[OSM-talk] db and user statistics update

Robert (Jamie) Munro rjmunro at arjam.net
Thu Jan 18 23:13:18 GMT 2007


Jochen Topf wrote:
> On Wed, Jan 17, 2007 at 12:26:43PM +0000, SteveC wrote:
>> Date: Wed, 17 Jan 2007 12:26:43 +0000
>> From: SteveC <steve at asklater.com>
>> To: David Earl <david at frankieandshadow.com>
>> Cc: OSM <talk at openstreetmap.org>
>> Subject: Re: [OSM-talk] db and user statistics update
>>
>> * @ 17/01/07 11:50:45 AM david at frankieandshadow.com wrote:
>>>> I would have to say that the first impression of the wiki is one
>>>> of confusion over:
>>>> 1) Where to look first
>>>> 2) Where to find the map
>>>> 3) What is all this anyway?
>>> The map was originally reached by www.openstreetmap.org, but was suppessed
>>> last year because it was a poor advert for the project.
>>>
>>> But it has improved enormously since then, so maybe the time has come to
>>> reinstate the map on the main URL and the wiki comes second.
>> A number of people have asked for this. Any major objections?
> 
> I don't think the map is ready yet. It is very confusing for an outsider
> to have four different maps, especially if the default one is most
> outdated, as it seems to me.

I agree.

> I am all for having a slippy map on the home page, but
> 
> * it should only contain one map (we can have the other layers somewhere
>   else for testing)

We could have 2 maps, Mapnik DB and Osmarender, but no more.

> * it should be complete, i.e. all tiles should be there (99% of the world
>   does not have OSM data, so that shouldn't be too hard to do with just
>   the continents and country borders.)

Although Mapnik WMS-C is in many senses the best layer (it generates in
real time - so always has data), we can't use it because it uses a
complicated strict spheroid version of the Mercator projection, rather
than the simple sphere-based one that everything else uses. While it
makes a difference of a few km in the UK, I see very little advantage to
the projection in practical terms, and the maths is much harder. I think
that either it should be changed to match the other projections, or it
should be removed. I know the author of the WMS-C layer will probably
say Christopher Schmidt will say "But my projection is right and yours
is wrong", but in fact they are both wrong. In reality the earth is
round and doesn't fit on a 2d screen. Neither projection is "right", nor
is any other.

> * It should show some kind of indication when it was last updated. Maybe
>   each tile can be labelled in a corner with the date it was rendered?
>   At least there should be a notice below the map saying something like
>   "This map does not always show the most current data, but nothing you
>   are looking at is older than three days."

No! Please don't label each tile - that would be horrible. I don't think
an indicator of last update will be necessary once we get to the stage
of the map basically being up to date. Based on the fact that WMS-C is
able to keep up (subject to the planet dump), I think there is no reason
Mapnik DB won't be able to keep up, and we should soon have the ability
to pull updates rapidly into Mapnik's PostGIS database.

Robert (Jamie) Munro

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 249 bytes
Desc: OpenPGP digital signature
URL: <http://lists.openstreetmap.org/pipermail/talk/attachments/20070118/3ebbd392/attachment.pgp>


More information about the talk mailing list