Nils, On Thu, 2008-03-20 at 16:14 +0100, Nils Faerber wrote: > What I am missing though, and this I cannot very well define myself, is > more of the mobile or mobility aspect. I think Gnome/mobile can do much > more than "just" tweaking things on the user interface side and very > shallow waters underneath. I think there are also aspects that > could/should be discussed within the Gnome/mobile context that reach far > more below in the stack - as one example: power manegement. The upper > layer stack elements can IMHO do a lot to at least help power > management. One project in that direction that comes to my mind is OHM, > which is a pretty nice mixture of upper and lower layer bound together > using dbus. And this is important since, like in this example, power > management on mobile devices comes very high in the priority list of > features. OHM is listed on the GNOME Mobile site as a technology under consideration, and yes along with HAL it is very useful. The low-memory-notify patches which have been floating around LKML are interesting too, it's been on my plan to test them and write a wrapper so that applications can receive gobject signals when memory is low. > Besides this I also see a lot of work in the areas of network management > (connection and link management, parts being now done using network > manager) which will get more and more important also for mobile devices. > And on mobile devices this has the extra attribute that the mobility > causes additional "troubles": Connection management must be seamless, > with the least possible user interaction, roaming of technologies (GSM > to WiFi to 3G, etc.). Also in this realm we find the issues of security, > i.e. all the different authentication mechanisms that need to be taken > care of for all the different technologies and networks (SIM card, WEP, > WPA, etc...). Network Manager handles the backend of this quite well, especially the development version in svn which handles modems, 3G cards, and so on out of the box with multiple active connections. For seamless transitions, this presumably means applications will need to monitor the connection and disconnect/reconnect automatically as the network layer changes. This is probably about 10 lines of code. > And last but not least there are the "voice networks", i.e. telephony of > all sorts for which we would like to have a common solution and API. > From my recent experience telephony API (short TAPI) is one of the most > discussed topics in all sorts of consortiums and standardisation groups > - without much success, as far as I know. Even bigger institutions like > LiMO, OHA, etc. have no big master plan for that, the LiPS TAPI was a > neat approach but not exactly what we would like for Gnome/mobile I > guess, etc. I think there was a workinggroup with Linux-Foundation and > OMA (?) but also without much success. We already have Telepathy in the platform, which handles IM and VOIP currently. Voice calls are not handled yet as far as I know, but I imagine that is possible. Ross -- OpenedHand Ltd. Unit R Homesdale Business Center / 216-218 Homesdale Road / Bromley / BR1 2QZ / UK Tel: +44 (0)20 8819 6559 Expert Open Source For Consumer Devices - http://o-hand.com/ ------------------------------------------------------------
Attachment:
signature.asc
Description: This is a digitally signed message part