Advertisement · 728 × 90

Posts by Mark Kop

Curio

I've always been a big fan of @everynoise.bsky.social and revisited it after a few years and found out this amazing Curio project everynoise.com/curio.html that @glennmcdonald.bsky.social created.
I remember I played around Every Noise and Spotify API some time ago
github.com/Markkop/disc...

7 months ago 1 0 0 0
tl;dr
'use client' is a strongly-typed <script>.
'use server' is a strong-typed fetch().

tl;dr 'use client' is a strongly-typed <script>. 'use server' is a strong-typed fetch().

by common request there will be a tldr

11 months ago 84 2 5 0

React compiler e Turbopack em Prod é um sonho que tenho já tem uns 2-3 anos

11 months ago 0 0 0 0

New Claude LRM en route!!!

This is gonna be *great* for coding in Cursor!

1 year ago 1 1 0 0

    The GPT-4 barrier was comprehensively broken
    Some of those GPT-4 models run on my laptop
    LLM prices crashed, thanks to competition and increased efficiency
    Multimodal vision is common, audio and video are starting to emerge
    Voice and live camera mode are science fiction come to life
    Prompt driven app generation is a commodity already
    Universal access to the best models lasted for just a few short months
    “Agents” still haven’t really happened yet
    Evals really matter
    Apple Intelligence is bad, Apple’s MLX library is excellent
    The rise of inference-scaling “reasoning” models
    Was the best currently available LLM trained in China for less than $6m?
    The environmental impact got better
    The environmental impact got much, much worse
    The year of slop
    Synthetic training data works great
    LLMs somehow got even harder to use
    Knowledge is incredibly unevenly distributed
    LLMs need better criticism
    Everything tagged “llms” on my blog in 2024

The GPT-4 barrier was comprehensively broken Some of those GPT-4 models run on my laptop LLM prices crashed, thanks to competition and increased efficiency Multimodal vision is common, audio and video are starting to emerge Voice and live camera mode are science fiction come to life Prompt driven app generation is a commodity already Universal access to the best models lasted for just a few short months “Agents” still haven’t really happened yet Evals really matter Apple Intelligence is bad, Apple’s MLX library is excellent The rise of inference-scaling “reasoning” models Was the best currently available LLM trained in China for less than $6m? The environmental impact got better The environmental impact got much, much worse The year of slop Synthetic training data works great LLMs somehow got even harder to use Knowledge is incredibly unevenly distributed LLMs need better criticism Everything tagged “llms” on my blog in 2024

Here's my end-of-year review of things we learned out about LLMs in 2024 - we learned a LOT of things simonwillison.net/2024/Dec/31/...

Table of contents:

1 year ago 651 148 28 46

Ei, #bolhadev.
Se você for num meetup em 2025, quais tópicos de desenvolvimento você gostaria de ver/saber mais?
Responde aqui em reply plz, se já tiver algum pode dar like
cc @samsantosb.bsky.social @sseraphini.bsky.social

1 year ago 2 4 0 0

In 2010, anti-copyright activist Aaron Swartz hanged himself after being prosecuted to the full extent of the law for making copyrighted academic publications freely available. Today, academic publishers are insisting those copyrights be lifted to feed the corporate AI slop machine.

1 year ago 3264 1658 44 50
Advertisement
Preview
My Top Cursor Tips (v0.43) I've been using Cursor extensively for 4 months in both professional and hobby ways, mostly on...

Alô #bolhadev!
Acabei de criar um blog post com as minhas melhores dicas sobre o Cursor! Saca só:
dev.to/heymarkkop/m...
cc @samsantosb.bsky.social @sseraphini.bsky.social

1 year ago 7 4 0 0

Acabei de testar bolt new e replit agent depois de uns meses e eles tão bem melhores.

Mas ainda não conseguem ter o contexto inteiro da codebase e pensar o suficiente pra sugerir mudanças mais estruturais em vez de if/elses ou outras pequenas alterações.

Quanto tempo até lá?

1 year ago 0 0 0 0

Vale a pena pagar o ChatGPT pra várias coisas. No mínimo, vale o teste. Pra treinar inglês acho que pode ser bom, mas sinto falta de um serviço mais avançando pra isso. Já usei o Elsa Speak e era bem bom

1 year ago 1 0 1 0

Sistema brasileiro supremacy

1 year ago 3 1 0 0
Login • Instagram Welcome back to Instagram. Sign in to check out what your friends, family & interests have been capturing & sharing around the world.

