[OSM-talk] Blue sea tiles for tiles at home

[Frederik Ramm]

Hi,

>> In the long run we should change
>> back to running Frollo after close-areas but it works for now.
>
> Silly question: Why?
> What are the advantages and disadvantages of either order?

The thing is that close-areas.pl takes a plain standard OSM file  
without extra fancy attributes, and writes out the same.

So any software that is designed to work with plain standard OSM  
files will always work with the output of close-areas.pl.

Osmarender (as used in tiles at home) however is designed for working  
with "frolloized" OSM files. There are some things that will not  
render properly unless Osmarender sees special tags inserted by  
Frollo. (At the moment, the only such thing I am aware of is areas  
with holes, but both Osmarender and Frollo might change in the future  
to increase that list.)

Now, since close-areas.pl emits plain standard OSM, the combination  
"first close-areas, then Frollo, then Osmarender" will always work,  
no matter how many things are changed in Frollo and Osmarender, as  
long as Frollo takes plain standard OSM files.

The other way round, "first Frollo, then close-areas, then  
Osmarender", means that whatever extra ways and segments close-areas  
creates, they will always have to look like ways and segments pre- 
processed by Frollo. So if Frollo is changed to add some kind of  
extra information to ways, and Osmarender then depends on that extra  
information, then close-areas needs to be changed as well so that the  
ways it creates also carry that extra bit of information.

It is really not *that* big deal but it is an unneccessary potential  
trouble spot.

> Nice work on coastlines by the way... it is a massive improvement. We
> just need more of the coast to be actually in the DB, although even
> "level-12 tile as a pixel" level of detail is better than nothing :-)

In many areas coastline is actually present, but level-12 tiles  
haven't yet been re-rendered so that the level-7 tiles you see now  
are not as detailed as they could be. Give it a few more days...

Bye
Frederik

-- 
Frederik Ramm  ##  eMail frederik at remote.org  ##  N49°00.09' E008°23.33'