On Mon, Jan 16, 2012 at 12:03 PM, SomeoneElse <li...@mail.atownsend.org.uk> wrote:
> >> If I was following some documentation that says to use a particular > version of something (say, Mapnik 0.7.1) I wouldn't expect to be able to > use some random newer version and expect to still be able to follow the > instructions. If the documentation was updated to say "and here's where > you get the version of Mapnik that these instructions were written for", > and that was somewhere that wasn't at the mercy of a third-party project > changing version control systems, then we'd be in the best of all possible > worlds. > If I was following some instructions that pointed to a SVN repository that was moved to git, and then did not contain any version prior to 2.0 of the said software, I would also try the next available version since 0.71 isn't available. What I had to do was finally find a PPA for Ubuntu 10.04 that has the 0.72 branch of Mapnik, because I couldn't get any of the 2.0 versions to compile. > ... and a pre-configured VM would all be ideal ways to prevent this sort >> of thing from happening. >> >> which would work, except that given the hardware needed to serve a map, a > VM is probably not ideal for most people. Who are "most people"? Are you really saying that most people will build things on static hardware? Especially people who are trying out new software to see if it works for them? I heartily disagree, and I think an OVF of a preconfigured map server that already HAS the planet_osm file imported into postgis would be not only extremely useful but would get you a lot more people taking the mapserver for a spin. It also doesn't explain how to get OSM running on a server that's already > installed doing something else. There are off-the-shelf black-box > solutions out there already, but they don't explain the integration process > any - which is exactly whathttp://en.flossmanuals.net/** > openstreetmap/setting-up-your-**own-tile-server/<http://en.flossmanuals.net/openstreetmap/setting-up-your-own-tile-server/> > and > before it > http://weait.com/content/**build-your-own-openstreetmap-**server<http://weait.com/content/build-your-own-openstreetmap-server> > do > extremely well. I hate to disagree, but I don't think they do this extremely well. I think the documentation is feasible but has flaws and every time something changes in the build infrastructure, this will happen again. Today, if you follow these instructions, you can't actually build a map server based on them. On the other hand, if you provided a modified version of these pretty good instructions, along with a chef recipe / juju charm that builds it for you; or if you provided the instructions and a VM that lets you take them and start up a map server with less effort and hacking than is now required, you would be in a much better place. Roger Weeks
_______________________________________________ dev mailing list dev@openstreetmap.org http://lists.openstreetmap.org/listinfo/dev