From mboxrd@z Thu Jan 1 00:00:00 1970 Content-Type: multipart/mixed; boundary="===============7398251705421960590==" MIME-Version: 1.0 From: Denis Kenzior Subject: Re: [RfC] Retry registering packet radio to home PLMN when available. Date: Mon, 26 Oct 2009 19:06:58 -0500 Message-ID: <200910261906.58614.denkenz@gmail.com> In-Reply-To: <1256577664-4459-1-git-send-email-andrew.zaborowski@intel.com> List-Id: To: ofono@ofono.org --===============7398251705421960590== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Hi Andrew, So I took this patch and the two earlier patches and applied them. Thanks. > Instead of setting DataConnectionManager.Powered to false after > failed attempts to register to home PLMN with packet radio when > RoamingAllowed is false, wait for the home PLMN to become available > retrying everytime the circuit switched radio manages to register > to home PLMN. The downside is if the circuit switched radio is off, > we wait indefinitely. So I decided to simplify things even further. Before we can attach we must= be = registered to the network (e.g. COPS issued), otherwise attach fails. Sinc= e = we need netreg for this anyway, we might as well always rely on netreg's = registration status to base our attach / detach decisions on. To me this seems to be the most sensible way to approach this. However, th= is = assumes that the modem can't be registered on netreg and roaming on gprs. It also might break down in low-signal conditions where packet radio has = signal but circuit-switched radio doesn't (which is interesting case where = COPS reports an operator, CGREG shows registered, but CREG shows unregister= ed) Comments? Regards, -Denis --===============7398251705421960590==--