On Feb 20, 11:03 pm, "John Rowland"
wrote:
wrote:
On Feb 20, 2:40 pm, "John Rowland"
wrote:

I'd like bus shelters to bear accurate mathematically distorted
geographical maps, where, for instance, distance from the centre of
the map is proportional to the square root of the actual distance on
the ground, and any super-long routes have an arrow at the edge of
the map listing further destinations.

How do you handle the case where two bus routes share the same stretch
of road, diverge, and then rejoin at another point?

You either give them different colours, or you show them as separating and
converging like the Northern Line - whichever makes the map easier to
understand (which would depend on what else was going on in the map).

But then this would destroy the accurate mathematical distortion
property.

--
Abi

wrote:
On Feb 20, 11:03 pm, "John Rowland"
wrote:
wrote:
On Feb 20, 2:40 pm, "John Rowland"
wrote:

I'd like bus shelters to bear accurate mathematically distorted
geographical maps, where, for instance, distance from the centre of
the map is proportional to the square root of the actual distance
on the ground, and any super-long routes have an arrow at the edge
of the map listing further destinations.

How do you handle the case where two bus routes share the same
stretch of road, diverge, and then rejoin at another point?

You either give them different colours, or you show them as
separating and converging like the Northern Line - whichever makes
the map easier to understand (which would depend on what else was
going on in the map).

But then this would destroy the accurate mathematical distortion
property.

Even Ordnance Survey 1:50000 maps contain local distortions to cope with the
fact that roads are shown wider than they really are.

On Wed, 21 Feb 2007, John Rowland wrote:

wrote:
On Feb 20, 11:03 pm, "John Rowland"
wrote:
wrote:
On Feb 20, 2:40 pm, "John Rowland"
wrote:

I'd like bus shelters to bear accurate mathematically distorted
geographical maps, where, for instance, distance from the centre of
the map is proportional to the square root of the actual distance
on the ground, and any super-long routes have an arrow at the edge
of the map listing further destinations.

How do you handle the case where two bus routes share the same
stretch of road, diverge, and then rejoin at another point?

You either give them different colours, or you show them as
separating and converging like the Northern Line - whichever makes
the map easier to understand (which would depend on what else was
going on in the map).

But then this would destroy the accurate mathematical distortion
property.

Even Ordnance Survey 1:50000 maps contain local distortions to cope with the
fact that roads are shown wider than they really are.

As does the A-Z - i cycled from Kilburn to Finsbury Park last night, along
a route i've never used before, with only the A-Z-derived London Cycle
Guide for company, and came across a number of junctions where the map
indicated i should be able to go straight over but actually had to make a
substantial dogleg, or turn down one road before i could turn onto the one
i wanted, etc.

tom

--
A complex system that works is invariably found to have evolved from a
simple system that worked. -- Gall's Law

On Wed, 21 Feb 2007, James Farrar wrote:

On Tue, 20 Feb 2007 09:21:58 +0000, Tom Anderson
wrote:

What's the typical deliver size? Or rather, what weight would you say 80%
of deliveries are smaller than or equal to? Would it be small enough to do
by bike (using a freight bike of some sort, rather than a courier's
panniers)?

Most jobs go in (as a minimum) an A4 box (i.e. a box that would contain
five reams of A4 80gsm.

That kind of thing would easily be doable by freight bike - may i ask how
you deliver it at the moment? Presumably you don't have a van trip for
each delivery; pile multiple deliveries into one van and go on a tour
round all the customers?

Not to mention large scale deliveries. As I was leaving work this
morning we had 50+ reams of paper turn up. How are they supposed to
deliver that without a lorry?

I assume you get your paper in quite big sheets - 50 reams of A4 at 80 gsm
is 125 kg, doable on a trike or 8-freight or something. If it's A0,
though, that's two tonnes, which i would certainly agree requires motor
power!

Actually, I got that wrong. It was 20 boxes of A4 which is 100 reams.
Plus some A3 and other stuff.

A0 paper comes in rolls - 200m long, works out to maybe 10cm across at a
guess. 24 rolls on a pallet.

Okay, any of that'd be rather hard to shift by bike, i think. Although i
did start wondering if you could build a tandem derivative of an 8-freight
(a 16-freight?) which could handle a pallet - a problem to leave to a real
engineer, i think!

