Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
didip
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
15 ms
·
271.
▲
by
didip
3y ago
Go is both simple and easy. For people who knows both, I don't know why you would choose Python (except if your domain doesn't have the libraries, like scientific/AI/ML stuff).
272.
▲
by
didip
3y ago
Since KVM already exists, who would renew their ESXi licenses? Broadcom would surely rise license costs while at the same time disinvesting. Also, what will happen with VMWare’s Kubernetes investments? I am guessing all of their open source
273.
▲
by
didip
3y ago
Let’s be real here. At the end of the day, what matters more is commercial success and a big payout. AGI is still very far away and the fear mongering is nothing but PR stunt. But the devs need their big payout now. Which explains the mutin
274.
▲
by
didip
3y ago
Stability.ai, please make sure your board is sane.
275.
▲
by
didip
3y ago
I wonder if 1 of the board would love it if customers switched to Poe.
276.
▲
by
didip
3y ago
If I were OpenAI employee, I would have been uber pissed. Imagine your once-in-blue-moon, whatsapp-like, payout at $10m per employee evaporated over the weekend before Thanksgiving. I would have joined MSFT out of spite.
277.
▲
by
didip
3y ago
What will happen to employee’s stock options if they all mass quit and moved to Microsoft? The options will be worth $0, right?
278.
▲
by
didip
3y ago
It’s not hard to imagine why relevant people want him to stay. They all want to get the big payday, including Microsoft. People are in the startup game for the big payday. And Sam Altman is the best person given where OpenAI is at right now
279.
▲
by
didip
3y ago
Sam needs to ink a deal with Netflix to tell the story of this saga.
280.
▲
by
didip
3y ago
I still don’t understand DuckDB use case. To handle huge amount of data, you need a large database that’s connected to the network. DuckDB will eventually hit a ceiling.
281.
▲
Ask HN: Why Datadog succeed but New Relic failed?
7 points
by
didip
3y ago
|
14 comments
282.
▲
by
didip
3y ago
The post is super light on details, it's hard to visualize if it's worth it or not. For examples: - How much data are they working with? What's the traffic shape? Using NFS makes me think that they don't have a lot of da
283.
▲
by
didip
3y ago
I like the spirit of the article. My personal preference is server side templating using as much default HTML as possible with JS as progressive enhancement (using HTMX or something similar).
284.
▲
by
didip
3y ago
I don't understand authors who criticized the acquirer post acquisition. There is only one reason why you would sell: lots of money. You understand this going into the transaction. Once the company is acquired, it's no longer your
285.
▲
by
didip
3y ago
Why side income? Why not focused on maximizing the salary of your primary job?
286.
▲
by
didip
3y ago
It used to haunt me to the point where I was worried about the longevity of my career. It took many many years and several steps to address it. 1. My walking posture was bad with flat feet. So I addressed it with physiotherapy to change the
287.
▲
by
didip
3y ago
To quickly solve some of your pain, pay a VPN node in America and just proxy requests through it. Small amount of money compared to the stress.
288.
▲
by
didip
3y ago
You should submit your benchmarks to ClickBench.
289.
▲
by
didip
3y ago
I prefer writing a normal server-side with Go template and use HTMX on the client side.
290.
▲
by
didip
3y ago
The number on the performance improvements are pretty good. I like this release.
291.
▲
by
didip
3y ago
Micro anything is only useful if your company is big and has a lot of separate departments. It solves a real organizational issues. If your company is small, don't bother with any of these micro trends. The overhead of micro architectu
292.
▲
Ask HN: Are we heading towards World War 3?
10 points
by
didip
3y ago
|
8 comments
293.
▲
by
didip
3y ago
Project management is hard because people are unpredictable. Each person heard the same thing and understood differently. It's probably best to pick and choose a few tools that are complementary and focus on the team building itself.
294.
▲
by
didip
3y ago
I think the perfect no-code tools has yet to appear. Ideally, no-code tools should generate a good code (using LLM) into a git repo complete with the following: - good code structure. - test code. - database migration code. - CI/CD cod
295.
▲
by
didip
3y ago
Did you go with D7 visa or golden visa?
296.
▲
by
didip
3y ago
Does Motherduck have distributed storage, query, and caching layers?
297.
▲
by
didip
3y ago
Will humanity go extinct if our birth rate continues to plummet?
298.
▲
by
didip
3y ago
I am surprised people reduced this article as just another pet rock. Dude's journey has some interesting nuggets. He knows how to time box himself and build the smallest MVP that will actually make money. And his journey is not an over
299.
▲
by
didip
3y ago
This looks great, I can put it on my kid.
300.
▲
Ask HN: Is LLM advanced enough to generate optimized database indices?
1 points
by
didip
3y ago
|
0 comments
More ›