Frederik Ramm wrote:
> Hi,
> 
>     so we know that the minutely diffs are broken (and we have created
> the minutely-replicate diffs instead; something that I had completely
> missed until it was mentioned casually a few days ago - can osmosis
> --rci be run on them or do they require special treatment?).

There are corresponding --rri and -rrii tasks for the replication based 
diffs. I got these to work by doing this:

Initialiasing:

osmosis --rrii workingDirectory=.

You'll have a configuration.txt at this point. It has a wrong URL for 
the minute-replicate diffs, so correct that.

Then run the main task once:

osmosis --rri workingDirectory=. --wxc diffs.osc.gz

That fetches a state.txt from the server. It's likely you'll want to 
pick up the diffs at some earlier point in time. Go to the diff server 
and find the right state.txt with a timestamp for the time at which you 
want to pick up the diffs. Download that and put it as state.txt in your 
workingDirectory.

Brett has stated that it's best to have an overlap, so personally I 
start from a diff 3 to 4 hours before the timestamp I actually want. 
It's also possibly I'm misunderstanding, or that Brett was overly 
cautious, and you can keep the overlap to a minimum.

 From that point on, you can run osmosis --rri and output to a change 
file, or pipe it into other tools:

osmosis --rri workingDirectory=. --wxc | osm2pgsql -a .....

By default, it fetches one hours worth of diffs at each run. You can 
change that in configuration.txt.

> But until now I had assumed that at least the hourly diffs were correct
> which probably was naive - I guess that it is just *less likely* for
> them to miss out on something than the minute diffs, but not impossible?

Yes, even the daily diffs can miss some objects, because of the same 
reasons, although it's even less likely than with the hourly diffs.


-- 
Lennard

_______________________________________________
dev mailing list
dev@openstreetmap.org
http://lists.openstreetmap.org/listinfo/dev

Reply via email to