Future of location services on OM

Stefan Schmidt stefan at openmoko.org
Mon Jan 12 22:20:46 CET 2009


On Mon, 2009-01-12 at 21:10, Nick wrote:
> I have to work-out the license issue as you mentioned. I really started from
> the OSM one...
> even if I have seen all the discussions about it.
> I am ready to consider all the solutions to solve this license problem.

I have no concrete example of a license that will keep the spirit but avoids the
problem. Jan, have you more insight here?

> >So you get the confirmation that the opencellid data is also under
> cc-at-sa?
> >We did not found a concrete statement about it and Daniel already sent them
> a
> >mail to clarify it.
> When I started http://www.openBmap.org and until a few weeks ago,
> there was no mention about the data license on opencellid website.
> (and impossible to reach them...)
> The other projects mentioned on
> http://en.wikipedia.org/wiki/GSM_localization
> did not mention their license too.

This is bad as it means you can't use this data. No license means you have no
rigths granted. I would suggest to hold back with your merge of the opencellid
data until this is clarified. I now it sucks to deal with these things, but it
is needed. :(

> As the logo creative commons appears now on the opencellid website, I have
> decided
> to import their data in my database. I will still able to erase their data
> if needed...

The logo is placed next to the copyright for the website, so it could also stand
for the license of the website...

> Thanks! and please have a look at the following url for zones in Germany
> http://realtimeblog.free.fr/maps/with_google_earth.php?mcc=262&mnc=7&step=2
> or
> http://realtimeblog.free.fr/cell_map.php?mcc=262&mnc=7&lac=&cellid=&display=Display
> (please zoom out on Google maps to see Germany coverage for O2 operator)

Heh, I tried with 1 and 3. :)

> > Do you merge not only the data but also the projects? Having them
> co-exists with
> > the same goal is kinda pointless.
> Because
> 1) the CC license seems "all new" to opencellid.org
> 2) the way to calculate cellular "average" position and attributes are not
> the same...
> 3) the signal strengths are not stored at opencellid.org side
> I 'll wait a little before considering a merge.

Well, it could still turn out that they are using the same license and are fine
with storing more infos like you. Have communication is the critical point

> > I like to have more data available. Can you give us some insight how big
> such
> > collections can grow. For example your current db size and the current
> number of
> > entries? I would like to have the option to have such a db on my SD card
> for
> > offline use and syncing when on wifi or similar.
> From the information I have collected, there must be 50000 towers in France
> (spread
> on 550000 km2) and I evaluate from 4 to 8 cells per tower for one operator.
> It means from 200 000 to 400 000 entries in a database for one country (as
> large as France)
>  for one operator (to have the full country in local SD database)
> For these 400 000 entries, you need to store
> mcc/mnc/lac/cid and two double for latitude and longitude
> (and perhaps other attributes such as max radius...)

Ok. Good that big microSDs card are coming and getting cheap. :)

> Probably you will not need the full country all the time on your SD card. We
> probably can design
> a way to sync the local openmoko database with the openbmap database... On
> LAC basis for instance...

Yeah, could be an idea.

> If you want to have a closer look at the current local database
> (I extracted the local database from the website), please look at
> http://realtimeblog.free.fr/latest/realtimeblog_23_12_2008.zip

1,9MB is still really lite. Before thinking to much about the storage problem we
have some time then.

Let's wait for the developer of the openmoko client showing up and discuss
further integration steps then.

Stefan Schmidt

More information about the devel mailing list