On 04.06.2012 17:20, Marco Martin wrote:
On Monday 04 June 2012, Aaron J. Seigo wrote:
On Monday, June 4, 2012 11:49:41 Marco Martin wrote:
yeah, seems that connman or networkmanager support is statically linked
in the packagekit daemon and not a plugin, so doesn't seem possible to
have both backends compled and installed at runtime :/
so here's the next obvious question: why is conman still in Mer? :)
it really does not make much sense to have 2 backends mantained unless
there is a significant advantage unique to each one. right now, the only
advantage i know of is that conman is not actively developed? :)
have a packagekit here that uses networkmanager:
https://build.pub.meego.com/package/show?package=PackageKit&project=home%3Amart%3Abranches%3AProject%3AKDE%3ADevel
about connman i think is still alive as part of tizen, but not exactly
developed in the open..
Yes, connman is still used in Tizen. Question: how much more open it
could be developed?
Upstream webpage http://connman.net/
Git tree: http://git.kernel.org/?p=network/connman/connman.git;a=summary
Also there seems to be work done in Ubuntu as well.
https://wiki.ubuntu.com/ConnMan
personally i think moving to networkmanager at mer level as well would be more
future proof, i guess the resiztence at the moment is that the other ui
projects have ui only for connman?
What makes networkmanager more future proof? It has been developed much
longer, yes, and connman still lacks some of the features sure. But I
would say that it is getting there. Everything takes time when it is
started from scratch.
What is the main thing why Plasma Active people do not want to use
connman? Is it because of lacking some crucial features at the moment?
Or because one has done code for network manager that one don't want to
throw away and redo for connman? Or something else?
Regards,
Marko
