[OSM-talk] Organisation of Wiki pages

[Lars Aronsson]

Etienne wrote:

> So how long do we wait for these improvements?

The key is Steve's attitude.  So far he acts as if it doesn't 
matter if the slippy map works or not.  That has got to change.
When I try to point this out, I'm told I'm trolling.  That's sad.

I cannot demo the slippy map, because I can never trust it to 
function correctly and timely.  The only people I can recruit to 
OSM are programmers who understand that some software doesn't 
always work.  Then I lie to them and say that, hey, it's open 
source and they can make a change.  In fact, nobody has made any 
change, because Steve is blocking this.  Imi is right.

The way we get anything useful out of OSM is through the dumps.

This is the problem with the slippy map.  I wish we could find a 
solution for it.  However, I don't quite understand how the wiki 
could be the solution.  Sure, it is faster (actually, the 
MediaWiki software developed for Wikipedia was very slow and 
inefficient all through 2002 and 2003, but has since improved 
immensely), and the wiki can be a useful tool for a handful of 
towns that you need for a nice demo.  But how many pages for each 
little placename are you going to create?  And how are you going 
to maintain that data?  When I update the map, who updates the 
static images on the wiki for those cities? The wiki has no 
function for finding other wiki pages that describe places nearby.  
Are you going to create all links between [[Oxford]] and 
[[Cambrige]] and all other pages where this is relevant?  Could 
you provide an estimate of how big this project will be?

The GEOnet names server database contains 5.37 million names 
describing 3.90 million features.  The subset for Sweden contains 
97,000 names.  The English Wikipedia has 1.3 million articles and 
the Swedish Wikipedia has 180,000 articles.

> With almost zero effort we can improve the experience of the 
> casual visitor from an 8 minute wait to a result within 15 
> seconds.

Say again, how is this almost a zero effort?

> We have seen prototypes of better solutions, but they are *only* 
> prototypes

The slippy map is a prototype too, and so is your wiki ideas.

> Lars, I know that you have contributed enormously to the OSM 
> project, but I have never, ever, seen the result of anything 
> that you have done.

Yes, it is frustrating not being able to see the fruits of ones 
work.

> Does anyone have access to the web-server stats?

This is the key.  Getting the stats, counting how many calls we 
serve, how many succeed, and how many fail.  That's acting as if 
errors matter.  Right now, we're not acting that way.  Steve is 
blocking access to the server, and doesn't publish logs or stats.
It's a black hole where nothing comes out.


-- 
  Lars Aronsson (lars at aronsson.se)
  Aronsson Datateknik - http://aronsson.se