google maps petition

Tom Anderson said:

This was posted on Bike Forums. It is a petition to Google asking for a
'bike there' feature on Google Maps

http://www.petitiononline.com/bikether/

And if you want to see how really usuful such a feature is go to:

Quoted message said:

http://www.camcycle.org.uk/map/

User input data here means the mapping in the Cambridge Area is very good.

You can choose Fastest, Shortest or Quietest route, and you get a 'run
through' of photos on the way.

Isn't part of the issue here that to show 'cycle routes' you need a
database of information that is probably only complete in Cambridge?

Jim Chisholm

J. Chisholm said:

Tom Anderson said:

This was posted on Bike Forums.  It is a petition to Google asking fora
 'bike there' feature on Google Maps

Quoted message said:

http://www.petitiononline.com/bikether/

And if you want to see how really usuful such a feature is go to:

Quoted message said:

http://www.camcycle.org.uk/map/

User input data here means the mapping in the Cambridge Area is very good.

You can choose Fastest, Shortest or Quietest route, and you get a 'run
through' of photos on the way.

Isn't part of the issue here that to show 'cycle routes' you need a
database of information that is probably only complete in Cambridge?

Jim Chisholm

I've used the Google route planner a couple of times recently for car
journeys and it have ended up abandoning its suggestions and using a
map. It wanted to send me down mazes of back streets when there's a
perfectly straightforward main-road way of getting to the same place.

The TFL cycle route planner in cycle mode does quite a good job at
providing quieter back-street routes for London so I guess it must
have a lot of cycle-specific data to work from.

OpenStreetMap is starting to build up a lot of cycle route information
that a route planner could use, not just numbered cycle network routes
but useful stuff like where roads are gated at one end but allow
cyclists through. There's a lot of potential there and the best thing
is that if the data is wrong you can fix it.

POHB said:

J. Chisholm said:

Tom Anderson said:

This was posted on Bike Forums. It is a petition to Google asking for a
'bike there' feature on Google Maps
http://www.petitiononline.com/bikether/

And if you want to see how really usuful such a feature is go to:

Quoted message said:

http://www.camcycle.org.uk/map/

User input data here means the mapping in the Cambridge Area is very good.

You can choose Fastest, Shortest or Quietest route, and you get a 'run
through' of photos on the way.

Isn't part of the issue here that to show 'cycle routes' you need a
database of information that is probably only complete in Cambridge?

Jim Chisholm

I've used the Google route planner a couple of times recently for car
journeys and it have ended up abandoning its suggestions and using a
map. It wanted to send me down mazes of back streets when there's a
perfectly straightforward main-road way of getting to the same place.

I suspect it it is too crude. To make such a model work it needs far
more that just link lengths.
Just think of the Manhatten Problem (how many possible routes of the
shortest length are there between to nodes on a grid network) No one but
a Mathematician would do anything other than a trip with a single right
angle turn to get between the two nodes.
The route you take also varies according to your percieved costs of
'distance' and 'time'. If you pay for your petrol, and have plenty of
time you may take the short and slow route, but if its the firms car and
you're in a hurry...

Quoted message said:

The TFL cycle route planner in cycle mode does quite a good job at
providing quieter back-street routes for London so I guess it must
have a lot of cycle-specific data to work from.

OpenStreetMap is starting to build up a lot of cycle route information
that a route planner could use, not just numbered cycle network routes
but useful stuff like where roads are gated at one end but allow
cyclists through. There's a lot of potential there and the best thing
is that if the data is wrong you can fix it.

And can you 'legally' port data collected on a Google map to the Open
Street Map you've just spent months creating?

Jim Chisholm

J. Chisholm said:

And can you 'legally' port data collected on a Google map to the Open
Street Map you've just spent months creating?

I'm not sure I understand the question, what do you mean by "collected
on a Google map"? Google doesn't have a way for users to load data
into its maps, but lots of sites use Google maps as a background for
data they've collected. It certainly wouldn't be legal to trace
streets or copy their names from a Google map onto an OpenStreetMap
but you could take your own data collection that you previously
displayed on a Google map background and enter it into OpenStreet
instead. However, if your data set was created by locating places on
a Google map rather than e.g. using a GPS I'm not so sure.

You'd be on dodgy copyright ground publishing cycle routes based on
Google data even if you could get access to the underlying routing
information rather than just the generated map image tiles.

I suppose the moral is that it's best to start off with copyright-free
data sources in the first place so you don't have to worry about
getting caught out later.

POHB said:

J. Chisholm said:

And can you 'legally' port data collected on a Google map to the Open
Street Map you've just spent months creating?

I'm not sure I understand the question, what do you mean by "collected
on a Google map"? Google doesn't have a way for users to load data
into its maps, but lots of sites use Google maps as a background for
data they've collected. It certainly wouldn't be legal to trace
streets or copy their names from a Google map onto an OpenStreetMap
but you could take your own data collection that you previously
displayed on a Google map background and enter it into OpenStreet
instead. However, if your data set was created by locating places on
a Google map rather than e.g. using a GPS I'm not so sure.

Correct. Users clicked onto a Goggle map to determine location, so
although the Lat and Long are determined it USED the Google base
Hence the uncertainty.

Quoted message said:

You'd be on dodgy copyright ground publishing cycle routes based on
Google data even if you could get access to the underlying routing
information rather than just the generated map image tiles.

I suppose the moral is that it's best to start off with copyright-free
data sources in the first place so you don't have to worry about
getting caught out later.

I hope others considering similar excercises consider that.

Jim

In article said:

Tom Anderson said:

This was posted on Bike Forums. It is a petition to Google asking for a
'bike there' feature on Google Maps
http://www.petitiononline.com/bikether/

[...]

