leonidasv

7 ms·

61.
Why a $500 Steak Dinner Only Yields a $25 Profit(wsj.com)
5 points by leonidasv 10mo ago|1 comments
62.
by leonidasv 1y ago
It costs up to ~$9m per year, processing 63 billion transactions per year: https://oglobo.globo.com/economia/noticia/2024/10/03/bc-gast... Banks pay a small fee (less than a cent) per batch of trans
63.
by leonidasv 1y ago
Which models do you use instead of Anthropic ones? I've only tried Claude Code with an external model once (Kimi K2) but it performed poorly.
64.
by leonidasv 1y ago
Most consumer are just like that. I'm the one who don't bother ordering directly to get better value, but I started noticing less and less restaurants (that I order from) having this option anymore. They probably saw too little pe
65.
by leonidasv 1y ago
> A browser is not that special anymore and there are plenty of alternatives All Chromium forks. Browser engines are special. Firefox is the only non-WebKit derivative with relevant market-share currently.
66.
by leonidasv 1y ago
SVG started as an open competitor to Shockwave/Flash Player and also an application format for PDAs. It almost got networking support once.
67.
by leonidasv 1y ago
Question: ChatGPT voice mode seems to have too much tolerance for mispronouncing. Sometimes, it understands you even you mispronounce something in a phrase, and it's not aware enough to correct you - it even says your pronunciation is
68.
by leonidasv 1y ago
I always saw MCPs as a plug-and-play integration for enabling function calling without incurring API costs when using Claude. If you're using the API and not in a hurry, there's no need for it.
69.
Deciphering Stuxnet, the Most Menacing Malware in History (2011)(wired.com)
7 points by leonidasv 1y ago|0 comments
70.
Deadlocks in Go: the dark side of concurrency (2021)(craig-wood.com)
20 points by leonidasv 1y ago|5 comments
71.
by leonidasv 1y ago
I somewhat agree, but I think that the language example is not a good one. As Anthropic have demonstrated[0], LLMs do have "conceptual neurons" that generalise an abstract concept which can later be translated to other languages.
72.
Bing Search API to be retired on August 11th(azure.microsoft.com)
3 points by leonidasv 1y ago|0 comments
73.
by leonidasv 1y ago
> Our performance evaluation shows up to 2.7% overhead for the microcode mitigation on Alder Lake. We have also evaluated several potential alternative mitigation strategies in software with overheads between 1.6% (Coffee Lake Refresh) a
74.
by leonidasv 1y ago
https://web.archive.org/web/20250513173440/https://peyre.42w...
75.
by leonidasv 1y ago
I use the same setup, works like a charm.
76.
by leonidasv 1y ago
Fun fact: adjusted for inflation, that's equivalent to $2,017.82 in today's money [0] [0] https://data.bls.gov/cgi-bin/cpicalc.pl?cost1=595&year1=1982...
77.
by leonidasv 1y ago
They're giving out 1 month free if you're paying for their IDEs already. I've tried it last year and it was very limited, not "agentic". Now they've launched an agentic version called Junie and also gave anothe
78.
by leonidasv 1y ago
Not exactly answering your question, but as a user of JetBrains IDEs the Windsurf Plugin with Cascade[0] for JB IDEs is the best solution I've found so far to get a good agentic AI integration without giving up JetBrains goodies. [0]
79.
by leonidasv 1y ago
The thing is: we should not need standalone editors just to use AI coding agents. They could be just plugins, but Microsoft does not want to bend the plugin API enough for that. Windsurf has a "plugin edition" for JetBrains IDEs t
80.
by leonidasv 1y ago
Location: Sao Paulo, Brazil (GMT-3) Remote: Yes (only) Willing to relocate: No Technologies: full-stack web development (back-end biased), distributed systems, Go, Typescript, Kotlin, Node.js, React, Next.js, Python, gRPC, Postgres Résumé&#
81.
by leonidasv 1y ago
Apparently you can use a P.O. Box as address for this purpose[0] when registering for AppStore, which is substantially cheaper. However, Reddit says Google does not accept P.O. Boxes [1], so the only option is a "virtual" office a
82.
Fujihack Doom Port (2023)(fujihack.org)
1 points by leonidasv 1y ago|0 comments
83.
by leonidasv 1y ago
LLMs are great for exercising skills, especially ones with a lot of available data in the training corpus, such as leet code. The prompt below, put in the System Instructions of Gemini 2.5 Pro (using AI Studio) summons the best leet code te
84.
by leonidasv 1y ago
The assistant is limited to a 32k context window (consisting of up to 7+1 messages) and usage quotas varies depending on model pricing[0]. Although I'm a Kagi user, I don't see much value in it, especially now that virtually all L
85.
by leonidasv 2y ago
The periods of my life I was the most productive and learned the most were when those when I managed to sleep >=8 hours every single night, in a dark, silent environment. I still manage to sleep >=7 hours most nights, but I feel like
86.
by leonidasv 2y ago
The CPF (persons) and CNPJ (companies) numbers has two check digits.
87.
by leonidasv 2y ago
The UX is standardized by the Central Bank and they mandate a step showing the name of the person/company associated with that Pix key before you make the payment. So you can always double-check.
88.
by leonidasv 2y ago
> How much do you trust your government with your money? When you use Pix, your money is not in custody of the government. The money goes to/comes from a standard commercial bank, just like a wire transfer. What the government (Cent
89.
by leonidasv 2y ago
Well, Pix is not free from those too... Besides being operated by the Central Bank, you still have to use your commercial bank account to send/receive money and, even though the Central Bank do require some minimum UX implementation st
90.
by leonidasv 2y ago
For market research, Exa.ai may be a better option. It's a search engine that crawls the web and index pages in LLM embeddings. For my searches, it delivers good results that are not found in Google's first page because the SEO sp

More ›