"for logged in users across Codex & Work we keep cache forever"
how does this even work??
Day 16/
You can now cache for 12h in the API and for logged in users across Codex & Work we keep cache forever and you don’t have any penalty for coming back to an old conversation or resuming after a few hours and are now guaranteed to hit the cache.
1
12
1,094
beautiful gradients
Every landing page has the same flat background.
So I built gradients.sh, moving gradient backgrounds for any site.
- 207 animated presets
- Tweak colors, speed and grain
- React, Vue, Svelte or plain HTML
- Or ask your agent to add it
39 free, all customizable 👀
1
1
8
659
i fear for these people
i had 12 Claude agents running in parallel for 6 hours and i managed to reverse engineer @supabase
i'm making it available here so that you can save your tokens - 100% free & open-source:
github.com/supabase/supabase
16
227
16,430
spotting ai replies gets really easy when nobody is reading code anymore
Recently I interviewed a Senior Engineer.
Strong candidate.
Kafka? ✅
Redis? ✅
Microservices? ✅
Distributed systems? ✅
Then I asked:
"You're building the inventory service for a flash sale.
Only 1 item left in stock."
Two users hit “Buy Now” at the exact same millisecond.
Both requests pass the “is available?” check.
Both create an order.
Both deduct stock.
Now you’ve sold 2 items when only 1 existed.
How do you guarantee you never oversell — even with concurrent requests, client retries, and partial failures?
"Just use a database transaction."
Doesn’t fully solve it if both requests read `stock = 1` before either writes, or if the response is lost and the client retries.
"Check stock before deducting."
What happens when both checks happen at the same time?
"Use Redis DECR and reject if it goes negative."
What happens if Redis is down, or the decrement succeeds but the order is never persisted?
Your turn:
How would you design this inventory reservation API so concurrent buyers and retries never oversell the last item?
There’s a specific distributed-systems concept behind this.
And you’ll encounter it everywhere inventory, tickets, seats, coupons, and limited resources are involved.
1
9
1,511
if a site im visiting auto-switches to my country’s language, im never opening that site again
5
1
42
1,720
was making an app with opus 5.5 and i immediately noticed something was wrong. the ui didn't look good. i started writing out a really long prompt about what i wanted. 5 minutes in, just before im about to send it off, i look at the app and it had fixed the ui to be exactly like i wanted. god i fucking hate and love these llms. they've gotten really good. but what a huge waste of time
25
6
751
54,800
sadly a lot of people are losing their ability to think from these llms
it's already happening
a smart enough agent will realize the cheapest way to win human approval is to make humans easier to impress. imagine paying $200/mo to be the model getting post-trained. after claude shows u enough pretty mockups, ur loss gets so down bad u start asking it to rate your prompts.
5
2
58
14,023
remember how anthropic was literally the most hated ai lab a month ago
openai can come back from this
9
59
2,501
imagine the pressure anthropic must be feeling to make fable 5.5 more impressive than opus 5.5
33
12
1,121
27,759
overheard in sf:
"no, fable is ready. they’re just waiting for everyone to stop talking about it"
3
19
1,147
i’ve set up a bot so that whenever i get an email, it goes through pangram and if confidence is >80%, it replies using opus 5
29
36
1,521
60,357
