Starlink has bufferbloat. Bad.
 help / color / mirror / Atom feed
From: Dave Taht <dave.taht@gmail.com>
To: starlink@lists.bufferbloat.net
Subject: [Starlink] a bit on reddit
Date: Wed, 21 Jul 2021 14:42:23 -0700	[thread overview]
Message-ID: <CAA93jw60btZbDu+Q+S4KG0yBDtLN=g+7SaQDXDfwk1=wf1f=RQ@mail.gmail.com> (raw)

... despite my hope in one day hitting the front page of washington
post with something like

"Starlink achieves 5ms latencies under load on the rrul test! Gamers Rejoice!"
or "comcast cuts latencies by 8x by using AQM!"

I've grown weary and cynical today about anyone ever deeply
understanding what a consistently low latency internet could really
mean for the world's productivity or "how we could restore the edge to
innovation!" or other glowing, hippie-dippy, bright futurist thoughts
I too often have.

I just wasted the morning on reddit putting in some good links to
older hard to find material here:

https://www.reddit.com/r/Starlink/comments/okmx3x/bufferbloat_and_beyond_talking_with_dave_taht/

I need to go off and work on one of my main (paid) projects and some
grant proposals for the rest of the week. For anyone interested: nlnet
has 50k available to just about anyone that can write a good one, ICSI
is looking for new blood, and ampr is also looking for suitable
non-profits to apply for nearly anything that applies to wireless. If
you got something good that fits into https://www.ampr.org/giving/
submit soon!

The FOSS show did cut our starlink segment down to a digestible size:
Fixing Starlink's Latencies: https://www.youtube.com/watch?v=c9gLo6Xrwgw

And they also extracted my cluebat segment:
https://www.linkedin.com/feed/update/urn:li:activity:6823375976098074624/

And for the record, yes, I did have an email exchange with elon in
2013 (via dan whaley) about icei.org and bufferbloat, I'd worked on
the mothballed spacedev's trailblazer sat's linux-rt and usb
subsystems in 2004 - (I have a few spacedev, and for that matter
rotary rocket stories I should tell), and yes I was bitter about the
falcon 1's 3rd launch that needed a slightly longer timeout (I figure
a single shift left) on stage sep (with enough bugs all eyeballs are
shallow), and when I interviewed at spacex in 2012 after we'd first
succeeded with fq_codel and our project was essentially bankrupt (I
was sleeping on paul vixie's couch), I had a choice between working on
bufferbloat or calling spacex back to do embedded there (and didn't),

and I do my best to make it sound fun and funny now. But it wasn't. A
lot of it wasn't. I'm now on listening to the third season of
"hitchiker's guide to the galaxy", and
that's helping me a lot!

I'm glad to see all the new members here, and we hope to have some new
sch_cake data to report by the end of the week. I was enjoying the
speculation on how to route
stuff well around starlink but haven't time to speculate, please keep
at it! cya sunday or so!

--
Dave Täht CTO, TekLibre, LLC

                 reply	other threads:[~2021-07-21 21:42 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: https://lists.bufferbloat.net/postorius/lists/starlink.lists.bufferbloat.net/

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to='CAA93jw60btZbDu+Q+S4KG0yBDtLN=g+7SaQDXDfwk1=wf1f=RQ@mail.gmail.com' \
    --to=dave.taht@gmail.com \
    --cc=starlink@lists.bufferbloat.net \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox