From mboxrd@z Thu Jan 1 00:00:00 1970 Content-Type: multipart/mixed; boundary="===============4598636137695472992==" MIME-Version: 1.0 From: Denis Kenzior Subject: Re: IPV6 question Date: Thu, 29 Sep 2016 11:07:28 -0500 Message-ID: <57ED3C40.5020705@gmail.com> In-Reply-To: <6C3C91E4-F4C0-41B8-B6F0-B4A9484BBF90@gnumonks.org> List-Id: To: ofono@ofono.org --===============4598636137695472992== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Hi Harald, On 09/29/2016 10:01 AM, Harald Welte wrote: > This might be off topic, but AFAIR there are PPP type PDP contexts where = not raw IP, but actual PPP frames are transmitted over GPRS inside SNDCP/LL= C. Not sure if any operators use that, but its perfectly valid. > Yes, I think this was introduced in Release 7. However, it would = require the use of "PPP" pdp type in CGDCONT, not "IP". Also, for EPS = contexts I don't believe that "PPP" type is possible (nor wanted, as = you'd fry your CPU keeping up with the high-bandwidth connection). = Probably a legacy feature that might still be used in very narrow = circumstances. I've certainly never seen it used in practice. > In this case PPP doesn't terminate in the modem, so "hardware" support sh= ould be irrelevant. > True. But I'm probably missing your point. PPP packet compression is = still an optional protocol feature, so it should not preclude = interoperability. Regards, -Denis --===============4598636137695472992==--