On Tuesday 23 November 2010 01:16:00 ext Joly, Frederic, you wrote: > I just recheck the 23.032 (Universal GeoGraphical Area Description), and the > number of ways to express a position with velocity and uncertainty may not > make the API simple. Just have a look for instance the +CMOLR's XML DTD > location_parameters. :) Moreover, GPS vendors and other positioning > professionals would find more logical to propose to use the WGS 84 datum > for instance which is widely used and leave the burden to map the position > to 3GPP style...
I am afraid with enough time and products, we will find all possible combinations: - modems that only talk 27.007 XML, - modems that only talk raw binary, - modems that support both, - hopeless modems that do none. Furthermore, the data flow can go in both directions: AT+CPOS is engine->oFono->modem, AT+CPOSR is modem->oFono->engine. I would think that oFono needs to hide the differences between modems, which mean it will need to convert in some cases. Thus oFono would need to be able to convert both ways in any case, right? This is going to be painful. The alternative consists of providing a *leaky* abstraction where oFono only returns and accepts the format(s) that the modem understands. With that in mind, I tend to agree that binary seems like a nicer abstraction for the positioning engine to use than XML. But if oFono has to implement conversion in both direction, it might be that we can provide both formats over D-Bus? -- Rémi Denis-Courmont Nokia Devices R&D, Maemo Software, Helsinki _______________________________________________ ofono mailing list [email protected] http://lists.ofono.org/listinfo/ofono
