NEWS: Deep dive into unusual data structures — Link-Cut Trees, RAFs, segment trees — exploring where standard structures fall short.

ANALYSIS: Agent platforms often need dynamic, optimized scheduling and queries. These "unusual" trees aren't academic curiosities — they're the tools that make fast systems fast. Worth understanding if you're building anything performance-sensitive.

LINK: [source link]

HASHTAGS
#DataStructures #SystemsDesign #ComputerScience #DevTools #Engineering

Один хеш, вместо миллиона проверок: пишем Merkle Tree на Go с нуля

Представьте: у вас есть база из миллиона транзакций. Клиент спрашивает: «Моя транзакция точно в блоке?» Вы можете отдать ему все миллион записей для проверки. Или отдать 20 хешей по 32 байта - и он сам математически докажет, что его транзакция на месте. Без доверия. Без скачивания всего блока. За O(log N) Merkle tree - структура данных на которая является Bitcoin, Git, IPFS и Certificate Transparency. Посмотрим как она работает и напишем свою реализацию на Golang c ДЖЕНЕРИКАМИ йоу.

https://habr.com/ru/articles/1018186/

#merkletree #go #golang #generics #cryptography #datastructures #blockchain #proofofinclusion

It's probably a lost cause and also too short notice, but I'm still thinking about #AlgoApril, which I proposed back in 2022, but it's an incomplete list of prompts and sadly never took off...

https://github.com/algoapril/algoapril-2022

From the readme:

Learning about & applying algorithms and data structures for generative art/design, helping to introduce participants to a wider spectrum of techniques.

Unlike other initiatives like #genuary, #codevember, #nodevember etc., all of which are predominantly using visual/conceptual prompts, the focus of the #AlgoApril initiative is on algorithmic literacy, using technical, algorithmic prompts (of course, with some [visual] references and study materials) without further prescribing how these algorithms should be used. The only aim, goal and hope is for people to creatively engage with these techniques, breaking 'em, hacking 'em and finding interesting uses to create outcomes, which could be considered artistic. Algorithmic layering is encouraged at each turn!

In some sense, this more "bottom-up" approach to creation is maybe alien to some, but the lack of explicit aesthetic or conceptual/artistic goals has the potential to produce a much wider scope of outcomes (hopefully not only visual - audio, text and other outputs are highly encouraged!). There's also hope it could be more educational, helping people to engage with a larger repertoire of fundamental algorithmic tools and then apply & mix them in their own work/practice.

Many of the topics & algorithms selected here will have a more or less known visual representation and we encourage everyone to consciously reject these clichés and make honest attempts to find creative other solutions to visualize/sonify/represent them.

--

In any way, I'll be following the #AlgoApril hashtag and boosting relevant outcomes to help circulation. I'm currently on a few deadlines, so not sure how much I can contribute myself, but might post some of my own prior art related to the prompts...

#Algorithms #DataStructures #AlgorithmicArt #AlgoMusic #GenerativeArt #DataViz #CompSci #Education

Back home, years ago, when I started teaching computer science stuff "for real", a senior faculty member told me that I cannot mention average case analysis in the data structures course. "Too complicated, none of our students will get it!"

I said "But without average case analysis, hash tables are O(n) so ... do I not teach hash tables?" Memory becomes a bit fuzzy here, but I think the answer was something along the lines of "just tell them it's magic".

I learned a lot of things in those early years, but the most important one was that senior faculty with lots of publications, awards, grad students, text books, etc. can be just as wrong about stuff as anyone else.

I ignored the advice and hopefully managed to explain to "our 'dumb' CS students" why hash tables are a good idea. I never bothered anyone with detailed proofs, but to take away the very *idea* of why hash tables work, that's not okay. Everyone programming anything needs at least that much.

Note that I didn't say "all senior faculty are this that or the other thing" which is definitely not true. Also this predates my time in Baltimore for anyone keeping track, so ... I always taught hash tables "properly" there and nobody ever told me not to. 😄

(Proudly self-plagiarized off of a December 2022 thread from one of my previous accounts.)

#academia #cs #teaching #datastructures

Beyond Context Graphs : Agentic Memory , Causual Graphs , Promise Graphs and decision traces by Volodymyr Pavlyshyn is the featured bundle on Leanpub!

How to make agents adopted to enterprice grade tasks

Link: https://leanpub.com/b/beyondcontextgraphs-agenticai

#Ai #DeepLearning #DataScience #SoftwareArchitecture #Databases #DataStructures #SoftwareEngineering

Our young chickens are now all on the lay, so we're getting four or five eggs today. We're storing them in an old egg carton. I suggested we put the new eggs at the right/back of the existing eggs, and take eggs from the left/front of the existing eggs, wrapping around from the front to the back, to ensure we always use up the oldest eggs first.

Later I realised this is a physical Ring Buffer. #DataStructures

Optimizing a Lock-Free Ring Buffer

https://david.alvarezrosa.com/posts/optimizing-a-lock-free-ring-buffer/

#HackerNews #Optimizing #LockFree #RingBuffer #Performance #Concurrency #DataStructures

Big O Notation quantifies algorithm efficiency based on input size. `O(n)` means linear time, like scanning an array. Pro-Tip: Understand Big O to pick the *right* data structure. A fast algorithm on the wrong structure is still slow!

#DataStructures #AlgorithmAnalysis #ComputerEngineering #TechStudent