Uninen

I've been working on a streaming platform for musicians as a passion project for the last decade partly because the analog radio shows have become a dying breed. Web-based indie radio shows are lots of fun because they're very close to the best radio shows from my own youth when most radio DJ:s still were curators and experts in the music they played. And they have the feeling of community and genuine passion for the hobby.

There's definitely still demand for quality radio shows but I guess it just makes more money if you play whatever the big studios pay you to play. But its the same thing with television as well; many indie channels on YouTube produce MUCH better content than anything you can see on the traditional TV or Netflix.

I hope someone builds an open API around this metadata. I'd love to have alternatives to the big player APIs.

I'd say Ghostty is a pretty big success story as well.

The problem with the Small Web list is that it's English-only! They definitely need a multilingual one as well.

I think you can either have the memory on or off but according to the docs the projects have their own separate memory so it wont leak across the projects or from non-project chats:

"Each project has its own separate memory space and dedicated project summary, so the context within each of your projects is focused, relevant, and separate from other projects or non-project chats."

https://support.claude.com/en/articles/11817273-using-claude...

You're absolutely right!

Vite (which is used by a large portion of modern frontend frameworks) by default builds the production bundle for baseline widely available: https://vite.dev/guide/build.html#browser-compatibility

Off-topic but this is one of my pet peeves of the modern web; some of the images on the page are ridiculously gigantic. We've had awesome publishing software for two decades with the very basic feature of optimizing published images but somehow nowadays it's totally acceptable to publish an article with several 10+ Mb images.

This is wild!

"when assessed by Claude 3.5 Sonnet’s production-grade RM, our unsupervised assistant policy wins 60% of head-to-head comparisons against the policy trained with the human-supervised RM." So now the models can even post-train the new models better than a human can

That's a good question. This might be some kind of a "fireplace delusion" [1] kind of a thing, I've never really even considered it. I've never seen anyone have a fire extinguisher in a flat here in Finland, that might also be one reason. I guess that wouldn't be the worst purchase to make but I'd still much rather try not to put myself in a situation where I'd need one :)

[1]: https://www.samharris.org/blog/the-fireplace-delusion

I used to not care that much about fire risks of cheap electric devices (apart from lithium batteries like here because they seem to be inherently volatile/dangerous) until one day few years back when a faulty wire of a $3 iPhone charger from AliExpress caught fire on my desk.

I probably wouldn't have noticed it until way too late if my cat hadn't happened to sleep next to it on the same table. He had a sudden scare of the fire and jumped quickly off the table. It looked so weird that I went and looked what he was afraid of and saw flames coming of the half-melted charger and the wire. The desk was full of paper and junk, it was seconds away from catching fire in a way that I probably couldn't put down anymore as I live in a flat and don't own a fire estinquisher. (We only have a fire blanket in the kitchen but that wouldn't have helped much.)

I will never save few bucks from charger wires or chargers or power banks like these ever again -- it's just not worth it!

That all said, don't have any experience from Anker devices myself but in my experience you typically get what you pay for when buying cheap.

Haven't seen that documentaty but unfortunately it seems to be geoblocked in Finland :(

The Anthropic models comparison table has been updated now. Interesting new things at least the maximum output tokens upped from 8k to 64k and the knowledge cutoff date from April 2024 to October 2024.

https://docs.anthropic.com/en/docs/about-claude/models/all-m...

One thing about the reply gives away why Claude is still basically clueless about Actual Thinking; it suggested me to move the HTML sanitization to the frontend. It's in the CF function because it would be trivial to bypass it in the frontend making it easy to post literally anything in the db. Even a junior developer would understand this.

I'm somewhat impressed from the very first interaction I had with Claude 3.7 Sonnet. I prompted it to find a problem in my codebase where a CloudFlare pages function would return 500 + nonsensical error and an empty response in prod. Tried to figure this out all Friday. It was super annoying to fix as there's no way to add more logging or have any visibility to the issue as the script died before outputting anything.

