Connection details ------------------ * Date: October 17th 2018 * Time: 8-9AM DST, 17:00 CEST: https://www.worldtimebuddy.com/?qm=1&lid=2988507,1816670,5391959,5128581&h=2988507&date=2018-10-03&sln=17-18 * Webex Link: https://cisco.webex.com/ciscosales/j.php?MTID=m7933acfc5befd0ac211e03ee678e5a1d * Meeting number: 203 710 782 * Meeting password: JM3Y2Kqf * To join by phone or other means, please go to the Webex Link for details. Agenda ------ * [17:05] Administrivia [ 5min] o Note-Well, Scribes, Agenda Bashing o Status of drafts * [17:10] SCHC Updates since last Interim - Dominique [15min] * [17:25] Open discussion on Fragmentation [25min] * [17:50] AOB [ 0min] Minute takers ------------- * Dominique Barthel Orange * Carles Gomez * Arunprabhu Kandasamy * Ivaylo Petrov Attendees --------- * Ana Minaburo * Laurent Toutain * Pascal Thubert * Alexander Pelov * Carles Gomez * Dominique Barthel * Juan Carlos Zuniga * Ivaylo Petrov * Arunprabhu Kandasamy * Vincent Audebert Past Attendees --------- Action Items from last time ---------------------------- * Chairs: find reviewers for drafts Minutes ------ * [17:05] Administrivia [ 5min] o Note-Well, Scribes, Agenda Bashing No objection, so minutes of the last interim meeting are approved. Doc cutoff Oct 22nd. Meeting slot Tuesday 9h-11h local time. Please request slots. Dominique on Hackathon: starting new project using micropython. openSCHC. Splitting the API to be delegated easier to other people for long term. Web: www.openschc.net Package existing code with clean architecture. No LPWAN connectivity is promised at the hackathon. Just use UDP. Only 2hrs this time. Rechartering o Status of drafts * [17:10] SCHC Updates since last Interim - Dominique [15min] Dominique presenting: Author list modified, included Juan Carlos. Comments on the commits welcomed. Haven't published the new version yet. 3 work sessions between authors and contributors. 27 commits since last interim. commits has log messages. new ACK-on-Error mode, normative text and FSM drawing. Supports varying MTU and out-of-order delivery. Alex: confident about Out-of-order delivery gauranteed ? Dominique: side effect of the design. Still NoACK relies NO OOO Delivery. Alex: It is a very nice feature, maybe it's worth nothing in the text. Juan Carlos: receiver can bank the fragments, but it doesn't mean that we are encouraging the sender to do OOO. Laurent: When you are doing retransmissions, you are doing OOO (out of order) Dominique: In NoACK mode, each successive data is assembled by order that it received. In Ack-On-Error, because of retransmission, using FCN, W, we can do OOO. Upto the implementation to restrict it if needed. Juan Carlos: Sender doesn't know about the reciever's freedom to support OOO. More text , may be profile, that clarifies about the OOO. Dominique: Each mode says if it supports OOO delivery.No longer in the abstract, introduction as we had before. Introduces profiles. Parameters will be defined in profile. specfic to technology, device class under technology. which ruleIDs denoted to ACK-on-ERR, etc. Definition of Tiles, Windows with the introduction of ACK-on-ERR. When MTU changes, we have retransmission messages transmitted in pieces of data. FCN decoupled from Tile number. FCN and Tile number are not same. eg: all-1 fragment. Ack-always new text still to be done? Single ACK request replaces All-1 and All-0 ACK requests. It has window field, the receiver can tell to which window it belongs. Single ACK format, C-bit is present always. It is called integrity check bit. Improved text and drawings. Tiles must be contiguous. Must be of same size on ACK-on-ERR except the last fragment. NoACK shall have varying tile size when MTU changes since it doesn't have retransmission. JCZ: I did not understand this part while reading the text. Juan Carlos: suggest to make tile description as minimum denominator supported by the technology. will read again and come back. Dominique: protocol is much more flexible now. common tool box, and deriving tool box for each mode. Multiple tiles should be ordered. Sequential and concurrent process. Until all-1 fragment is received, the receiver doesn't know that it is receiving last window. Until integrity check succeeds, receiver has no idea that it has all tiles. Last tile's size is not known. size is infered from schc frag payload size. extreme small payload:when no room for tile and MIC. send MIC in separate fragment. ** HEADACHE** please fill in here? Next steps: discussion to solve the headache. charlie's answere to be included. rewording ack-always. oct 22nd ietf103 cutoff. Laurent: Optional MIC to be taken to other document. needs more knowledge about this behaviour. Dominique: have the MIC mandatory and leave the profile define MIC size, MIC polynomial. Could be as small as 1 byte and not computed at all. But MIC stays mandatory in SCHC. Worried that IESG argues about MIC! ** leaving now ** please fill in the minutes after it. * [17:25] Open discussion on Fragmentation [25min] * [17:50] AOB [ 0min]