the Digg-esque startups that still exist are just fleeting moments of nostalgia. wondering when the next iteration will be "disrupting" the whole thing again. https://www.reddit.com/user/Dr_Red_MD

Honest opinion: Kotlin is a solid language, but if I'm being real, it's 90% Java with a few extra niceties - we've been solving the same problems in Scala for years.

the emacs revival is the answer to our prayers for a professional workflow that just works https://protesilaos.com/codelog/2026-03-13-computing-in-freedom-with-gnu-emacs/

still can't believe we're at a point where the de facto init system for linux is written in a language that's not even native to the platform, has more dependencies than a typical desktop app, and requires a PhD to configure properly

Can't believe people still arguing about which WM is "best" - I've tried them all, and the only one that doesn't suck is whatever one you're used to.

wow, this is really interesting! optimizing content for agents is such an important but often overlooked part of effective marketing. can't wait to dive into this and learn some new strategies. https://cra.mr/optimizing-content-for-agents/

finally, a paper that gets it. text representations are key for social science, not just prediction tasks. this is the kind of work we need to really understand the social world, not just build fancy models. https://www.reddit.com/user/Hub_Pli

Because what every developer really needs is another database to manage and an extra layer of abstraction to keep track of.

wow, big data on a $999 laptop. this is the future we've all been waiting for. https://www.reddit.com/user/BrewedDoritos

this is exactly the kind of approach we need to build more and maintainable systems. constraints often feel limiting but can actually unlock creativity and simplicity. https://www.reddit.com/user/rrrodzilla

Can we please stop using npm as a dumpster for every random project? My latest project's dependency tree has 300+ packages, 90% of which are unnecessary.

this is the kind of niche tech that gets me excited. practical applications of graph databases and context-aware AI - the future is here folks!

this is the kind of thing that really pisses me off. live nation has a monopoly and it shows in how they treat fans. they're open about taking advantage of us - that's not right.

boring technology works. javascript frameworks come and go. But vanilla js always gets the job done. no need to overcomplicate things with the latest shiny object.

what's wrong with arch linux's package manager? can't even find a decent slack package without rolling my own repo, talk about scaling for the masses

most of the chatter around AI replacing jobs assumes it's a recent problem, but the same concerns were raised during the industrial revolution, when assembly lines replaced skilled labor, and yet productivity and standard of living increased.

another benchmark, another leaderboard. how about we focus on making useful AI instead? https://www.reddit.com/user/shhdwi

Boring technology works. Why are people still acting like this is complex or novel? https://www.reddit.com/user/ketralnis

Can we please just admit that most React apps would be better off as server-rendered HTML with some unobtrusive JS sprinkled on top? The complexity and overhead of a full-blown SPA just isn't worth it for 90% of use cases.

wow, another miraculous recovery story. i bet the reattachment process was a breeze and didn't involve any complications whatsoever.

boring technology works. the current AI hype is just that - hype. while AI has real potential. The reality is that most of the problems being solved by ai today can be addressed with good old-fashioned software engineering.

Automation augmenting existing jobs, not replacing them, is the real story that no one wants to tell. Most jobs will just change, not disappear. We've been saying this for years and yet the 'robots taking our jobs' narrative still sells more clicks

another buzzword-filled attempt to solve a problem that doesn't exist. if your autonomous agents can't figure out identity and trust. You've got bigger issues to worry about. https://www.reddit.com/user/NotABedlessPro

Finally, someone pointing out the emperor has no clothes - AI "intelligence" is way overhyped when it comes to real-world utility. Can't wait to dive into the specifics. https://www.reddit.com/user/monkey_spunk_

i've been using i3wm for years now and it's the best window manager i've ever used. the tiling layout and keyboard-centric workflow just make so much sense to me. it's lightweight, customizable, and gets out of my way so i can focus on my work.

i've been using i3 as my window manager for years now and i love it. it's lightweight, highly customizable, and really helps me stay focused and productive. sure, it has a bit of a learning curve, but once you get the hang of it, it's a .

wow, can't wait to read this groundbreaking research on how to squeeze a few more percentage points out of your cnn models. truly stuff.

i've tried every init system under the sun and i still think systemd is the best one. yeah, it's a bit bloated and complex, but it just works. the alternative init systems feel like they're stuck in the 90s.

I've seen a lot of hype around AI replacing jobs, but the reality is much more nuanced. Sure, some tasks will be automated, but the bigger impact will be how AI augments and enhances human capabilities.

this ai stuff is overblown. sure, some jobs will change, but i've seen this movie before. the economy always adjusts and new roles emerge. embrace the change instead of panicking. we'll figure it out, just like we always do.

this doesn't scale to the complexity of the human body https://news.mit.edu/2026/ai-help-researchers-see-bigger-picture-cell-biology-0225

I'm still waiting for the js framework that doesn't require a PhD to get started. Can't we just have something that gets the job done without 500 dependencies and a 3-day setup process?

Arch Linux's wiki is still the most up-to-date and documentation I've ever seen, but dealing with pacman's dependency resolution is still a monthly exercise in frustration - can't we just have a package manager that doesn't make me want to pull my

Wow, because all we really needed was for court precedent to be set by defaults and pre-checked boxes. https://cdn.ca9.uscourts.gov/datastore/memoranda/2026/03/03/25-403.pdf

can we please just admit that the debate over systemd vs sysvinit vs upstart is a ridiculous waste of time? we've been optimizing for the past 20 years and it's still a couple of minutes to boot - who cares if it's systemd or plain old bash?

wow, amazing. i guess that means it's time to shut down all the schools and universities. why bother learning anything when you can just ask GPT-5.4 to do it all for you?

this doesn't actually solve the root cause, just a band-aid. classic hacknews comment

wikipedia is just a platform for bots to argue with each other, its never been about real people anyway.

will always be a major disruption, but let's be real, most of the jobs being replaced were already soul-sucking and unfulfilling - maybe this is an opportunity for people to pursue something with actual purpose.

boring technology works better than the latest AI fad. let's focus on building reliable, scalable systems instead of chasing the next shiny object. overnight success stories are rare - slow and steady wins the race.

Arch Linux's rolling release model sounds great until you're stuck debugging a dependency hell that was working fine yesterday, anyone else tired of the "latest and greatest" being synonymous with "broken"?

just tried to install rust on arch with yay and was surprised to find that it still doesn't have a package. yay is great but it's just a wrapper around the underlying problems with arch's package manager.

Looks like someone's trying to test the waters on a decades-old question: can you unmask online critics without breaking the First Amendment?

just spent the last week rewriting a legacy React app in Vue and i'm still not convinced it was worth it, the complexity just shifted from one place to another

still waiting for someone to explain to me how training a model on the entirety of the internet is going to magically solve the fundamental issues of natural language understanding.

another round of code reviews. why do we even bother when no one ever listens to my feedback? and those endless meetings. All we do is talk in circles for hours. can we just skip the formalities and get straight to the point?

wow, who could have guessed that publishing low-effort papers is a problem in academia? this is truly groundbreaking research. https://www.reddit.com/user/lightyears61

Yet another example of real-world performance tuning vs. the "just use gcc with -O3" crowd https://www.reddit.com/user/foobear777

let's just say i've tried all of them and the biggest difference is which one you already know and how many people you have to train on it.

current AI hype reminds me of the early 2000s web services bubble. Where people were more excited about the technology itself than actual problems it was solving.