different
30 stories and discussions about different, aggregated from every source we track.
If you've ever felt like a fraud in an interview in English, this is probably what you need to...
<p><strong>TL;DR: Should I give a "time out" to 'newbies' who have been 'newbies' for (subjectively) too long, or am I being too harsh?</strong></p> <p>Let me start with a graph.</p> <p><a href="https://raw.github.com/gist/9e8f8b837c1eb7e68c90/6332a8a13e51ae71adf2cdfad04d8dbb1839f791/skill-vs-time.png" rel="ugc">https://raw.github.com/gist/9e8f8b837c1eb7e68c90/6332a8a13e51ae71adf2cdfad04d8dbb1839f791/skill-vs-time.png</a></p> <p>Now let me explain what the graph <em>means</em>.</p> <p>The red and yellow lines in this graph are two people who have been in the Ruby on Rails community <em>about</em> the same period of time. The person represented by the red line has gained more skill over a shorter period of time than the person represented by the yellow line. Let's say the total time for this graph is 18 months. Both people, for all intents and purposes of this story, are gainfully employed as Ruby on Rails developers. Where they work, what they are paid and what they're building aren't important to this story. What is important is the role that these two people play in the #rubyonrails channel on Freenode.</p> <p>Yellow is in the channel nearly every single day continuously asking questions. Red (not me, by the way) is there, again, nearly every day, answering questions from people like Yellow. Occasionally, these questions take anywhere from mere seconds all the way up to (in more than one case) <em>5 hours</em> to answer. These questions are usually relatively simple things, like defining custom validations, working out nested resources or figuring out syntax errors in ERB code.</p> <p>Now, I don't personally have a problem with people who take up the channel's time by asking question after question after question, month after month after month. That's what the channel's there for. If we didn't allow all those questions then why even have the channel in the first place? My problem is that I think that a lot of the answers to these problems can be solved in a better way than asking anonymous strangers on the internet for help. Like reading <a href="https://railstutorial.org" rel="ugc">free</a> and <a href="http://manning.com/bigg2" rel="ugc">not so free</a> about the topic, for instance. Or finding an anonymous stranger who can consistently volunteer their time to help build the thing that they're having issues with, and become not-so-anonymous, perhaps even a friend.</p> <p>Basically, IRC is a shitty medium for discussing big programming architecture problems, but a great one for teeny-tiny problems.</p> <p>In IRC-land, Red only has so much patience to help Yellow in their quest for learning. If it appears that after a long time that Yellow isn't actually learning anything from what they've been told and they aren't applying any actual critical thinking to the problem space, then Red may get annoyed at the "slow-goingness" of the progress. I've seen this happen. Hell, I've been annoyed myself that some people just don't 'get it' sometimes. But what I realised is that, as the above graph shows, different people learn at different speeds. As an extension of that, different people learn better from different resources. Some people may learn better from just brute-forcing a problem. Others may need hand-holding and a walkthrough of the problem several times before they finally understand.</p> <p>I don't know what the "correct" answer is to this problem. Do you maintain the status quo and let Yellow keep asking questions, day in and day out and hope that, over time, there is actually something to show for it; that they actually reach a higher form of competency? Or do you encourage them (perhaps forcefully so) to look into other means of improving their knowledge in the field? I just don't know.</p> <p>WDYT, Lobsters?</p>
<p>It's about BeOS: a microkernel-based, multimedia OS w/ pervasive multithreading. The demo clip at the end shows what load that architecture could handle without slowdowns.</p>
Why and how I am making a game for two very different platforms.
Stop thinking of Python variables as boxes. This visual deep-dive reveals how Python really stores your data, manages memory, and cleans up after you. Packed with diagrams, code experiments, and aha moments.
Cambridge scientists have shown that girls and boys differ in what they pay attention to, even at birth. Since these differences are present so early, it is possible they emerge due to prenatal factors. The findings…
Have you ever wondered how it’s possible to call C code from other languages like Python? Here are the technical details — and also why this topic is important for security and sustainability.
Fourier transforms are a tool used in a whole bunch of different things. This is a explanation of what a Fourier transform does, and some different ways it can be useful.
The GPT-6.1 Sol release offers 5 models, each with different intelligence, performance, and pricing characteristics. Below is a comparison of the key metrics across the 5 models. For intelligence, the top model of…
Compare the prices of different AI models before starting your larger project. - NikclasTech/nikclas
Performing inference on a Transformer can be very different from training. Partly this is because inference adds a new factor to consider: latency. In this section, we will go all the way from sampling a single new…
Why and how I am making a game for two very different platforms.
Why robotics RL is a different problem than LLM RL, what EXPO-FT gets right and wrong, and what a universal post-training recipe for robotics needs. By Perry Dong, PhD student in Computer Science at Stanford University.
A different approach to forecasting: Jev, a classification LLM, weighs live radar and NOAA models for the next ninety minutes of rain, and every forecast is scored against what fell.
Vulnerability-Lookup - Fast vulnerability lookup correlation from different sources.
If you ever noticed the price was different online than in-store, this affected you
Composite Machine: Automatic Calculus via Dimensional Arithmetic... and a bit more. - tmilovan/composite-machine
Paste a public GitHub repo. Meet its lockfile personality. Bring receipts.
I ran five different coding agents at the same local model, on the same task, with the same frozen test suite — and then I counted why they failed. The answer wasn’t subtle. About 90% of the …
Learn how prompt evaluation noise can make a small prompt tweak look better than it is, even with temperature 0 and provider pinning.
One model writes the work, a different one reviews it, and neither decides whether it passed. Cross-provider AI review with verifiable evidence, for DeepSeek Harness/Cordis hosts. - WDahah/portable...
A request arrives at a web application. Before it reaches the application, it passes through a...
Certain tasks, and their complexity, change how impactful an interruption is. Research also discovers a disconnect between perception and physiological data.
Entity-level git merge driver. Resolves false conflicts git invents when independent agents edit the same file. ~95% reduction vs. line-based merge. - Ataraxy-Labs/weave