Hi Rodney,

Thanks for the feedback.
We will certainly take these notes into consideration in the next version.

Can you please elaborate which protocols use the 80-bit timestamp format
(other than PTP, obviously)?

Thanks,
Tal.


On Mon, Sep 25, 2017 at 6:01 PM, Rodney Cummings <[email protected]>
wrote:

> Hi Tal,
>
> This is an excellent draft. I have a couple of suggestions.
>
> 1. Add the PTP 80-bit Timestamp, since there are protocols using that as
> well, some with hardware support.
>
> 2. In section 4, add informative text to explain one of the benefits of
> specifying a small set of fixed format specifications: hardware support.
> The lower down in the interface stack that timestamping occurs, the more
> accurate and precise the timestamp. This is well known for PTP, for which
> timestamping is often done at the MAC or PHY level. Hardware support is
> potentially a reason to avoid the control field mentioned in section 7,
> because hardware typically cannot support "any" format... ideally this RFC
> will narrow it down to 2-3 at most.
>
> Rodney
>
> From: TICTOC [mailto:[email protected]] On Behalf Of Tal Mizrahi
> Sent: Sunday, September 24, 2017 8:38 AM
> To: [email protected]; [email protected]
> Subject: [TICTOC] Guidelines for Defining Packet Timestamps
>
> Hi,
>
> We have revised the draft based on the comments received in IETF 99, and
> based on the comments from Yaakov (thanks Yaakov).
>
> https://urldefense.proofpoint.com/v2/url?u=https-3A__
> datatracker.ietf.org_doc_draft-2Dmizrahi-2Dintarea-
> 2Dpacket-2Dtimestamps_&d=DwMFaQ&c=I_0YwoKy7z5LMTVdyO6YCiE2uzI1jjZZ
> uIPelcSjixA&r=WA71sf2o7Dw7CbYhFt24DPjt3lJuupswWYdnboKbZ8k&m=
> yUZsj6L0ib9J4wu3kPNa1OQJK4MJPBl2vMca6Rpvx_Q&s=xdFj3xTP-9EZqma_
> 84T6eWQ0p0VlwMrdXil6FFU3FW0&e=
>
> The main changes compared to the previous version of the draft:
> - We have extended the discussion about the factors that may affect the
> choice of the timestamp format.
> - A new section has been added, called "Timestamp Use Cases".
> - The sychronization aspects have been separated from the timestamp
> format, allowing the timestamp format to be independent of how time is
> synchronized.
>
> Any further comments are welcome.
>
> Cheers,
> Tal, Joachim, and Al.
>
_______________________________________________
TICTOC mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/tictoc

Reply via email to