MDX allows to write markdown with JSX. Nice :) I think I will use it for a new version of my blog written in svelte.

and a big post about markdown: the bad of it for developers and content editors. The feature of markdown are the one's needed in 2000, but not for the modern web. There is no designated successor to markdown., but some needed features are mentioned in the post.

A open AI should have the code open + the data used to train it

TODO: texte à recopier

After looking at the available options I definitely prefer number 2 in the list above: the name of the language as text in the language itself, possibly followed by the name of the language in the language of the current page.

After thinking about it, this is what is the most explicit, even if it is not shiny 👍‍

Summary:

• Use Boyer-Moore (and unroll its inner loop a few times).

• Roll your own unbuffered input using raw system calls. Avoid copying
  the input bytes before searching them. (Do, however, use buffered
  output. The normal grep scenario is that the amount of output is
  small compared to the amount of input, so the overhead of output
  buffer copying is small, while savings due to avoiding many small
  unbuffered writes can be large.)

• Don't look for newlines in the input until after you've found a match.

• Try to set things up (page-aligned buffers, page-sized read chunks,
  optionally use mmap) so the kernel can ALSO avoid copying the bytes.

The key to making programs fast is to make them do practically nothing. ;-)

TL;DR;

MySQL is a pretty poor database, and you should strongly consider using Postgres instead

TL;DR;

I want to support the creators that I care about to earn a living through advertising, but remembering to whitelist all of these is a faff.

I changed my mind recently, however, when browsing the web without an ad blocker became unbearable.

I’d estimate that the amount of resources I downloaded from ads far outweighed the actual webpage content on the sites I visited. In a world where we all need to seriously conserve resources and minimise the carbon footprint of our sites, ads and trackers should be the first things to go.

I’m aware that the conversation around monetisation on the web is far from simple. But this state of affairs is surely serving no one.

So I think the right path is something like this:

• Try to generalise from your experiences, but don’t hold your opinions too strongly.
• Listen to other people’s conclusions, but try to learn as much as you can about the context that formed them.
• See the value in expertise and approaches that have a limited scope of application.

*de l'architecture

28 red flags listés

and a start to improve the situation.

Critiquer Rust sous prétexte que c’est un langage complexe conduit à passer à côté de ce qui fait sa force : il est pensé pour être expressif, ce qui signifie qu’il y a beaucoup de fonctionnalités, et dans la majorité des cas c’est ce que vous attendez d’un langage de programmation.

L'article est bien complet !

Mon avis: utiliser Go pour le code qui doit être déployé demain, Rust pour le code qui devra rester en production tel quel pour les 5 prochaines années
—Grzegorz Nosek

Au niveau de la liberté des licences:

WTFPL > MIT > Apache

Rendre le code libre mais limiter sa distribution est un contre-sens.

Je comprends parfaitement l’idée de RMS d’obliger les distributions d’un code libre a rester libres tout au long de chaîne : ça permet à toutes les modifs d’êtres publiques et disponibles par tous.

Mais cette obligation de liberté est en soi un privation d’une liberté (celle de faire du non libre). Si ton code était réellement libre, cette liberté ne me serait pas retirée.

Quand on est vraiment pour la liberté, on doit pouvoir accepter que son code soit utilisé pour des choses qu’on ne cautionne pas, ou distribués d’une façon qu’on ne cautionne pas.

Je ne dis pas que tout le monde devrait faire ça. Je dis juste que tout code qui ne respecte pas ça n’est pas véritablement libre.

I bookmark this post especially for this code snippet:

body {
    margin: 0 auto;
    max-width: 40rem; // can be bigger but not too much ! (<= 80 I think)
    font-size: 1.2rem;
}

It makes a HTML website directly readable :)

+1

Use git to push and pull to get content, share them.

The author compares git repositories with the RSS feeds: a git repository is more versatile.

text is small.

👍‍

Les arguments relevés semblent pertinent:

• les tests passent avant le typage → ok mais pourquoi ne pas avoir les deux et laisser ce choix par projet ?
• la lisibilité des en-têtes de fonction peuvent très bien être améliorée
• les annotations requirent leurs propres imports, ce qui peut conduire à des imports circulaires (cf: https://stackoverflow.com/questions/22187279/python-circular-importing/22187343#22187343)

1. ... true at some point
2. IDK
3. TRUE. I see it between Typescript and Javascript. If you know how TS works, then
4. I don't have enough experience about this to know if it is right to me
5. Except with regulations
6. True, but the API seems to be easier and more powerful over time. You can do now things that are more powerful than before.
7. pairing +1
8. YES
9. Maybe. I am currently satisfied with the new cli tools (fd, exa, bat, nvim ...)
10. Maybe
11. They can be improved
12. True as the companies does not want to get involved. But how to blame if your products are going to be more expensive than others.
13. True. We need regulations. The GDPR in Europe or the accessibility laws in the US lead the way.

Un point de vue plus que justifié sur le mode de financement du logiciel libre.
Ce billet fait suite à la faille dans la bibliothèque Log4j.