baq

18 ms·

661.
by baq 7mo ago
the beauty of it is there's not much to show; I use a crude jjui approach where I have an octopus merge working tree commit (in command line terms, jj new PR_A PR_B PR_C) and either use native jj absorb (S-A in jjui) which guesses wher
662.
by baq 7mo ago
it's actually git that makes you think backwards - in jj the working tree is a commit, in git it isn't until you at least stage it. the working tree being a commit has wide ranging implications as all the commands that work with c
663.
by baq 7mo ago
does trivially working on 3 PRs in a single checkout and pushing focused changes to each one independently without thinking twice count? if you don't need this, you might not see any value in jj and that's ok. you might use magit
664.
by baq 7mo ago
flying really close to the dropbox comment, sir
665.
by baq 7mo ago
if I can't read TFA because of its formatting it isn't tangential
666.
by baq 7mo ago
yeah reader mode it is, didn't know it's different on Linux than on Windows and the support article listing it is here: https://support.mozilla.org/en-US/kb/keyboard-shortcuts-perf...
667.
by baq 7mo ago
+1 Firefox users: press F9 or C-A-R
668.
by baq 7mo ago
The problem is implementing anything approximately twice is a hard sell… this is no longer true, though - TLA+ models are cheap now. You should be using them when writing any sort of distributed systems, which is basically everything nowada
669.
by baq 7mo ago
Just when I’ve gotten used to having 3 or more PRs in parallel with a local octopus working tree with jj. Maybe my colleagues will see the light at least.
670.
by baq 7mo ago
There’s no closed source software anymore, clankers are mighty good at decompiling.
671.
by baq 7mo ago
> One loss sets you back 10-20 wins. didn't look at the numbers, but this one sentence reminds me of selling options for 'passive income' (don't do that)
672.
by baq 7mo ago
Linux is missing good vm defaults (dirty_bytes etc.) - out of the box settings on the distros I tried are abysmal; both windows and macos are much saner. Other than that, yeah, it's a royal pain in the ass. It's treating the user
673.
by baq 7mo ago
jane street is always hiring!
674.
by baq 7mo ago
at least you won't burn $200 of tokens in an hour now
675.
by baq 7mo ago
You can vibe code minor fixes to some annoyances including the clanker managing the whole fork/pull request flow if you want to contribute back for $20/mo on codex or claude (though $20 is the free trial tier there, codex is nearl
676.
by baq 7mo ago
Dear sir, please think of the shareholders, they need a fair exit.
677.
by baq 7mo ago
Pretty sure there’s at least a couple orders of magnitude in purely algorithmic areas of LLM inference; maybe training, too, though I’m less confident here. Rationale: meat computers run on 20W, though pretraining took a billion years or so
678.
by baq 7mo ago
Technically all ships crossing the strait matter as their cargos end up being bought, sanctioned or not, Iran-friendly or not; they wouldn't have crossed otherwise. If we're talking about avoiding a global recession and worst case
679.
by baq 7mo ago
Remote MCP solve the delivery and update issues just like saas and browsers did for human users. Not much more to it really
680.
by baq 7mo ago
LLMs are essentially pure functions.
681.
by baq 7mo ago
AIS over there is jammed and spoofed. Nobody wants to get bombed and certain Greek and Chinese owners run the strait dark with their crews apparently very happy about the danger pay.
682.
by baq 7mo ago
You need to send an analyst there with binoculars, a box of cigars and $15k in cash to get realtime trustworthy data
683.
by baq 7mo ago
It's a fallacy to think all insiders are big players. Small fish want a piece, too, and are not constrained in the same ways the big ones are.
684.
by baq 7mo ago
> TL;DR: We tested Anthropic Mythos's showcase vulnerabilities on small, cheap, open-weights models. They recovered much of the same analysis. AI cybersecurity capability is very jagged: it doesn't scale smoothly with model siz
685.
by baq 7mo ago
Futures are sometimes informed hours earlier than Twitter was the point
686.
by baq 7mo ago
Nobody is forcing anybody to place the bets, but if you see credible money flows into eg. Middle East cease fire bets, you can decide to fill up your tank today, or not. We can’t think for you here, if you don’t see anything useful in the p
687.
by baq 7mo ago
On prediction markets there are plenty of inefficiencies; e.g. an MM could take both sides of the bet when odds don’t add up to 1, which apparently happens more often than we all think
688.
by baq 7mo ago
Come to think of it, what insider trading in oil even is…? Oil isn’t a company and doesn’t have any capacity for material non-public information. The biggest players are all insiders on this market anyway. Why should Barron be prosecuted an
689.
by baq 7mo ago
Exactly. Barter is super insanely expensive in comparison with the most manipulated paper markets; thus, if you closed the current markets, the participants would be very highly incentivized to create a new one, which would likely be worse
690.
by baq 7mo ago
Who do you think places those bets

More ›