[OSM-talk] Things People Say

[Michal Migurski]

On Dec 30, 2011, at 7:42 AM, Frederik Ramm wrote:

> In fact, many of those arguing for OSM becoming a nice map portal haven't even touched the subject of editing in their argument.
> 
>> 
> ...
> 
> Secondly, and this touches on something from my "looking forward" post a few days ago, we have always made it clear that there are no official tags and no official list and no promise that anything gets rendered anywhere. This has many advantages, decoupling editing from rendering, and brings many freedoms, but if we were to push that "one true map" or maybe these "ten true maps" and try to be the map portal for everyone then that would be the end of saying "well the Mapnik map is just a showcase and you cannot expect us to render everything". We would clearly make a much stronger bond between editing and rendering; fewer and fewer people would be willing to map things that are not on our main map(s), and we'd be pushing specialist maps to the sidelines. Let's not kid ourselves: Competing with Google Maps *will* make us more like Google Maps.

For what it's worth, I didn't really start editing until the Mapnik loop closed and I could see my edits on the same day I made them. This was a huge difference for me, and a step up from making an edit, waiting several days for it to show up, and then making another one. The "main" Mapnik layer is one of the only products OSM makes that's visible to regular humans, and even though someone like me *could* stay up-to-the-minute with the replication diffs it's more often the case that it doesn't happen.

Serge's mention of closing the circle is critical, because it keeps the promise inherent in the name "Open Street Map" - somewhere, somehow, people should be able to see a decent map, and it would be best if it were a reliable enough resource to be generally usable by the world at large.

At the same time, the current home page is 99% map and 1% anything else, so we're *already* implicitly competing with GMaps through our homepage design. Here's my opinion of what the front page should look like, based on a pastiche of elements pulled from various OSM sites including .de, the wiki, and the current page:

	http://mike.teczno.com/img/osm-homepage-sketch.jpg

First, there's still a map across the top of the page, but now it's smaller. It's kept for two primary reasons: visitors must be able to see that ultimately all the work leads to a usable map, and six years of permalinks to specific locations should not be broken. The overall tab layout across the top (with the Edit tab) stays, but I expect that each tab might lead to a second page with a taller, more page-hogging map on it for roomy editing.

The space below the map is there for the project to explain itself. We should quibble about the choice of sections (I've borrowed these wholesale from .de) but this area is in place to say something about what OSM is *for*: it's something you can join, there's data you can use, etc.

In the bottom-right corner is the wiki Image Of The Week, which is so often the home of solid gold output from the OSM community, whether it's new renders or photos of mappers. This part will change, and will reinforce the dynamic nature of the project. There will be cool and weird pictures there.

In the sidebar, I've just copied some stuff from the wiki. Honestly, I don't know what should go here—I never pay attention to sidebars on websites, but presumably if someone is absolutely scratching their head then being able to scan the page for the word "Help" or "Blog" will get them out of a jam.

Back to the map:

The OL layers menu is replaced with an old-Gmaps-style set of buttons, because if you don't know OpenLayers how will you know that the Blue Plus will give you something interesting? Buttons encourage pushing, which addresses the need to show a variety of cartographic outputs beyond the default Mapnik layer. The current choice of layers is good: "Mapnik" is what editors need to see what they are doing, "Cycle" and "Transport" both show what it means to highlight entirely different sets of tags, and "MapQuest" shows that third parties with recognizable names should get involved. I can't think of a rationale for keeping Osmarender in the list.

"Data" is a special case - I think it should be implemented in the form of a combined thinline raster layer and a backing data layer driven by a data format similar to Mapnik's new UTF Grid feature to support clicking on features. If we do this smartly, then we can simultaneously solve clickable POI's which is one of the OSMF's list of Top Ten tasks that they want handled.


> I believe SWG are having a discussion about "core values" at the moment. Suffice it to say: There are core values of this project, and if you don't share them than you can care for OSM as much as you want, you're in the wrong project. Now what exactly these core values or important goals are, is open to discussion. *My* vision is that by providing excellent map data, we put everyone in a position to make the map *they* would like to have. I know that this is more of a hurdle than "let's provide a drop-in replacement for Google Maps tiles", but I believe that the end-user stands to benefit from that.

My vision is that tiles are a form of excellent map data. Now what?

-mike.

----------------------------------------------------------------
michal migurski- mike at stamen.com
                 415.558.1610