GENTE fiz um post no Instagram falando sobre semântica e acessibilidade, mostrando na prática qual o impacto de ter um código bem organizado com a tags certas

também mostrei uma nova forma de usar o devtools pra ver a aba de acessibilidade, vai la ver ✨

#bolhadev

www.instagram.com/p/DDXfER0x91...

1 year ago 2 3 0 0
Preview
Ian Neves Um canal dedicado à informação, ao debate e à análise política.

Link youtube.com/@iannevesofi...

1 year ago 0 0 0 0

Confere @ianneves.bsky.social, melhor conteúdo que encontrei nos últimos meses

1 year ago 0 0 1 0
Video

Naming things is hard. If you don’t know what to call a component in your design system, you can start by looking at what other teams are doing.

Component Gallery is a great resource for this: component.gallery/components/

…But don’t assume that something is correct, just because it’s popular.

1 year ago 135 35 15 4

My greatest pain after building a nextjs project without vercel is the deployment time change from seconds or a few minutes to ~10 minutes every time.
Kinda breaks, or delays, the reward loop system

1 year ago 1 0 0 0
Advertisement

Todos esses títulos são contextuais ao ambiente que você está, especialmente dependendo do tamanho da empresa. Em uma empresa grande, você vai ter papéis muito especializados, em empresas menores, você vai ter gente fazendo de tudo porque ou faz ou as coisas não acontecem. Não se apeguem a nomes.

1 year ago 8 2 0 0

Testei o Replit Agent ontem e hoje e assim: é muito massa, é meio que o agent mais avançado e faz muita coisa. Mas ainda é meio burro e não entende o que é pra arrumar e é ruim em refatorar.
Acho que ainda estamos 1-year-away de ter um agent que preste pra codar

1 year ago 0 0 0 0
Post image

Testing v0's "Composer" mode

1 year ago 1 0 0 0

Eu tô surfando d++ na onda de IA. Minha produtividade tá off charts, entrego task pra krl, flexibilizo melhor as horas de trabalho no meu dia, tô usando e testando ferramentas de IA a rodo.
Sorte de estar numa empresa que me permite trabalhar assim

1 year ago 0 0 0 0
Post image

O @eltonminetto.dev falando de decisões aqui no meetup da @codecon.dev

1 year ago 3 0 0 0
To make sure, I'll look up the spelling of "strawberry" in a dictionary. Yes, it's spelled S-T-R-A-W-B-E-R-R-Y, with three R's. So, despite the pronunciation sounding like two R sounds, the actual spelling contains three R's.

To make sure, I'll look up the spelling of "strawberry" in a dictionary. Yes, it's spelled S-T-R-A-W-B-E-R-R-Y, with three R's. So, despite the pronunciation sounding like two R sounds, the actual spelling contains three R's.

the DeepSeek R1 model is wild

does it actually have access to the dictionary? does it just have an overactive imagination?

1 year ago 18 2 3 1
Post image

Estou chegando @codecon.dev

1 year ago 0 0 0 0

Será que no futuro será possível pegar e pedir pra ia algo tipo "faça um jogo completo, porém simples, sobre a cultura e origem de Florianópolis"?
Seria uma forma muito massa de aprender sobre algo jogando um RPG que você não sabe direito o que esperar

1 year ago 0 0 0 0
Post image

Every day, I go to bolt new and send a few prompts within my daily token limit in my hobby project.
Not worth to pay for yet

1 year ago 0 0 0 0
Post image

I hate you @v0

1 year ago 0 0 0 0
Advertisement

@eltonminetto.dev No fim, qual To Do list app você escolheu?

1 year ago 0 0 1 0
Post image Post image Post image Post image

algumas fotos (lindas) da Select Experience ontem em SP

1 year ago 18 3 1 0
Post image

Olha ai na mesma foto um gerente do Google que gosta de programação funcional, o autor de um dos meus blogs preferidos ( @eltonminetto.dev ) e o mestre do PHP.

(não achei os outros @)

#codecon

1 year ago 31 3 1 0

Muito boa conversa com @coprodu.to e @jessilyneh.bsky.social sobre performance de equipes de software.

Galera mandando muito muito bem na @codecon.dev

1 year ago 7 2 1 0