On Monday, May 1, 2017 7:16:34 PM AEST Slava Monich wrote: > but that didn't quite fix it. It does eliminate one scenario (described in > the merge request) but there must be something else there. The whole > connman bearer plugin thing looks generally prone to race conditions, I > would've done it differently but we have what we have.
I don't think 'race' is the right term for what is happening. The whole networkAccessible thing was for Symbian, so the network could manually be put into a not accessable mode, when the network was actually still accesible. It was never really intended to be "am I online" check for developers/apps. Symbian had network access per app. Other platforms do not have this concept, it's the system that has network access. The 'race' issue is really just that connman is async and QNAM basically expects low level to be synchronous and that settings are always instantly available when the QNAM object is contructed. connman doesn't work like that. and sometimes, connman and wpa_supplicant are just plain buggy as well. As well, connman has a concept of 'network working' and 'internet is working', both of which are buggy. So things are happening in an unexpected order sometimes. To top it off, all these settings go through dbus. so all the wireless AP's that are nearby go though this and need to be handled in some way. So the dbus is getting pounded at times. _______________________________________________ SailfishOS.org Devel mailing list To unsubscribe, please send a mail to devel-unsubscr...@lists.sailfishos.org
