John Lindquist

26.1K posts

John Lindquist banner
John Lindquist

John Lindquist

@johnlindquist

Codex Power User Workshops - https://t.co/xPtDG0Inyh Previously @vercel AI DX The OG egghead - https://t.co/rxAeU2DHfm

home Katılım Temmuz 2008
2.2K Takip Edilen28.6K Takipçiler
Sabitlenmiş Tweet
John Lindquist
John Lindquist@johnlindquist·
At the request of many, I've put together a "Claude Code Power User" 🤓workshop on Aug 1st. We'll cover: - Automated workflows (cron, pr, keyboard shortcuts, etc) - Claude Hooks - The SDK - Subagents - Containers - Orchestration Want ULTIMATE POWER? Sign-up here 👇
English
4
4
68
36.2K
John Lindquist
John Lindquist@johnlindquist·
Finally watched Pluribus. Always love how Gilligan explores morality. It's nice to have a show where you can chat with others about "what would you do in that situation?" and there are no right or wrong answers.
English
0
0
6
709
John Lindquist
John Lindquist@johnlindquist·
@kenwheeler If the perspective shift helps, many people would say the same about running a 5k
English
1
0
2
94
patagucci perf papi
patagucci perf papi@kenwheeler·
i actually don’t know where i stand on this. i’ve been shocked a number of times when i thought grown men could do 135. on the other hand i’ve seen dudes that look like video game bosses and absolutely mog the shit out of me.
English
5
0
11
1.7K
John Lindquist
John Lindquist@johnlindquist·
@aarondfrancis @wesbos @ryanvogel I want to resurrect all my old iPads, phones, and any old wifi-enabled hardware and link up their cameras/mics/etc to do something crazy. I just don't know what 🙃 (Other than some useless, artsy-fartsy talking piece)
English
0
0
1
75
Aaron Francis
Aaron Francis@aarondfrancis·
Codex is currently trying to reverse engineer the Amaran Bluetooth mesh network protocol. It told me to buy this board, now it's flashing it with something I have no idea what it's doing but it's doing a great job
Aaron Francis tweet media
Aaron Francis@aarondfrancis

Codex was trying to install something that needed sudo and since it's running in Solo it just spawned a new terminal, sent it the install commands, and waited for me to type my password. Pretty cool flow!

English
22
1
188
37.8K
John Lindquist
John Lindquist@johnlindquist·
@debs_obrien @ryancarson Would you elaborate on "helping people grow"? Is the expectation that everyone is self-motivated and should develop their own skills?
English
0
0
0
39
John Lindquist
John Lindquist@johnlindquist·
@aarondfrancis @wesbos @ryanvogel Frickin love this. These are the sorts of AI unlocks that I wish more people realized was possible. We're going to need a new term for "Internet of Things" that describes making connections you never thought possible thanks to agents.
English
1
0
2
140
Aaron Francis
Aaron Francis@aarondfrancis·
@wesbos @ryanvogel github.com/aarondfrancis/… boom, done. You can import a backup of your iPad and it will read the mesh keys and set it up to join your existing mesh network instead of creating its own. Now you can have iPad + CLI control!
English
2
0
8
1.9K
Frederic Souvenir
Frederic Souvenir@FredDSouvenir·
gmcli v0.1.4-alpha Google Messages for OpenClaw It pairs via QR, syncs texts into local SQLite, supports read-only search/query, and includes an Agent skill heavily inspired by @steipete wacli and uses mautrix-gmessages/pkg/libgm AGPL-3.0 github.com/fdsouvenir/gmc…
English
1
0
0
78
Jack Wotherspoon
Jack Wotherspoon@JackWoth98·
@johnlindquist @shengzheyao My future is still undecided. But probably not, I'll be helping folks migrate from Gemini CLI and doing my best to make things smooth during the 30 day transition period. After that who knows.
English
4
0
11
395
Jack Wotherspoon
Jack Wotherspoon@JackWoth98·
Antigravity CLI v1.0.2 shipped over the weekend 🚀 Run /changelog within the CLI to see all the changes. The release is mostly improvements based on feedback since launch🙏 P.S. @shengzheyao is the eng lead for Antigravity CLI, definitely give him a follow to stay up to date.
Shengzhe@shengzheyao

Antigravity CLI 1.0.2 is out! - Hide your email & plan tier by setting ⁠AGY_CLI_HIDE_ACCOUNT_INFO=1⁠. - Fixed a bug where main agent tool permissions timed out after 60s. - Improved ⁠/help⁠ to display full shortcuts. Next up: documentation & examples. Thank you all for providing feedback! antigravity.google/docs/cli-getti…

