Retire MapTiler

[camerons]

As per request from Klokan Petr Pridal, the MapTiler point of contact, we should retire MapTiler.

Of note, MapTiler has remained at version 1.0beta2 for a few years now (since OSGeo-Live 3.0), which suggests a diminishing open source community behind the project.

Todo:

• Update maptiler Project Overview and Quickstart to retired status in main docs.

• comment out install_maptiler.sh from main.sh installers

• comment out maptiler from metrics.rst docs

[camerons]

I've updated the docs to retire MapProxy, committed in version 10469.

Still needs to be commented out of main.sh installer.

[hamish]

I find the MapTiler program to be still useful and relevant, and would by willing to maintain it on the live dvd on an ongoing basis. (it's tiny and zero maintenance anyway so I don't mind much)

Since it just works, it doesn't matter that there hasn't been a new version in ages. It just means it is mature; the GDAL backend it uses regularly updated so it "gets updates" that way.

regards, Hamish

[hamish]

Cameron wrote:

I'd like to hear thoughts on whether we should be retiring MapTiler.

Preface: I view the live disc as "the best of open source mapping", not just "the latest developments in open source mapping". And so I don't mind looking after some oldies but goodies if they're not impacting anything else. I also don't think we should be in the business of picking favourites in the case of projects with parallel functionality- instead we should be providing options a, b, and c for our users to try out to decide which is best for them without having to go to the trouble of installing and learning how to get started with all the available options on their own. (2c)

My thoughts are that we should be helping new users find established Open Source projects, which typically are projects which have a strong community behind them.

I note that MapTiler hasn't changed version number since OSGeo-Live 3.0,

you'll notice that PROJ.4, grep, and ls don't change much either. It doesn't mean they aren't used or are no longer useful.

and is still reported as a beta status.

I see no reason for it to be classified as such.

I also note from the Ohloh metrics, there seems to be very little development: ​http://live.osgeo.org/en/metrics.html

As above, that's somewhat of an orthogonal measure, just like SLOC can be a rather poor measure of a programmer's output. Indicative, perhaps. But definitive? nope.

I suspect that users might be moving to another project instead? Eg: GeoWebCache?

GeoWebCache is a geotile-aware network proxy. MapTiler is a front end to the gdal2tiles command line program which takes a raster image as input and creates an openlayers web page and directories of image tiles for each given zoom level. Gdal2tiles comes with GDAL (and so is updated as needed). As long as the command line interface to gdal2tiles doesn't change, then MapTiler doesn't need to: it keeps going just with an upgraded GDAL engine each time. One helps serve map tiles, the other creates them.

It's a nice little tool, only 440kB (compressed), and AFAIK both the fastest and simplest tool available to get a live web map up on the internet. A couple of clicks, copy the result over to a web server, and you're done in 3-5 minutes with very little training needed. That "holy cow I'm done!?" + "I made that :)" user reaction is the main reason I want to keep it. It makes FOSS4G look easy and good.

regards, Hamish

[camerons]

As decided in OSGeo-Live weekly meeting, MapTiler will stay on OSGeo-Live, in line with Hamish's comments. Docs changed back in version 10482. (Maptiler application wasn't uninstalled).