Doc Cottle!
Doc Cottle!
That's a sad looking refrigerator
gotta get that Proteini
Grimly humorous
the all-stick and no-carrot administration
Espresso Proteini is a genius name
www.usatoday.com/story/money/...
King
The spice must flow
damn... I really gotta stop listening to my gut
Hopefully it's about the ballroom
Ahhhhh
Oh no, not the pre-Texas baseline!
and they are already saying they are going to do it to Cuba in a few weeks. Insanity.
So Philip Glass just joined TikTok. He only has like 4 posts and theyβre simply him in what I assume is his living room playing the piano. First comment on one I saw was βyour piano is too close to the fireplace, youβll dry it outβ and if thatβs not the internet in a nutshell I donβt know what is.
thatβs some real 1890s shit
the cast of Star Trek Voyager in street clothes awkwardly posing with late-90s computer hardware and consumer electronics
this is what it feels like to post on bluesky
Col. Jessep did say that though. Easy to confuse the two.
The folks happily walking into moving traffic is pretty realistic though
The Linear MCP server seems to need re-authentication at least once a week. Seems aggressive.
The talking points have been disseminated
maybe Dodge Rams can be converted to run on "beautiful clean coal"?
bsky.app/profile/chri...
This guy isn't even worried about bringing harmony and peace to the United States
I really don't think I should have to verify that I am human to look at the Cheesecake Factory menu online. Dogs and robots should be able to look at 57 types of cheesecakes too.
in my personal experience, Claude Code with Opus is best at following my directions and they seem fine tuned to work together. I use Codex and Gemini occasionally, but they are more prone to make big unexpected changes that I then I have to reverse.
in the coding communities I follow, Opus 4.6 and the latest Codex models from OpenAI are considered the best for programming tasks. But preferences seem to change month to month
Claude Code lets you pick the model, and you have usage limits for Opus and Sonnet that depend on your account level and/or willingness to pay per token.
This is the exact guidance I gave to the team!
The rules of PRs are unchanged. If you submit a bad PR - buggy, untested, too big, poorly documented, poorly isolated to the task - then you did that as an IC on the team, and you're gonna lose points on employee battlechart & have to do a hunger games
Similar thing happened to me, I found a brand and size that fit, and ordered a bunch of stuff on eBay for 1/3rd the price