English
21
31
356
39.6K
sunil pai
sunil pai@threepointone·
hmm (for tomorrow)
sunil pai tweet media
English
7
1
46
15.4K
John Lindquist
John Lindquist@johnlindquist·
@DavidWells I'm honestly just grateful he shares as much as he does. gogcli, bird, etc are all massive time savers for me.
English
0
0
0
72
John Lindquist
John Lindquist@johnlindquist·
I have more ideas than tokens
English
5
1
15
1.9K
John Lindquist
John Lindquist@johnlindquist·
@ThePrimeagen Some of my most impressive work was scrapped by execs moments before release. I try not to get too attached.
English
0
0
1
321
ThePrimeagen
ThePrimeagen@ThePrimeagen·
I have written so much software in my life, half the things I thought were throw away ended up sticking around for years, and half the things I thought were permanent were thrown away within the month. I largely think the reason why we see so much crap software is because people think software they are writing will only be around for a moment.
English
129
49
2.4K
120.9K
David K 🎹
David K 🎹@DavidKPiano·
This. Understand the code you're responsible for. "My AI agents one-shotted this app overnight and I didn't read any code and just shipped it" is not impressive; it's irresponsible.
Lee Robinson@leerob

You might believe you should spend less time thinking about code because of AI. I strongly disagree! We’re watching this play out live where tons of AI generated code becomes a liability. At the end of the day, an engineer needs to be responsible / on call for code that gets shipped to production. If you don’t understand the system you’re trying to debug, you’re probably going to have a bad time. Yes, AI can help with all of this, if you set up the proper systems. You can have agents triage prod logs, look at errors, etc. You can speed up parts of the investigation, but an engineer needs to make the call. There might be serious customer or financial implications from that change. I expect the trend continue for trimming dependencies, vendoring code so you can modify it directly, preferring simpler systems with fewer abstractions, and spending waaaay more time thinking about system design and code maintenance. I’ve said this before, but it’s a great time to get familiar with CS fundamentals and some of the history behind what great software looks like. Many parts will be different in the coming years as AI progresses, but also a lot more than people realize will stay the same.

English
19
14
187
14.1K
vogel
vogel@ryanvogel·
@trq212 Would be interesting to see how certain models could decompile that code and see how well they could extract the meaning for different programs and processes
English
3
0
10
1.3K
Thariq
Thariq@trq212·
my main takeaway from the Bun rewrite is that legacy codebases will be incredibly valuable as a source for "distilling" code into new forms every game should be crossplatform, all legacy software should work on the web, we don't need COBOL anymore
English
114
57
1.7K
142.8K
John Lindquist
John Lindquist@johnlindquist·
agreed. I'm mostly hypothesizing about a future where agents replace modules wholesale. Instead of refactoring code, you drop in an agent that assumes the role of the legacy code's intent. This obviously depends on a lot of things going right (especially perf). But we've proven that agents can replace IDEs and I fully believe that trend will continue into every nook and cranny of software. I recognize I'm uber-optimistic with this stuff, but I'm just having so much fun building things I never thought possible.
English
0
0
0
144
Sarah Drasner
Sarah Drasner@sarah_edo·
Too complex for a tweet to do justice- yes you can modularize pieces and leverage agents, there are certain tasks they excel at (creation of unit tests for instance)- but back to Theo's original point- the scale and determinism I'm talking about necessitates that codebases we work in today still matter. What you won't see in a mini-demo: we have to be intentional about massive refactors, have extremely good test and conformance checks, benchmarking, etc. e2e is still too non-deterministic, for instance across many millions of lines of code. We can leverage AI for across the SDLC for many different tasks! But we're not in YOLO mode.
English
1
0
4
446
Theo - t3.gg
Theo - t3.gg@theo·
A lot of people are building with the assumption that the codebases we work in today will still matter next year. I’m not sure if that’s the case.
English
66
67
2.4K
248.7K
John Lindquist
John Lindquist@johnlindquist·
@LyalinDotCom @Disney It's an interesting conundrum. Darth Vader, Yoda, Death Star, etc are such icons that it's impossible to do them justice. The stakes of the original were already set to maximum, so only spin-offs/lore/etc work. I don't know what a timeline reset would look like
English
1
0
1
244
John Lindquist
John Lindquist@johnlindquist·
Oooh, I love this conversation. I find discussions around Agentic Architecture fascinating. I fully believe that every module of a monolithic codebase could be represented by an agent that handles/heals any fail scenario. APIs evolve into natural language bridges between task-optimized agents. Instead of refactoring code, you drop in agents that infer intent from the legacy module. This is mostly wishful thinking based on mini-demos today, but holy crap things are moving fast.
English
1
0
2
567
Sarah Drasner
Sarah Drasner@sarah_edo·
It really depends on the scale and amount of existing userbase you have. I completely agree on smaller apps. But something with millions of lines of code where any non-deterministic refactor could cause major outages, we’re not at all there. “Everything is greenfield” depends on your brownfield.
English
4
1
99
10.1K

Keşfet

@kenwheeler @aurorascharff @aarondfrancis @wesbos @ryanvogel @debs_obrien @ryancarson @FredDSouvenir