Quoted message said:

Isn't part of the issue here that to show 'cycle routes' you need a
database of information that is probably only complete in Cambridge?

Presumably part of the point of the petition is to encourage Google
to gather such data where possible.

Just adding a "this blockage is a through road for me really" feature
to the existing route planner would be an improvement - for example
allowing you to go through the rising bollards in Cambridge, whether
because you want to plan a cycle route or to estimate a taxi cost.

If it's done as a "don't reroute this line segment even if it isn't on
a known road/track/path" function, then you could also use bits of off-road
tracks you know about even when they aren't in the database, and let the
normal existing routing handle the on-road bits.

Obviously it would be _more_ useful if it did know about shortcuts,
off-road cycle paths, and contraflow cycle lanes in the database (and
at least for some users knew which roads had shared use paths besides
them), but it could be better than it is currently even without that.

J. Chisholm said:

Correct. Users clicked onto a Goggle map to determine location, so
although the Lat and Long are determined it USED the Google base
Hence the uncertainty.

You can make your own website which overlays data derived from a
Google map (or anything else) onto an OpenStreetMap base. This is a
"Collective Work"; the OSM licence expressly allows it.

What you can't do is feed that data into OpenStreetMap itself.

Quoted message said:

Quoted message said:

I suppose the moral is that it's best to start off with copyright-free
data sources in the first place so you don't have to worry about
getting caught out later.

I hope others considering similar excercises consider that.

Oh, how I wish they would...!

On the wider point of the Google petition, I concur that such a
feature wouldn't be useful without the relevant data - which Google's
suppliers (TeleAtlas and Navteq) simply don't have. Away from
automotive routing, their data is pretty sparse.

So it has to be either user-generated or licenced from the OS. Google
are unlikely to go for a solution which is UK-only, so it'll be user-
generated. Getting back to your comment above, then, it's much better
for people to contribute to an openly-licensed project - such as
OpenStreetMap or similar - than to further enrich Google's coffers by
donating them your hard work.

Richard

POHB said:

J. Chisholm said:

Tom Anderson said:

This was posted on Bike Forums. It is a petition to Google asking for a
'bike there' feature on Google Maps

Quoted message said:

http://www.petitiononline.com/bikether/

And if you want to see how really usuful such a feature is go to:

Quoted message said:

http://www.camcycle.org.uk/map/

User input data here means the mapping in the Cambridge Area is very good.

You can choose Fastest, Shortest or Quietest route, and you get a 'run
through' of photos on the way.

Isn't part of the issue here that to show 'cycle routes' you need a
database of information that is probably only complete in Cambridge?

Jim Chisholm

I've used the Google route planner a couple of times recently for car
journeys and it have ended up abandoning its suggestions and using a
map. It wanted to send me down mazes of back streets when there's a
perfectly straightforward main-road way of getting to the same place.

The TFL cycle route planner in cycle mode does quite a good job at
providing quieter back-street routes for London so I guess it must
have a lot of cycle-specific data to work from.

OpenStreetMap is starting to build up a lot of cycle route information
that a route planner could use, not just numbered cycle network routes
but useful stuff like where roads are gated at one end but allow
cyclists through. There's a lot of potential there and the best thing
is that if the data is wrong you can fix it.

it is getting quite good isn't it. though doesn't work terribly well out
of london or other populated areas. nr my folks place it just lists the
main road, almost nothing else.

mind you google maps and the like are fairly poor there as well.

roger
--
www.rogermerriman.com

J. Chisholm said:

Just think of the Manhatten Problem (how many possible routes of the
shortest length are there between to nodes on a grid network) No one but
a Mathematician would do anything other than a trip with a single right
angle turn to get between the two nodes.

How big is Manhattan? I'm not a mathematician [1], but I can well
imagine attempting some form of limited zigzag if the time / effort
saving were large enough.

Cheers,
Luke

[1] although several of my family are

--
Red Rose Ramblings, the diary of an Essex boy in
exile in Lancashire <http://www.shrimper.org.uk>

Quoted message said:

I've used the Google route planner a couple of times recently for car
journeys and it have ended up abandoning its suggestions and using a
map. It wanted to send me down mazes of back streets when there's a
perfectly straightforward main-road way of getting to the same place.

Have to agree with you there; GM usually generates two or three times
as many steps in a route than (say) the RAC planner, plus it's not all
that anglicized - so you get tend to get un-numbered 'ramps' instead
of J16 etc.

It's unusual for Google, who normally do most things far better than
anyone else.

Regards,

Duncan

"POHB" <[email hidden]> wrote in message
news:[email hidden]...

J. Chisholm said:

Tom Anderson said:

This was posted on Bike Forums. It is a petition to Google asking for a
'bike there' feature on Google Maps

Quoted message said:

http://www.petitiononline.com/bikether/

I've used the Google route planner a couple of times recently for car
journeys and it have ended up abandoning its suggestions and using a
map. It wanted to send me down mazes of back streets when there's a
perfectly straightforward main-road way of getting to the same place.

I dont have a lot of faith in the Google route planner. Last month it sent
me down a lane in Rochdale that was blocked halfway along its length. Last
week it directed me to a cul de sac on the wrong side of Manningham Lane in
Bradford I was deposited around a 1/4 of a mile away from my destination.

J. Chisholm said:

And if you want to see how really usuful such a feature is go to:

Quoted message said:

http://www.camcycle.org.uk/map/

User input data here means the mapping in the Cambridge Area is very good.

You can choose Fastest, Shortest or Quietest route, and you get a 'run
through' of photos on the way.

That is really impressive. How did they do that ? Can they train the
rest of us ?

Paul

--
CTC Right to Ride Rep. for Richmond upon Thames