For those of you not familiar with the 8-freight:

http://www.velovision.com/mag/issue9/8freight.pdf

Funnily enough, it actually has been used to transport boxes of print -
ten boxes, 100 kg "quite happily".

I should add that i'm not seriously suggesting you replace vans with
bikes; i'm just interested in working out to what extent bikes could
replace motor vehicles for goods traffic in a real-world situation.

Also, bear in mind that the OP was only proposing closing one route to
cars; all you'd really have to do was pile everything into a hand-cart and
wheel it a few hundred metres to the nearest motor-accessible road!

tom

--
A complex system that works is invariably found to have evolved from a
simple system that worked. -- Gall's Law

On Tue, 20 Feb 2007, John Rowland wrote:

Dave A wrote:

The problem with that is that where there are long routes that can be
shrunk in a spider diagram but will not fit into a traditional map -
this is the case for many routes on the central London traditional
map. The most useful connections will be those outside central London,
which wouldn't be represented by the map I describe.

I (still) hate the spider maps.

I'd like bus shelters to bear accurate mathematically distorted
geographical maps, where, for instance, distance from the centre of the
map is proportional to the square root of the actual distance on the
ground,

Nah, logarithm.

and any super-long routes have an arrow at the edge of the map listing
further destinations. Each group of routes which serve the same local
stops would be shown as a single coloured line, which then branches into
the different routes towards the edge of the map....

http://www.openstreetmap.org/
http://www.tfl.gov.uk/buses/spiders/
http://www.gimp.org/

Get busy!

tom

--
A complex system that works is invariably found to have evolved from a
simple system that worked. -- Gall's Law

On Tue, 20 Feb 2007 wrote:

On 20 Feb, 01:05, Tom Anderson wrote:
On Mon, 19 Feb 2007, Dave A wrote:
Paul Corfield wrote:
On Mon, 19 Feb 2007 08:59:02 -0500, David of Broadway
wrote:

I will say that your spider maps are much easier to read and much more
useful than the maps we have posted at bus stops.

They are fine if there is a direct bus from the stop you are standing
at. They are hopeless if your journey requires interchange to another
service at some point. There is no sense of there being a network with
spider maps which I believe is counterproductive when you have a network
which is as dense as London's and where the move to shorter routes over
the last 4 decades means changing services is much more of a necessity.
There is little to guide people as to how to accomplish such journeys if
they are relatively unfamiliar with the bus network.

The only way I can think of to make a clear diagram like this is to
combine the spider and the traditional bus map - by using the
traditional map as a base, and overlaying buses from the current
location as individual coloured lines.

How about annotating the spiders to show interchange points, as on the
tube strip maps? So, for instance, on the Finsbury Park spider, the
Holloway Nag's Head stop on the 29/253/etc bundle would have a little box
saying "4 17 43 271 393", maybe with arrows pointing away on either side
labelled "Archway" and "Highbury & Islington" (or something, since not all
those routes go those ways). It wouldn't completely solve the problem, but
if you were at A, wanted to go to B, and knew what the routes serving B
were, you could look for a suitable C on the spider map at A. Even if you
didn't know the routes at B, you could perhaps make a reasonable guess
based on the destination hints. The key problem would probably be the
sheer number of boxes and arrows - there are a *lot* of routes in London!

Absolutely, but not only are there buses, but also tubes and trains and
streets too.

None of which i'd show. Well, tubes and trains would get an icon, as on
current spider maps, but that's easy; i wouldn't bother trying to show the
railway lines on the bus map. Not terribly integrated, i know, but i think
the railway lines are well enough mapped (and well enough generally known)
that they don't need to be laid out on the bus map.

You can't easily mix tube style diagrams (as seen in NYC or with spider
maps from TfL in London) with tubes or trains which also use this type
of diagram. theres only so many colours so spider diags are local
before all the colours are used. Thats why the bus map (Quickmap) being
discussed here is so useful/different. Getting all london on one sheet
is mega difficult but if you've got one in your pocket it allows you to
always get off the tube (when its not working) and immediately onto a
bus going in the right sort of direction.

