[Tech-board-discuss] Nomination: Dave Taht for TAB

[Tech-board-discuss] Nomination: Dave Taht for TAB

[Dave Taht]

My name is Dave Taht, and I am running for the TAB.

I am best known for my role in the Bufferbloat Project, rearchitecting
the Linux WiFi stack[1], contributing to multiple qdiscs (codel, fq_codel,
pie, fq_pie and sch_cake), and helping nudge along many other latency
reducing improvements to network stacks in
general, not just Linux. My chief success in this project was making
fq_codel (RFC8290) the default in most Linux distributions, OSX, and
iOS. I have been involved in embedded Linux since 1998.

I often feel a heavy burden of responsibility for so many packets
flowing through the algorithms I helped design and propagate, and an
even larger one, for so many packets that aren't, particularly those
in common home routers and edge devices. In this new age of everyone
working from home and endless videoconferencing, every time I see
someone else have a glitch, jitter or lag, I wish I could have done
something more, particularly for WiFi.

My beta noir is better home routers, still, and in finding ways to
accelerate the IPv6 deployment so the increasingly limited, CGNATed,
IPv4 based internet can finally be replaced.

In recent years, sanguine my work was mostly done in Linux...  I'd
moved onto multiple standardization efforts the IETF.

A few months back, I had a wake-up call.

I had an encounter with the SpaceX Starlink router, and "dishy", the
bug report story entertainingly (I hope) retold on Doc Searls' podcast:

https://www.youtube.com/watch?v=c9gLo6Xrwgw

... where they were shipping 9 year code without any of the improvements
and bug fixes of my project, and of the 10s of thousands of other
contributors to the Linux kernel.

In the months since I've done a survey of the embedded Linux world,
and I am dismayed at what I've found. Everywhere I look now in the
embedded Linux world the future seems frozen in place, with ancient
kernels locked to proprietary offloads, and once common courtesies
like a GPL drop, or an open bug reporting system and user engagement,
having fallen by the wayside.

In joining the TAB, I would like to return to my Linux roots, and the
Linux Foundation to return to its roots, for us all to work together
again to re-establish and re-invigorate efforts with the vendors and
community towards sharing well maintained, common codebases, to re-improve
vendors open source processes, improve developer outreach and education,
and most importantly find ways to shorten the cycles between
development and deployment, to improve response times to CVEs and
other bugs, in an age where Linux is ubiquitous and critical
infrastructure for the entire world.

[1] https://lwn.net/Articles/705884/

-- 
Fixing Starlink's Latencies: https://www.youtube.com/watch?v=c9gLo6Xrwgw

Dave Täht CEO, TekLibre, LLC