[OSM-dev] Proposal for T at H land/sea updates

[Alan Millar]

Thanks for everyone's feedback on my idea

>> The problem with this is that this data is distributed as part of the
>> code.

Sebastian Spaeth writes:
> That is correct and it is not perfect.

Martijn van Oosterhout writes:
> Distribution of the data was a problem considered at the beginning but
> never really solved.

I have now found that there is some update code built into tilesGen.pl. 
It checks a version file on the svn server, and if the version number has
changed, it does an svn update to retrieve the latest version, which
includes any changes to oceantiles_12.dat.  This does have some advantages
over my idea.  It keeps everything up-to-date, including osmarender style
sheets and, of course, the code itself.

It is based on a "version.txt" file on the server.  Anyone know who/what
updates that?

> Possible, but IMHO it should still be possible to use close_areas.pl in
> offline mode, as some use it to render local data files...

This is a very good point that I had not considered.  I think it is
important.

I think I will concentrate first on making the svn update work reliably. 
I found that when I delete everything from my work directory, which
includes the local version.txt file, the code doesn't always reset it and
pick up a new version.  The code looks right, so I'll have to do some
debugging.

While updates to the binary file may or may not be the most efficient way
to do it, they are relatively infrequent, and probably not significant in
the total traffic of the t at h client.  And it still allows offline use.

Dirk-Lüder Kreie writes:
> There is also the update mechanism via informationfreeway.org at z12,
> but I am currently not sure who maintains the updating of oceantiles.dat
> from there.

Yes, I need to investigate that.

> Basically all that's
> interesting to close-areas.pl is: do we need to make a coastline-way
> around the tile to render and create islands inside, or do we treat
> closed coastlines as "lakes"?

Sebastian Spaeth writes:
> oceantiles_12.dat is only consulted if close_areas.pl can't decide. So
> the issue could be with either close_areas.pl *or* with
> oceantiles_12.dat. There have been bugs in close_areas.pl previously, so
> it might be worth checking if it's really the oceantiles.dat that is at
> fault.

Good points.  I've found a few tiles that are drawn inverted regardless of
the hint in oceantiles_12.dat.   I will investigate that further.

> I'd say that a new layer *might* be a bit overkill, as blankness
> information is in the current default layer as well.

I didn't realize there was a default layer.  I'll go learn some more about
it.

So for now I'm going to concentrate on debugging existing features before
introducing new bugs, I mean features :-)

Thanks for the feedback and education.

- Alan