True - i'm talking about an improved local map, rather than a new global
map. To be honest, i think the London bus network is complicated enough
that a map of the whole thing just isn't going to be generally useful.

tom

--
A complex system that works is invariably found to have evolved from a
simple system that worked. -- Gall's Law

On Wed, 21 Feb 2007 15:44:20 +0000, Tom Anderson
wrote:

On Wed, 21 Feb 2007, James Farrar wrote:

On Tue, 20 Feb 2007 09:21:58 +0000, Tom Anderson
wrote:

What's the typical deliver size? Or rather, what weight would you say 80%
of deliveries are smaller than or equal to? Would it be small enough to do
by bike (using a freight bike of some sort, rather than a courier's
panniers)?

Most jobs go in (as a minimum) an A4 box (i.e. a box that would contain
five reams of A4 80gsm.

That kind of thing would easily be doable by freight bike - may i ask how
you deliver it at the moment? Presumably you don't have a van trip for
each delivery; pile multiple deliveries into one van and go on a tour
round all the customers?

Depends on distance and deadline. Sometimes, yes.

Also, bear in mind that the OP was only proposing closing one route to
cars; all you'd really have to do was pile everything into a hand-cart and
wheel it a few hundred metres to the nearest motor-accessible road!

Parking the vans is bad enough as it is!

Paul Corfield wrote:
On Tue, 20 Feb 2007 23:13:13 +0000, Dave A wrote:

Tom Anderson wrote:
On Mon, 19 Feb 2007, Dave A wrote:

Paul Corfield wrote:
On Mon, 19 Feb 2007 08:59:02 -0500, David of Broadway
wrote:

I will say that your spider maps are much easier to read and much
more useful than the maps we have posted at bus stops.
They are fine if there is a direct bus from the stop you are standing
at. They are hopeless if your journey requires interchange to another
service at some point. There is no sense of there being a network with
spider maps which I believe is counterproductive when you have a network
which is as dense as London's and where the move to shorter routes over
the last 4 decades means changing services is much more of a necessity.
There is little to guide people as to how to accomplish such journeys if
they are relatively unfamiliar with the bus network.
My impression of bus use in London is that it is broadly confined to
the use of single routes from origin to destination - ISTR a statistic
that only 4% of journeys involving buses, involved changing from one
bus to another.
Any idea if that includes night buses? I can almost never get home in
the wee small hours without changing.
On further inspection, it looks like I was lying my face off. The figure
I quoted is for all bus journeys in Great Britain. In London, it looks
like the figure is nearer 20%, which surprises me.

Source: TfL Interchange Plan (2002), Para 2.19 (primary source was
London Transport Planning in 1997)
http://cache.tfl.gov.uk/tfl/pdfdocs/inter_improve.pdf

Thanks for owning up ;-)

In my own experience I have to change buses quite a lot to make any
number of even quite local journeys. It is impossible for me to reach
the central area from where I live without changing buses - admittedly
only one change gets me onto a good spread of radial routes into zone 1.

My observations would also suggest that substantial volumes of people do
change buses in order to make their journeys despite the relative
richness of London's bus network. The easy availability of capped bus
fares via Oyster PAYG may inadvertently encourage this trend as would
the introduction of transfer tickets offering discounts. One simple
example of the extent of transfer between services is somewhere like
Silver St in Edmonton. Large numbers of people get off route 34 (east -
west) to change onto north-south routes at this point. This pattern is
repeated all over London. I'm actually surprised the figure is as low
as 20%.

I think that for many locations in inner London, changing bus is often
just not necessary, as the spread of routes from most inner London hubs
covers most of the surrounding area. Saying that, I should have done a
common sense check before quoting a silly statistic - there are
stampedes in the morning in Shepherd's Bush when a 207 unloads behind a
loading 49.

(Incidentally, I was very bemused to find that the 49 was receiving no
service enhancement at all from the C-charge extension, unlike
practically every other route in the area, and despite being completely
full in the morning peak from Shepherd's Bush and often skipping all
stops on Holland Road. The C1 extension is welcome, but it was never
intended to be routed along Holland Road, and it's hopelessly slow
around Earl's Court.)

Perhaps it would be better to limit it to important destinations which
are reachable by bus within a practical time frame - say an hour
(average journey time to work for those travelling by bus is 39 mins
across London; 47 mins in central London). For example, from Notting
Hill Gate there are 10 daytime bus routes covering most destinations
reachable within an hour by bus from there, except a few which could be
noted in the way you suggest - e.g. Clapham Junction, Barnes, Holborn.

