pier25

9 ms·

151.
by pier25 4mo ago
> Vinyl literally has less than half the dynamic range of CD A lot less than half. It's around 20-30db and every 10db is a factor of 10. The CD has between 100-1000x more dynamic range.
152.
by pier25 4mo ago
Every time I do something that's not supposed to be permanent until I have time to implement a proper solution it lasts for years.
153.
by pier25 4mo ago
So a company which has access to practically unlimited tokens and their best models makes crappy software. Huh who would've thought? /s
154.
by pier25 4mo ago
> Shipping tens of megabytes per web page is impolite, if not outright disrespectful to users. You're being generous with what I would consider negligence.
155.
by pier25 4mo ago
The vast majority of global users are not using iPhones.
156.
by pier25 4mo ago
The consequences of MBs of JavaScript can be perceived by anyone in terms of performance and mobile data consumption.
157.
by pier25 4mo ago
You definitely need other factors but societal passion seems like the cornerstone. The US has everything except this. Personally couldn't care any less about sports. I say this as an external observer.
158.
by pier25 4mo ago
It's the culture. I'm from Spain and living in Mexico. Football in these countries is almost like a religion.
159.
by pier25 4mo ago
> There's guaranteed persistence There's not. Browsers can delete "persistent" storage at any time. https://developer.mozilla.org/en-US/docs/Web/API/Storage_API...
160.
by pier25 4mo ago
Honestly you can POST to an API to the other side of the world and receive a response in less than 300ms. That's literally the blink of an eye. There are very few use cases where you'd need even lower latency than that while at th
161.
by pier25 4mo ago
For native apps this is less of an issue since they have access to persistent storage but with browsers there's no guaranteed persistence.
162.
by pier25 4mo ago
> Webpack - while maybe a win over Grunt/Gulp, was MASSIVELY complex. and slow
163.
by pier25 4mo ago
yeah SSR is still marked as experimental after like 3-4 years
164.
by pier25 4mo ago
nobody wants it
165.
by pier25 4mo ago
> Maybe they should invest money in competing with Vercel on UX/DX That's exactly what they are doing.
166.
by pier25 4mo ago
This is just my own impression but I feel that Evan might have distanced himself from Vue to focus on Vite and Void. IIRC Vapor mode was spearheaded by someone else. Same with Alien signals. To be clear, I don't think this is bad. Vue
167.
by pier25 4mo ago
afaik Void Cloud never went GA
168.
by pier25 4mo ago
Weird situation for Vue. The Nuxt guys and Eduardo (creator of vue-router, pinia, etc) are working at Vercel while Evan is now at Cloudflare.
169.
by pier25 4mo ago
> theres no evidence that they arent (or can't) use profitable inference to subsidise those other expenses as far as we know there's no evidence that they can produce any profits at all
170.
by pier25 4mo ago
AI companies have more expenses than inference.
171.
by pier25 4mo ago
The MSRC situation is really unbelievable. There are probably better sources but I think this video by The Primeagen is a good introduction. https://www.youtube.com/watch?v=9kxx5xp5nTQ
172.
by pier25 4mo ago
I owned a Surface Book v1 for a couple of days before returning it. It was a garbage device with tons of issues. I will say it had amazing keyboard though.
173.
Starbucks Abandons Borked AI Inventory Tool That Couldn't Count(gizmodo.com)
32 points by pier25 5mo ago|10 comments
174.
by pier25 5mo ago
support is not uniform though eg: https://jsr.io/@std/dotenv
175.
by pier25 5mo ago
yeah I wish Node had something like that
176.
by pier25 5mo ago
Many runtimes/languages rely on third party deps for that. Also plenty of devs think the stdlib should be as lean as possible. Personally, I think there should be a balance. The direct consequence of a barebones stdlib is NPM and havin
177.
by pier25 5mo ago
> And I suspect it's probably higher within Anthropic That probably explains why their uptime and reliability are so bad.
178.
by pier25 5mo ago
Yes, when he made that claim back in 2024 they were spending like $100M to train a model.
179.
by pier25 5mo ago
Yeah I'll believe it when I see it. Revenue is increasing but so are their costs. Back in 2024 their CEO claimed training costs would rise to $10-100B in the next years. https://www.tomshardware.com/tech-industry/a
180.
by pier25 5mo ago
For example, security. See all the supply chain attacks from the past couple of years.

More ›