How to achieve AGI in 2024:
1. Define a benchmark with puzzles and call it "The AGI Testing Benchmark."
2. Fine-tune a VLM to solve these puzzles.
3. Declare the AGI achievement.
Meet the new iPad Pro: the thinnest product we’ve ever created, the most advanced display we’ve ever produced, with the incredible power of the M4 chip. Just imagine all the things it’ll be used to create.
@santibanezdani Es lamentable sobre todo para los nuevos candidatos. Algunos de mis colegas han considerado irse a trabajar temporalmente a Colombia o Perú para no perder el trabajo.
Two web app testing approaches:
1. Emulated. (Typically via @jestjs_ or @vitest_dev with jsdom or happy-dom)
2. Real browser. (Typically via @playwrightweb or @Cypress_io).
Here are the tradeoffs:
Emulated:
✅ Typically faster since it avoids overhead of running a full browser.
🚩 It's fake, so may not behave like the browser.
🚩 It's incomplete. Only mimics a subset of the browser's features.
🚩 It's harder to debug failing tests. We can't watch the test run in a real browser.
🚩 Not ideal for end-to-end testing, since it doesn't run in a real browser. So team needs to learn and support both emulated and real browser testing tools.
Real browser:
✅ Easy to debug. Can "see" the test run in the browser.
✅ It's real. We have more confidence and power when our test runs in a real browser.
✅ Team can use one testing tool for all tests (e2e, integration, single component, etc).
🚩 It's typically slower than emulated since running a real browser adds overhead.
Any other tradeoffs you've noticed?
How can you detect a BS vendor (consultant): they use generic words like:
"(...) mission to solve/fix a problem"
vs
"We fix your bathrom drainage problem using plumber's snake, auger, and/or a high-pressure water jet to clear the blockage."
@RangeTester Lo que más destaco de TestCafé en nuestro proyecto:
- Corre en cualquier navegador (sin ningún hack).
- Soporta ejecución en diferentes ventanas o tabs.
- Soporta iFrames.
Lo malo:
- Es lento comparado con Cypress.
- La librería nativa de assertions es limitada.