kburman

7 ms·

61.
by kburman 11mo ago
I don't know why you would say juniors handle this better. From what I've seen, juniors who ship fast often do so by ignoring domain complexity and testing against a single "happy path" prompt. Seniors take longer becaus
62.
by kburman 11mo ago
These types of tests are fundamentally flawed. I was able to create perfect clock using gemini 2.5 pro - https://gemini.google.com/share/136f07a0fa78
63.
by kburman 11mo ago
You're trying to reframe this as a factual error about budgets and timelines, but you're missing the core argument. > (1) They spend more on browser development now... Spending more to achieve less is an indictment of efficienc
64.
by kburman 11mo ago
Mozilla has lost the plot. People support Mozilla because they want a strong, independent browser, not so staff can siphon money into side projects that exist mainly to look good on someone’s resume. Things like Transformer Lab have nothing
65.
by kburman 11mo ago
“Open source” doesn’t mean “independent.” Most of those labs are state-linked and operate under laws that require compliance with party policy. The CCP plays a long game, they want dependency, not donations. Once enough people adopt their s
66.
by kburman 11mo ago
Not defending anyone, but by that logic, you shouldn’t be using any Chinese models either.
67.
by kburman 11mo ago
There's a massive gap between that pattern and the real-world complexity of billing. It's too much to cover in a comment, but this link explains the actual nightmare - https://www.getlago.com/blog/why-billing-
68.
by kburman 11mo ago
Just because it's in their production code doesn't mean it's not a ticking time bomb.
69.
by kburman 11mo ago
This is a perfect example of something that looks good in a demo but fails in a real product. Business logic and 'packages' are never this clean or simple. Putting this kind of type-based 'magic' in the code is a bad dec
70.
by kburman 1y ago
For the correlation idea, you might take a look at how Sentry does it, they rely mostly on stack traces, error messages, and pattern matching to map issues back to code areas. It’s cheap, scalable, and doesn’t need an LLM in the loop, which
71.
by kburman 1y ago
It’s an interesting direction, but feels pretty expensive for what might still be a guess at what matters. I’m not sure an LLM can really capture project-specific context yet from a single PR diff. Honestly, a simple data-driven heatmap sho
72.
by kburman 1y ago
This looks very promising. Are there any downsides or potential gotchas?
73.
by kburman 1y ago
Phasing out Rosetta 2 seems like a reasonable move. Maintaining backward compatibility indefinitely adds complexity and technical debt. Apple has supported Intel-based systems for a long time, and this step aligns with their goal of keeping
74.
by kburman 1y ago
I get the appeal of moving away from Microsoft, but in my experience, Nextcloud is extremely bloated even for personal self-hosting. I wonder how well it will scale in a government setup.
75.
by kburman 1y ago
Any reason to not include native application given how important mobile experience was.
76.
by kburman 1y ago
I thought https://playground.cognition.ai/ was just returning some cached query results, but no, they’re actually spinning up real VMs and running live queries without any authentication or restrictions. That must be costin
77.
by kburman 1y ago
That's fair, but then the framing matters. The article criticizes serverless architecture rather than acknowledging an evaluation failure. "Serverless was fighting us" vs "We didn't understand serverless tradeoffs&q
78.
by kburman 1y ago
You're confusing network proximity with application architecture. Edge deployment helps connection latency. Stateless runtime destroys it by forcing every cache access through the network. The whole point of edge is NOT to make latency
79.
by kburman 1y ago
What's valuable about rediscovering that stateless architectures requiring network round-trips for state access are slower than in-memory state? This isn't new information, it's a predictable consequence of their architecture
80.
by kburman 1y ago
The takeaway here isn’t that serverless doesn’t work, it’s that the authors didn’t understand what they were building on. Putting a latency-critical API on a stateless edge runtime was a rookie mistake, and the pain they describe was entire
81.
by kburman 1y ago
I was referring specifically to popular embedding models like OpenAI’s and sentence-transformers, which (as far as I know) don’t reliably handle negation or emotional nuance, they mostly capture topical similarity. I don’t know enough of th
82.
by kburman 1y ago
I recently learned that semantic search embeddings mostly represent topics and concepts, but they don’t handle negation or emotion very well. For example, if you search for “paintings of winter landscapes but without sun and trees,” you’ll
83.
by kburman 1y ago
It feels like one user-hostile move after another from Microsoft lately. Forcing Edge, using dark patterns to block Chrome, resetting telemetry with updates, the whole Recall fiasco... and now this. It's exhausting. My working theory i
84.
by kburman 1y ago
An employee of Turso, a commercial fork of SQLite, is presenting a standard, safety-first feature of SQLite's WAL as a dangerous flaw. As many have noted, this behavior prevents database corruption, it doesn't cause it.
85.
Wealth Shown to Scale(hmijail.github.io)
5 points by kburman 1y ago|0 comments
86.
by kburman 1y ago
I’ve watched the demo video, and gone through the discussion here but I’m still not sure what the core use case for Darklang is. It feels like I'm missing something obvious. Can someone explain the practical problems this is solving be
87.
by kburman 1y ago
Exactly. It's Virtual Thread, not Magic Thread. Just because they’re lightweight doesn’t mean you can throw out all notions of flow control or resource management.
88.
by kburman 1y ago
Its a design problem not virtual threads problem.
89.
by kburman 1y ago
I only use LLMs later in the development stage to review my code and check if I’ve missed anything. I don’t use them for implementing features, that just feels like a waste of time.
90.
by kburman 1y ago
I'm not sure I fully understand the purpose of this tool. Modern IDEs are quite good at detecting these types of issues. And if they don't catch something, I often just copy and paste the method with some context-into an LLM and a

More ›