The problem with your suggestion is that it relies on various parameters
that have different meanings to people. What is an "important
destination"? The destination for each individual passenger is
"important" to them and an awful lot of maps will not show such places -
especially if a change of bus is needed.

Certainly true - but then the opposite is also true, where you can show
everywhere and end up with a map that confuses most for the benefit of a
few (which was the reason for switching to spider maps anyway).

What is a practical time frame? - this must vary depending on whether
you are time rich or time poor as well as the activity that you will do
when you complete your journey. Finally a time based parameter will
unravel given the variability in journey times over the operating day
and it again does not deal with peoples' willingness to travel for a
long period by bus if overall they consider the bus to the best mode for
them given other factors like affordability.

Very true. However, I would suggest that people who are set on using the
bus for affordability reasons may also be more willing to work out
routes for themselves - that's my experience from students, at least.

I sometimes travel by bus even though "logic" would dictate that the
tube or a train would be more "sensible".

Snap. Often I just like the view out the window!

--
Dave Arquati
www.alwaystouchout.com - Transport projects in London

(Responding to an old post...)

Paul Corfield wrote:
On Mon, 19 Feb 2007 23:57:05 -0500, David of Broadway
wrote:

Although London's rail network has pretty wide coverage, it has limited
capacity in comparison to NYC's. Our trains are wider and longer and
most of our major trunk lines (and some of the minor ones, too) have
four tracks. Given how crowded our trains get, if we had to give up our
express tracks and shorten and narrow the trains, the buses would become
a lot more popular, by necessity.

I'm a tad taken aback by your comments on the relative capacities of
London's rail network vs NYC's. Now I'm certainly not an expert on your
subway or rail network but surely your rail network (not subway) is but
a mere shadow of London's?

By "rail" I was including subway/Underground.

But, including everything, you still might be right. It's difficult for
an outsider to get a good sense of your rail network.

From memory PATH is only twin bore into both WTC (as was) and 33rd
Street. Metro North is twin bore into Grand Central or is that 4 tracks?

Metro-North has four tracks along Park Avenue.

I think that LIRR and NJT into Penn Station is 4 tracks under the river.

NJT has only two, shared with Amtrak.

Here's a track map of the rail systems in the NYC region, excluding the
subway, PATH, and everything west of the Hudson:
http://www.richegreen.com/NYCTrackMapV3.pdf (an 8.7 MB file)

Now OK some of your trains are pretty long but overall frequencies and
distances covered are nothing like the density of service that we have
on networks like Southern, South West Trains or One from Liverpool
Street. I've observed Grand Central and Penn Stations in the rush hour
and certainly large volumes of people are shifted but it didn't feel on
the same scale as London's main line networks.

From the little I've seen in London, you're surely correct. Victoria
may be no busier than Penn, but we only have one other station similar
to Penn while you have lots more similar to Victoria!

I confess I don't know how many people are carried on LIRR lines that
terminate in Queens and Brooklyn.

Nor do I, I'm embarrassed to say. Since fares are the same to LIC and
Brooklyn as to Penn, and more trains go to Penn than to LIC or Brooklyn,
many passengers who might find the LIC and Brooklyn terminals useful
(it's easy to get to East Midtown from LIC and to Lower Manhattan from
Brooklyn) go to Penn anyway.

On the subway you do have much longer and bigger profile (than our tube
stock) trains and the benefit of express lines. In my (albeit limited)
experience of the NYC rush hour you get pretty high frequencies on
common sections of route served by multiple services but if you want a
particular letter / number then frequency drops noticeably compared to
almost all of London's tube service pattern.

Yes and no.

Where we have multiple services sharing a trunk, you have a single
service with multiple branches. I would argue that a single service in
NYC is more akin to a single branch in London. For instance, the
Central line would probably be treated as two routes in NYC, and the
Metropolitan would be at least three. The Northern would probably be
four, assuming current service patterns (which somewhat resemble our
service patterns on the 2 and 5, except that no 2's go to Dyre and only
a handful of select rush hour 5's go up the branch to 241; our south end
is a bit more complicated, with off-peak 5's terminating at Bowling
Green and lots of rush hour 5's going to Utica or New Lots rather than
Flatbush). And don't get me started on the District!