Both o1, o3 and Claude 3.5 failed to help me in any way with this, but Claude 3.7 not only found the correct issue with first answer (after thinking 39 seconds) but then continued to write me a working function to work around the issue with the second prompt. (I'm going to let it write some tests later but stopped here for now.)

I assume it doesn't let me to share the discussion as I connected my GitHub repo to the conversation (a new feature in the web chat UI launched today) but I copied it as a gist here: https://gist.github.com/Uninen/46df44f4307d324682dabb7aa6e10...

moment.js has been obsolete for years. I've considered it as a sign of an unmaintained code for at least ~5 years by now.

There has been so many better and lightweight alternatives for years that if you haven't already refactored it away, you just don't care about your users or the bundle size. (And I have personally done that exercise a handful of times myself, I know it can be painful, but it's just one relatively small PR!)

I appreciate how truly awesome the Web APIs are nowadays but I still feel sad that the first submission for this was almost 4 years ago in March 2021.

Talk about slow turning ships!

Our team quite recently switched from GitLab Wiki to (a self-hosted) Outline.

The initial impressions were very good but after using it daily for a couple of months I really don't like the default theme(s) (dark nor bright) as the formatting options are very limited and the end result is less readable documents. You can probably tweak these things but we haven't gotten there yet.

The search is great and you can set explicit "edit mode" in your profile settings so both reading and editong becomes clearer.

I've been using Kagi for a while now and I love it.

My only gripe is the usability on iOS as it keeps logging me off every now and then (it feels like more often than many other sites) which then results in a login screen when you're trying to search something and after the re-login the query you already wrote once is now lost.

I hope they'll put these on Flickr!

Even more clickbaity title would have been "Python consumes 70x more energy than Rust"

The intro from README explains it well:

”MRSK deploys web apps anywhere from bare metal to cloud VMs using Docker with zero downtime. It uses the dynamic reverse-proxy Traefik to hold requests while the new application container is started and the old one is stopped.”

https://github.com/mrsked/mrsk

I’m looking forward testing this out. I already use Traefik and back in the day I used to use Capistrano to deploy Django apps, it feels like closing the circle :)

Here in Finland slot machines (and most types of legalized gambling) are controller by a government-owned nonprofit that distributes the revenue to various charities. I think that's the best way to run these kind of things if they need to exist.

On iOS devices Safari (WebKit) has 100% market share.

I confess, I was one of those people who believed that he’d try hard to make a positive change. The reality seems to be exactly what the most cynical takes were; it’s all about money and petty personal things. It’s a shame. The wasted potential is enormous.

According to https://www.postgresql.org/support/versioning/ the "Final Release" (whatever that means) for v14 is November 2026, and for v15 is November 2024. I wonder if it's a typo or some LTS thing.

I’ve been using Hckr news for years, it’s great! The only thing I really miss is dark mode.

That list is a perfect example why GitHub is so problematic.

Forget Copilot. Even without that you’ve put all of those services in one centralized basket, fully controlled by a for-profit company (with proven track record of unethical behavior).

And not only that, but this is true for the vast majority of FOSS projects!

FWIW this post specifically links to the v1.0 release announcement in the form of "Tauri in 100 seconds" video that is now linked from the homepage; this is essentially their launch blog post.

The linked URL nor the video has not been posted before.

I'm just pointing this out as many projects seem to post their 1.3.2 Beta 1 blog posts and that seems to be ok -- or at least I see them pretty often.

I find it fresh and positive that Tauri developers take security very seriously. Before this 1.0 release they ordered a full security audit for the codebase and published the report ( https://github.com/tauri-apps/tauri/blob/next/audits/Radical... ).

The project encouraged me to better my own workflows too, as even the awesome-tauri repo requires signed commits in the PR template :) ( https://github.com/tauri-apps/awesome-tauri/blob/dev/.github... )