RARE user and assistance email list

[mc36 <>]

hi,

On 2/19/21 2:50 PM, Simon Leinen wrote:
> Few small items about the IETF slideset:
>
> * Maybe a few words on what GEANT (the project) is before starting with
>    "WP6-T1 sub-task..."?
tbh i would appreciate if somebody from the team could join me and
present the geant and rare project slides, just because what i can
tell them about these, from my perspective, is maybe not even crosses
the official standpoint: geant is here to connect us, nrens together
and provide us cheap(er) internet, and rare is here to create a
dataplane and a playground to hack that on and destroy it...:)
so if anybody does not agree with the above, i warmly welcome
him/her as as a co-presenter on both events.... :)))))))


> * On the "targets" slide, "Tofino" is almost invisible next to the bold
>    "NPU".  And I wouldn't call the Tofino an NPU (well, technically it's
>    easy to say it *is* an NPU, but the P4 community avoids that term,
>    probably because NPUs have gotten something of a bad name in the past.
>    I would probably just write "Intel/Barefoot Tofino on Wedge ..." or
>    maybe "...Tofino switching silicon on ..."
thanks for pointing out, will update asap!


> "Key take-way" -> "Key take-away"
thanks again!

> In my experience, what an IETF WG really wants to hear from implementers
> of a new protocol is which implementation obstacles you found and how
> you surmounted them.  So you may want to expand on the "ECMP pitfall"
> item.
my plan is to stay longer on the two technical slides, and that ecmp
issue is something i have to explain them a bit longer, i just written that
there to not to forget to talk about... but it's a bit bigger thing that it
could be written to ppts slides... but agree that a drawing would help here
to explain them the issue... and if they won't get it, i'll detail them the
observations to the mailing lists to start the conversation...


> (Of course they will also appreciate feedback of the form "I found the
> BIER standard very readable and was able to implement everything just
> from reading the text.  But in general the IETF culture is based more on
> constructive criticism than on positive encouragement, for better or
> worse :-)
i already told them my opinion about the text when it was just a draft
and i had to read it throughly to implement it... i wont repeat that as

it's out now... anyway i just found one thing, the endian-ness question
is even not touched at all, and when asked about it, some said that as
usual, msb, nothing to see here, some agreed to my suggestion to clarify
it a bit... but finally they did not merged the sentences... and it turned
out that we interwork so it's crystal clear, at least for us and junos...:)

regards,
cs