[OSM-dev] Binary OSM; the first pass encoder

[Frederik Ramm]

Hi,

Chris Jones wrote:
> You cant sensibly compare the two if you apply different compression 
> methods.
> 
> Given how compression works I suspect you'd do better to tweak things 
> like the block size used at compression time than come up with some 
> binary format.

Depending on the design of the binary format, it might actually allow 
random access (i.e. seeking to a position in the file and grabbing what 
you need via an index) without first uncompressing it. It is probably 
impossible to create a binary format that is as small as a gzipped XML, 
but it might get *near* as small and could be processed without first 
unpacking it.

For stream applications like Osmosis or osm2pgsql, it is irrelevant 
whether you have do decompress the stream first, but if you want to run 
something on a mobile device with limited storage then an indexed binary 
format will probably be vastly better than compressed XML!

Bye
Frederik

-- 
Frederik Ramm  ##  eMail frederik at remote.org  ##  N49°00'09" E008°23'33"