Login

Recent searches

No recent searches

Search options

Only available when logged in.
social.coop is one of the many independent Mastodon servers you can use to participate in the fediverse.
A Fediverse instance for people interested in cooperative and collective projects. If you are interested in joining our community, please apply at https://join.social.coop/registration-form.html.

Administered by:

Server stats:

502
active users

social.coop: AboutStatusProfiles directoryPrivacy policy

Mastodon: AboutGet the appKeyboard shortcutsView source codev4.3.4

#stackoverflow

1 post1 participant0 posts today
Public
Tommi Nieminen

Imutin kaikki #Facebook'in julkaisuni – ainakin jos #Meta'a uskotaan. Pyysin #JSON-muodossa toivossa, että tulisi sutjakammin. Hieman ongelmia aiheutti JSONin koodaus: merkkijonot ovat validia #UTF8:aa mutta JSON ilmeisesti olettaa #UTF16:n, joten vaaditaan mukamuunnos eestaas; apua löytyi #StackOverflow’sta. Aikaleimat sentään olivat standardi-#POSIX’ia.

En tiedä, kuinka täydellinen ”arkisto” on, mutta ainakin jotakin saisi talteen, kun lähtee lätkimään. #some #atkjuttuja

Public
Csepp 🌢

I don't get how someone with lots of technical knowledge would think that an offline #LLM is a better solution to internet censorship than fucking #Kiwix . You know you can just, download all of #StackOverflow onto any device, right? And that it's basically the same data the LLMs are trained on? Except you won't get hallucinations?

Public *
Pomax

Look, I know my #React answer on #Stackoverflow has over 500 upvotes but I'm done with watching this tag.

The number of people who don't know that React is JS and that every problem they have is because they're not using JS but some weird-ass bespoke React-cascading-bullshit-library has reached "this is someone else's problem" levels.

No amount of Stackoverflow answers is going to educate an entire generation of programmers about the fact that React solutions are JS solutions 1st, React last

Public
Miguel Afonso Caetano

"I’ve been experimenting with ways to fix this (because let’s face it, AI isn’t going anywhere). Here’s what’s actually working:

- First, use AI with a learning mindset. When it gives you an answer, interrogate it. Ask it why. Sure, it takes longer, but that’s literally the point.
- Next, find your tribe. Reddit, Discord, Mastodon—wherever the smart people hang out. That’s where you’ll find the real discussions happening. The ones that make you go “huh, I never thought about it that way.”
- Do code reviews differently. Instead of just checking if the code works, start a conversation with your team. What other approaches did they consider? Why did they pick this one? Make understanding the process as important as the end result.
- Build things from scratch sometimes. Yes, AI can generate that authentication system for you. But try building one yourself first. You’ll write worse code, but you’ll understand every line of it. That knowledge compounds."

nmn.gl/blog/ai-and-learning

N’s Blog · New Junior Developers Can’t Actually CodeSomething’s been bugging me about how new devs learn and I need to talk about it. We’re at this weird inflection point in software development. Every junior dev I talk to has Copilot or Claude or GPT running 24/7. They’re shipping code faster than ever. But when I dig deeper into their understanding of what they’re shipping? That’s where things get concerning. Sure, the code works, but ask why it works that way instead of another way? Crickets. Ask about edge cases? Blank stares. The foundational knowledge that used to come from struggling through problems is just… missing. We’re trading deep understanding for quick fixes, and while it feels great in the moment, we’re going to pay for this later.
#AI#GenerativeAI#SoftwareDevelopment
Public
alpharee

Hey, I'm stuck with this PyTorch problem and no one has answered on Stack Overflow yet. Need help creating a new tensor by using values from a 3D tensor as indices for a 1D tensor, efficiently and preferably without loops. Any PyTorch experts around?

stackoverflow.com/q/79389115/2

Boosts appreciated - thanks! 🙏

Stack OverflowHow to map values from a 3D tensor to a 1D tensor in PyTorch?I'm stuck with a Pytorch problem and could use some help: I've got two tensors: A 3D tensor (shape: i, j, j) with integer values from 0 to n A 1D tensor (shape: n) I need to create a new tensor t...
#pytorch#StackOverflow#tensor
Public
Lou Franco

The main reason a lot of people will find ChatGPT more pleasant than StackOverflow is that if you ask a newbie question, it won’t immediately get downvoted, closed, and draw comments making fun of you.

I’m a little surprised that ChatGPT didn’t somehow learn that behavior.

#AI#StackOverflow#coding
Public
dr 🛠️🛰️📡🎧:blobfoxcomputer:

After having to implement my own #python mini #graph system inside another graph system for a particular THING, I've been looking at how to space axis tick values so they a) use "nice" numbers and aren't too dense

There's a few research papers out there, but they have iterated optimizations, etc. I want something simple and fast just to keep my tucks from running into each other or being uselessly far away

This #stackoverflow post made no sense to me (esp due to the awful variable names in the top reply). But after messing with the ideas a bit, I think I get it.

stackoverflow.com/questions/36

Basically, see how big your data range is in orders of magnitude. Then see how "full" the "last order" is. If the last order is very full, take big steps, medium medium, etc. The step sizes are chosen from nice numbers like 1, 2, 5 and 10.

But here's my question: What about angles?

I think Id want a two-tiered system. For angles < 30, the nice numbers are still 1,2,5,10. For larger angles, I think they are 1,5,15,30,45,60,90

I'll have to try it and see

Stack OverflowAlgorithm for "nice" grid line intervals on a graphI need a reasonably smart algorithm to come up with "nice" grid lines for a graph (chart). For example, assume a bar chart with values of 10, 30, 72 and 60. You know: Min value: 10 Max value: 72...
Public
AfterDawn

Ohjelmoijien suosikkisivusto StackOverflow on kuihtumassa - ja arvioiden mukaan jopa kuolemassa

Taustalla on - tietysti - tekoälyn nousu. Ohjelmointi olikin se ensimmäinen osa-alue, johon suuret kielimallit otettiin käyttöön. Ja nyt näyttää siltä, että käyttäjät kysyvät apua ohjelmointiongelmiin mieluummin tekoälyltä kuin StackOverflow'sta.

dawn.fi/uutiset/2025/01/13/sta

AfterDawn · StackOverflow kärsii, kun tekoäly auttaa ohjelmoijiaBy Petteri Pyyny
#stackoverflow#ohjelmointi#tekoäly
ExploreLive feeds
About