On Sun, Jul 28, 2013 at 12:55:13PM -0700, mapl...@light42.com wrote:
> Hi All -
> I noticed this
> Maintainer: Debian OpenStreetMap Team <email@example.com>
... so I decided to CC this team (isn't this obvious??) and also the
relevant bug report (#710039) ... which is admittedly a bit redundant
but anyway. The fact that there is no response to the bug report since
two monthes would tell me, that the two uploaders are busy with other
things / on vac whatever and are happy to receive patches (otherwise
they could have answered: "No, we intentionally keep this version.")
So you can decide: Well, that's the Debian maintainers problem - I'll
just keep on working for osgeo or you consider it a good idea to
checkout the existing packaging. If you
you actually see that Giovanni Mascellani has decided to leave the team
last year and the missing manpower issue becomes obvious.
Trying to verify the hypothesis that there is a problem with manpower
in pkg-osm team which leads to not up to date packages we check the
team statistics for this group:
What people are discussing
Who fixes bugs
Who commits code
And finally who uploads packages
you come to the conclusion that the main driver of this team is David
whom I have the pleasure to (virtually) know from Debian Med as very
active and resonsive. So if even David does not find the time to
respond in a reasonable time frame I'm afraid that several packages of
pkg-osm are in danger to be improperly maintained.
The statistics for pkg-grass / debian-gis are a bit better and finally
if you compare the amount of packages via an UDD query
udd=# SELECT maintainer_email, count(*) FROM (SELECT DISTINCT source,
maintainer_email FROM sources WHERE maintainer_email like '%pkg-grass%' or
maintainer_email like '%pkg-osm%') tmp GROUP BY maintainer_email;
maintainer_email | count
firstname.lastname@example.org | 16
pkg-grass-de...@lists.alioth.debian.org | 72
the number of packages in pkg-grass repository is four times higher than
in pkg-osm. So in case people would consider a merge moving from
pkg-osm to pkg-grass would mean less work.
As a consequence out of this analysis I would propose the following:
Create a patch against the latest status of osmosis Git repository. I'd
take over the responsibility to care for a move into pkg-grass team and
will sponsor the updated package as offered in SoB. We keep on
discussing about this plan on pkg-osm list which leaves enough time to
insist if the plan is not accepted.
> while looking around.. The reason I mentioned osmosis recently is,
> the version we have on the Live is
> Osmosis Version 0.40.1
> which is very stale.. many new features and filters have been added
> "The stable releases below should be used whenever possible."
> I verified current stable runs correctly on the Live beta1 with:
> ./bin/osmosis --read-pbf-fast file="$PLANET_FILE" workers=8 --b
> bufferCapacity=10000 --tf accept-ways "building=*" --tf accept-nodes
> "building=*" --tf accept-relations "building=*" --write-pbf
> I dont know about java packaging, but osmosis seems like a decent candidate
> for learning how ..
> I think it would be useful to have a more current version.
Yes, for sure are current versions useful. I just want to lay out a
plan how we could have better chances for a sustainable way to get
recent versions also in future.
> best regards from Berkeley, California
Best regards from way to hot Germany
PS: I'll be at DebConf13 and would be happy to talk about Debian GIS
Pkg-osm-maint mailing list