think this would be an interesting project

 

[JB_B]

I have actually thought of trying to write this at some point just as a demonstrator of feasibility.

I think that is a great idea. As currently expressed it sounds horribly restrictive to me.

 

[Bletchleyite]

I think that is a great idea. As currently expressed it sounds horribly restrictive to me.

Quite the opposite. For instance, Birmingham to London is valid via Rugby, High Wycombe or Reading. But there are WCML tickets that are valid via Birmingham but not all three of those. This is silly; in a recursive system they would be.

 

[crablab]

@Bletchleyite how would you intend to express the valid routes in a file? Would you apply a limit to the depth of recursion? Otherwise you would end up with cycles.

eg. (KGX -> FPK -> ENC -> SVG -> FPK...)

A common approach would just be to calculate the shortest path applying an upper bound to the number of nodes you check, but that might be that you then never evaluate routes you do intend to be be valid.

It's also important to decide how you weight the paths between nodes. Distance? Time? (since the general shortest path would simply be the shortest path between nodes, which isn't necessarily practically correct)

And routes which require changes or cross-London would need to incur a penalty.

Very much a 1st year CS problem

 

[Bletchleyite]

@Bletchleyite how would you intend to express the valid routes in a file? Would you apply a limit to the depth of recursion? Otherwise you would end up with cycles.

New thread: https://www.railforums.co.uk/threads/recursive-routeing-guide.240812/

 

[Watershed]

As I understand it, the Routeing Guide was created at privatisation as part of the whole TSA and Peterborough Agreement, to find some sensible way to issue tickets across TOCs/TOUs, agree settlement etc.

I believe that the Routeing Guide itself wasn't always available from ATOC/RDG until they were made to publish it, but perhaps I'm wrong on that?

Presumably the Routeing Guide, as created, was based on some internal BR equivalent? Was that accessible to passengers?

I'm aware of the existence of the NFMs and of course these have direct equivalents today.

I'm also interested in whether the 'anomalous routes' exist purely as a result of privatisation codifying bad data.

Fundamentally, there isn't a published predecessor to the Routeing Guide. It was only created on privatisation, as before that travel was permitted by "any reasonable route" under the BR Conditions of Carriage (as remains the case with paper Underground tickets).

The creation of the Routeing Guide was partly a result of the need to regulate the routes offered by the electronic journey planners and booking engines, which were just starting to be rolled out around the time of privatisation.

 

[xotGD]

What existed before the routing guide? Common sense.

 

[Surreytraveller]

Fundamentally, there isn't a published predecessor to the Routeing Guide. It was only created on privatisation, as before that travel was permitted by "any reasonable route" under the BR Conditions of Carriage (as remains the case with paper Underground tickets).

The creation of the Routeing Guide was partly a result of the need to regulate the routes offered by the electronic journey planners and booking engines, which were just starting to be rolled out around the time of privatisation.

Were there electronic journey planners when the Routeing Guide was invented?
When I went on the training course before it was introduced, it was explained the need for it was so that tickets were only valid on the trains of operators that would receive a portion of the money from those tickets

 

[jfollows]

What existed before the routing guide? Common sense.

Essentially, yes.
I once tried to return from Glasgow to Berwick on a day return ticket via Carlisle and Newcastle, the guard essentially said "come off it, that's not sensible" and so we paid up. In fact we paid for a day return Glasgow-Carlisle and went back to Glasgow before going back to Berwick on a more sensible route. No complaints by anyone.

 

[thedbdiboy]

Essentially, yes.
I once tried to return from Glasgow to Berwick on a day return ticket via Carlisle and Newcastle, the guard essentially said "come off it, that's not sensible" and so we paid up. In fact we paid for a day return Glasgow-Carlisle and went back to Glasgow before going back to Berwick on a more sensible route. No complaints by anyone.

Indeed, essentially it was 'use any route as long as you don't take the p*ss'. Which, given the games played over the last 28 years over finding loopholes and errors in the routeing guide, is entirely the opposite of the way it seems to work now.
The original ideology of privatisation, i.e. separate TOCs in competition with each other, quickly foundered in terms of getting the legislation through parliament, which is why through ticketing and interavailability were glued on top of it. Trying to simultaneously achieve both outcomes is what gave us such monstrosities as the routeing guide. As we slowly drift back to a centrally managed network it will hopefully eventually be put out of its misery and replaced by a joined up journey planning process that links fares with service validity.
That is the logical and best solution for ordinary people who simply want to make a journey and find out what ticket they need but I am aware will be unpopular with certain residents of this parish as it will take a lot of the fun out of ticket buying

 

[alistairlees]

The correct name of the Routeing Guide is in fact the “electronic National Routeing Guide” - or eNRG. It was coded by Fujitsu in the mid 1990s following discussions with BR pricing managers. Somewhere I have got the original specification - I think from 1995. As has been observed, it was intended to ensure that carriers did not end up being forced to carry the passengers of ‘other’ TOCs. It predates online journey planning, but could (I assume) have been used locally in travel centres and booking offices equipped with Journey Planner (the Fujitsu software product that showed what itineraries were or were not valid for a given ticket).

 

[Ashley Hill]

If a person took a logically reasonable route and could justify it there was no problem. If they were going via London when their ticket said via Not London then they were excessed. And so on…
When issued I never really understood the routing guide as it over complicated things. The day it fell out of my locker and spread itself over the floor was the day it was scooped up and placed in the bin.

 

[sheff1]

I believe that the Routeing Guide itself wasn't always available from ATOC/RDG until they were made to publish it, but perhaps I'm wrong on that?

It was available to purchase by members of the public but that fact was not widely publicised.

Re 'reasonable routes' under BR, the situation was as described above but (in the early 70s at least) there was also an internal document which listed some routes which might not be considered reasonable at first glance but which were specifically allowed. A number of these involved a double back to travel on a sleeper service but a local one I remember was that Birmingham to Leicester (and v/v) was valid via Derby. The latter is not allowed by the Routeing Guide.

 

[Watershed]

Were there electronic journey planners when the Routeing Guide was invented?

Journey planning software existed (I believe it's been a part of TRUST, and previously TOPS, for years). But I don't think there was anything that was publicly available, certainly not something that would sell you a ticket, until after the Routeing Guide came along.

The existence of the Routeing Guide (or at least, some other formalised definition of what the permitted routes are) was always going to be a prerequisite to having online ticket sales.

When I went on the training course before it was introduced, it was explained the need for it was so that tickets were only valid on the trains of operators that would receive a portion of the money from those tickets

I'm not convinced by this argument. ORCATS allocates revenue primarily based on the fastest journey opportunities (it also takes into account a few other factors such as overtaken itineraries with fewer changes).

Nowhere do the permitted routes come into those calculations - indeed, ORCATS was created way before the Routeing Guide came about, and hasn't (to my knowledge) been changed to take permitted routes into account.

It's thus possible to travel on an operator who receives no revenue at all from your ticket, if they operate services along a permitted route but these don't feature in any of the fastest journey opportunities.

The inverse is also possible, as an operator may receive revenue from your ticket, even though you would have to be using a non-permitted route to travel on their services. This could for example be the case with last itineraries of the day, which often involve non-permitted routes.

 

[plugwash]

Were there electronic journey planners when the Routeing Guide was invented?
When I went on the training course before it was introduced, it was explained the need for it was so that tickets were only valid on the trains of operators that would receive a portion of the money from those tickets

AIUI that is not true though, IIRC revenue distribution is based on non-overtaken journeys, so tickets are frequently valid on operators who will receive none of the revenue.