141 private links
Ranger les mots par index permet de décharger entre la demande de l'utilisateur et sa réponse.
Tout est dans l'image, plus efficace qu'une recherche humaine dans un dicitonnaire par ordre alphabétique (approximativement une recherche dichotomique): https://www.camilab.co/content/images/2021/12/dictionnaire.png
Cela se fait au détriment:
- du temps de construction de l'index au départ → à faire qu'une seule fois
- de l'espace de stockage pour cet index → OSEF car le stockage est peu coûteux
- l'accès au disque sera toujours moins rapide que conserver les données en mémoire -→comment compresser l'index ?
A case study
Do not compress the data directly but creates a program that can reproduce the data 💥
As an example: Aaprogram able to calculate the Fibonacci sequence is a "compressed" representation of the Fibonacci sequence. We replace the data by code, code capable of reconstructing the data.
They already manage to gain 20% on APKs (Android applications) with Superpack compared to the usual zip compression of APKs.
about tree data structures and their diversity
Example of morse and UTF-8 decoder, word counter, coroutine and generators, and a unicode-case folding state machine
How to use hexagonal grids. It is complete ! Wow well done.
Everything is here to create a game based on hexagons :)
And how good it is, even with some flaws.
Source code is available on Mozilla's Github
LZHAM is a lossless data compression codec with a compression ratio comparable to LZMA but with 1.5~8x faster decompression.
And LZMA is one of the best ! ❤️
Link to the repository
(shared by https://sebsauvage.net/links/?KsYL2w)
Compress images and returns a 20-30 bytes long string that represents an image. It is used to wait until the image is fully loaded.
About Huffman coding