Thomas Smith retweeted
Hyper-scale data centers have a strong need for optimal RPC performance. In this @netdev01 0x17 talk Xiaochun Lu and Zijian Zhang from @BytedanceTalk do a comparative study of gRPC's performance on Homa versus over TCP. Who is prettier? Come and find out #netdevconf
Thomas Smith retweeted
The more of Silicon Valley I’m exposed to the more I’m convinced that 40% of companies that raise money are fraudulent, 58% are real but just never yield returns, and the remaining 2% carry everyone on their backs.
Thomas Smith retweeted
from time to time I think about this post, which I think remains in the top 0.01% of everything I have ever read
Thomas Smith retweeted
Replying to @MarcJBrooker
queue.acm.org/detail.cfm?id=… is used in many places, outside of packets. Uber, for example...
Thomas Smith retweeted
- Engineer: I invented this new thing. I call it a ballpen 🖊️
- TwitterSphere: OMG, people could write horrible things with it, like misinformation, propaganda, hate speech. Ban it now!
- Writing Doomers: imagine if everyone can get a ballpen. This could destroy society. There should be a law against using ballpen to write hate speech. regulate ballpens now!
- Pencil industry mogul: yeah, ballpens are very dangerous. Unlike pencil writing which is erasable, ballpen writing stays forever. Government should require a license for pen manufacturers.
Thomas Smith retweeted
Insanely cool ring buffer trick: abhinavag.medium.com/a-fast-…
Thomas Smith retweeted
While stuck in a "backlog grooming" meeting reddit.com/r/programmerhumor…
Thomas Smith retweeted
When designing APIs, a good angle is "make it easy to use, hard to misuse". I sometimes think of this quote on why it's hard to design a trash container that's bear proof:
“There is considerable overlap between the intelligence of the smartest bears and the dumbest humans.”
Thomas Smith retweeted
tired of code review? just write the majority of your shit as a library on github and pull it in as third party
Thomas Smith retweeted
In Big Tech companies, technical decisions are often more nuanced than many engineers realize and are not always based on pure technical merit. Here's a list of things I've actually seen and had to deal with.
🧵
people talking about all the "maintenance" of running servers but my experience is it takes 10x as long to figure out why the nth Lambda in some bizarre chain invoked by api-g failed when I could've just looked at a stack trace in a log
23
Thomas Smith retweeted
my contrarian opinion is that doing things you don't want to do is incredibly bad for you. all the advice that says to grind is bullshit
you just end up teaching yourself that you don't value your intrinsic motivations, and over time those vanish, leaving you empty
Thomas Smith retweeted
What is ignored by the media -- but will be studied by historians?
8 more examples:
Thomas Smith retweeted
Meetings are the worst!
They burn people out and nothing gets done.
My tweet on meetings and burn out last week hit a nerve. It resonated with a lot of folks.
These are 7 steps I follow to aggressively reduce meetings.
A short 🧵
Thomas Smith retweeted
The food in the US is poisonous.
I wish it wasn't. I'm from the US and spent most of my life there.
And I'm not sure exactly what it is, but I'll share what I do know.
Thomas Smith retweeted
At @usenix NSDI 2023 today @WilliamSentosa presented DChannel, a clever way to accelerate interactive latency-sensitive apps up to 40% in mobile settings by combining two channels – one high bandwidth, one low latency – with fine-grained packet steering: usenix.org/conference/nsdi23…
Thomas Smith retweeted
awkward truth:
engineering productivity is far more correlated with culture than team size
there are startups w/ 10 engineers that ship more for their customers than big co engineering teams of 1000 (ex: Midjourney)
you cannot throw people and/or money at building great things





















