I received a lot of reactions about my previous blog. So here are some clarifications:

- I strongly believe that Open Data is the way to follow, but this way requires to have strong fundations. The goal of this blog was simply to highlight actual problems that can and should be solved. I have also read the nestoria success story, but I would like to see hundred of equivalents stories ;-)
- Yes, I use OSM. I'm a contributor, too ;-)
- Point #1: Cartographers have demonstrated that automatic generalization is, at least for now, not possible. If we take the example of the UK Ordnance Survey, Vector Maps are delivered with two scale ranges: local and district.  Google Maps use various sources of data depending of the scales: GeoBasis DE/DKG, TeleAtlas, Europa Technologies or their own data.
- Point #3: OSM has always a duality between cartography and GIS. The main goal of OSM is to make a map, but GIS data are used to create this map. From a GIS perspective, the fact that "You can assign tags ("keys" and "values") to any point (node), street (way), area (closed ways) or relations (containers of the previous ones). see" creates difficulty when you want to use the data.
- I'm really suprised that points #4, #5 and #6 can be questionned. These are simply facts. A standard GPS is able to measure with a precision of 2 to 5 meters. The digititalization over an orthophoto with a 20 cm resolution is able to measure with a precision of 20 to 50 cm. A professional GPS is able to measure with a precision of 2 to 5 cm.
- Point #8: this is due to two reasons: all is based on three primitives and tags can be associated with everything.
- Point #11: all is in english in OSM. A translation mechanismus should exist is order to be able to obtain localized data.
- Point #13: this is the KISS principle. Currently, it's not very simple (or user friendly) to acquire data for OSM.

Regarding the interesting blog of Steven Feldman, I will extract the sentence: "...for this level of trust to be achieved a more formal approach to quality assurance and a more structured and consistent approach to data capture (content, geography and attribution) will be needed.". I think that the majority of my points goes toward this direction. (small hint to the author: precision and reliability are mathematics concepts used to deliver information about measurments (introduction))

Reactions

@cedricmoullet Rebuttal: wherewithal1.wordpress.com/2010/08/03/why… and wherewithal1.wordpress.com/2010/10/14/tra…

@cedricmoullet Not sure about #3, #8, #10, #11, #13 Thanks for your points, more background about why you say that would be cool :)

@cedricmoullet 1, 7 and 9 are true also for Google Map

@cedricmoullet I can understand your stress -> goo.gl/07sIu but many of your points are valid also for google maps #openstreemap

"Why OSM fails..." by @cedricmoullet: points 1/3/4/5/6/7/8/10/11/13 really? have you ever used it? :) http://bit.ly/tF33HH

@cedricmoullet have you ever used #OSM?

Promotion for own products? Sadly no comments are allowed :( RT @cedricmoullet: Why #OpenStreetMap fails ... http://www.cedricmoullet.com/news/whyopenstreetmapfailstoreplaceofficialbasemapsinasustainableway

@pascal_n @cedricmoullet Some strange arguments, IMO. E.g. 99% of all maps are not multilingual. And what does Map Maker have to do with it?

@cedricmoullet Please allow comments

Not all wrong, unfortunately RT @cedricmoullet Why OSM fails to replace offic. or propr. base maps in a sustain. way ?

@puntofisso I think @cedricmoullet misses the point (pun there) #osm not meant to replace official maps but he shld see j.mp/rqNd15

@StevenFeldman @cedricmoullet as you say, but what strikes me is the number of wrong assumptions about #osm

Really ? Then, how do you account for this http://test.geobretagne.fr/mapfishapp/ and that http://geobretagne.fr/geonetwork/srv/fr/main.home?uuid=c098fa63-42a7-4c56-90de-1c0877ff32a8 ?