Pinned Tweet
I've decided to log off Elon's wild ride. I wasn't using it actively but it will suck to lose my timeline. I just don't want to even be a passive part of whatever the hell he's doing over there.
In case you needed to see the inside of a Texas Instruments LISP machine today: wtf.hijacked.us/~andrew/lisp…
Might do a thing this weekend:
wtf.hijacked.us/~andrew/94yj…
wtf.hijacked.us/~andrew/94yj…
wtf.hijacked.us/~andrew/94yj…
wtf.hijacked.us/~andrew/94yj…
wtf.hijacked.us/~andrew/94yj…
It does, remarkably, fit. Now to pull it all out again to finish the welding and paint everything I can get access to.
Saturday was 1 year to the day that my family and I moved to Maine, never having visited the state before. It has been a great success and we love it here. Here's some random pictures of stuff I did over the last year:
wtf.hijacked.us/~andrew/main…
Andrew Thompson retweeted
just released erlang-rocksdb v1.6.0, the #rocksdb key-value database binding for #erlang & #elixirlang applications. This version brings the support of OTP 23, last stable version of rocksdb 6.13.3, compression options and other fixes: gitlab.com/barrel-db/erlang-… Enjoy!
I don't know if anyone needs an implementation of Forgetful Bloom filters (dprg.cs.uiuc.edu/docs/fbf_ca…) for Erlang but I tried my hand at one anyway: github.com/Vonmo/erbloom/pul…
Andrew Thompson retweeted
update: i have been informed that Sun did it and now Oracle has a patent on the technique
fuck patents, and fuck everyone who defends patents with a rusty shovel
That magical time in a dog's life when you can pick them up by the rope toy they refuse to relinquish...
The last time I dealt with Google's v8 was 5 years ago, it seems to *still* have wildly unstable APIs and a crazy build system, just EVEN MORE OF BOTH
XOR filters are pretty cool things: lemire.me/blog/2019/12/19/xo…
Andrew Thompson retweeted
No big deal. Just a @helium
LongFi-enabled GPS tracker being picked up by **12** Hotspots in NYC and NJ. #PeoplesNetwork #IoT
Andrew Thompson retweeted
It's a time to move away from (bounded) synchrony assumption if one is building secure distributed (>2 parties) protocols.