That said, your trains do tend to be more frequent than ours, it seems.
Especially off-peak. And some of our branches are fairly infrequent
-- I ride the B, for instance, which runs at about 6 tph all day except
for a brief period northbound in the AM rush and southbound in the PM
rush. I could also ride the Q, but I have to select one or the other in
advance, since they run on different lines in Manhattan.

One problem with express services is that they tend to dilute service to
local stations. Most of our expresses run on weekends, and a few even
run overnight!

I'm interested to get your feedback on what I've not noticed about NYC's
trains compared to ours here in London.

I think you've noticed quite a bit!

Also, most NYC neighborhoods not near the subway developed in the
automotive age. Most people in those neighborhoods use their cars for
all of their trips except into Manhattan. In those neighborhoods, the
only major demand for bus service is to the nearest subway station.
(And to nearby schools.) From what I've read here, London has a lot of
local travel by bus outside the central area.

I think we're sort of back with history here in that the rail and subway
networks are typically strongly radial links with little local traffic
(relative to flows into the centre) and virtually nothing offered for
orbital flows. Buses have always had a strong purpose given those gaps
in the rail network.

But my feeling is that the demand for local and orbital bus links is
much greater in London than in NYC. Car ownership and use is high in
parts of NYC not served by the subway.

In older times when we had less congestion many bus routes were very
much longer than today and lengthy radial journeys were also possible
into the suburbs or across the central area. Sadly this is now
relatively rare with few radial routes stretching from Zone 1 to beyond
Zone 2. There are more longer radial routes in South London that North
of the river - probably reflecting the influence of the tube network
north of the Thames.

We also have relatively few bus connections between the boroughs. There
are many routes running between Brooklyn and Queens (which does not
generally entail crossing a body of water) and between the Bronx and
Manhattan (which entails crossing the narrow Harlem River with its many
short bridges), but any other borough-to-borough trip involves a major
river crossing. We do have some -- the B39 and B51 between Manhattan
and Brooklyn, the S53 and S79 between Staten Island and Brooklyn (the
S93 is really just the S53 Limited), the M60, Q60, and Q101 between
Manhattan and Queens (also technically the Q102, which runs from Queens
into Roosevelt Island, formally part of Manhattan), and the Q44 and QBx1
between the Bronx and Queens -- as well as the (heavily subsidized)
express routes. But the vast majority of interborough trips are taken
on the subway.

You might find these links interesting:
http://www.nymtc.org/data_services/HBT.html
--
David of Broadway
New York, NY, USA

On Sun, 11 Mar 2007, David of Broadway wrote:

Paul Corfield wrote:
On Mon, 19 Feb 2007 23:57:05 -0500, David of Broadway
wrote:

Although London's rail network has pretty wide coverage, it has limited
capacity in comparison to NYC's. Our trains are wider and longer and most
of our major trunk lines (and some of the minor ones, too) have four
tracks. Given how crowded our trains get, if we had to give up our
express tracks and shorten and narrow the trains, the buses would become a
lot more popular, by necessity.

I'm a tad taken aback by your comments on the relative capacities of
London's rail network vs NYC's. Now I'm certainly not an expert on your
subway or rail network but surely your rail network (not subway) is but
a mere shadow of London's?

By "rail" I was including subway/Underground.

In that case, yes. But our overground trains are as big as, or bigger
than, NY subway trains.

But, including everything, you still might be right. It's difficult for
an outsider to get a good sense of your rail network.

It's pretty hard for an insider! The handful of lines north of the river
are simple enough - they're all basically like tube lines that happen to
stop at the Circle line, rather than continuing into town (barring the
North London and Gospel Oak to Barking lines). South of the river, though,
it's a different story - there's an untamed thicket of lines, all
criscrossing and interconnecting, and it's hard to believe anyone has a
solid grasp of it all. They're shown on this map:

http://www.tfl.gov.uk/tfl/pdfdocs/lon_con.pdf

Which i think gives some idea of the complexity of the topology. I'm not
sure how they wound up like that; for some reason, 19th-century railway
bods decided it was a good idea to build lines that were halfway between
radial and orbital, so now there's this matrix of overlapping spirals,
plus some more sensible radials. Maybe it was because the main station for
Kent, which is in the east, is Victoria, which is in the west. No idea how
that happened.

tom

--
Let us learn to dream, gentlemen, and then perhaps we will learn the
truth. -- Friedrich Kekule