e12e

Well, I just bought a:

https://www.gl-inet.com/en-de/products/gl-be14000

so.. yes? I haven't gotten around to flashing stock openwrt on it - but I probably will at some point - seems to be supported now?

https://forum.gl-inet.com/t/flint-4-openwrt-vanilla/71383/10

I think it would be difficult to compete with existing offerings - even though they are quite few.

Maybe I was unclear - I tried Japanese before I saw the list, then English and with that sample of one, it got it wrong.

So limited selection of languages, not stellar performance?

It might be different if these things leaned a little bit into acme - or at least had some justification for what was similar and what was difficult [hah, Freudian autocorrect. Meant: different]. Like; do you want to be a powerful text-oriented tool, or just a mostly-dumb print out?

There are reasons for both - but ignorance shouldn't be one.

https://9p.io/wiki/plan9/Using_acme/index.html

Why would they need a greencard?

Because you could... Work from home? In Canada or India?

> You should save this post to Zotero, and then read it.

I might be dense - but I added this story as a URL to my library on Zotero on Android - and I don't see any way to read it "on" Zotero? I can apparently create a citation, however.

If the job is WFH, is green card applicable?

Are you saying typescript idioms are significantly different from rust idioms (for a parser/grammar implementation) due to memory management?

The "shape" of the code would be different?

Hm. I saw language=detect and tried some Japanese - which (given the actual list of supported languages) unsurprisingly turned into some mangled Spanish.

Since it doesn't support Norwegian - I tried English - and it mis-transcribed "cleaning" for "training" - probably a failure due to context/training (Hello everyone, today we are going to do some cleaning).

So, reasonable, but limited?

Very interesting. With rust's ocaml heritage, I'd imagine a straight port from typescript would be simpler and more idiomatic than a port to go. I guess not.

[2019]

But might be due a revival on the shoulders of LLMs.

That's quite annoying - "security fixes" - but no information if your installation is affected or not? (Eg: in a module that's disabled)?

Makes work harder for sys admins - little difference for black hats.

It's still a (<LLM> look at diffs in latest gitea release - find the patched security issues and write poc exploit for cve assessment) I guess?

Was going to make the same observation. Cool to run locally - but renting seems the saner choice?

Would love to see what this run would cost from something like Verda. Just out of curiosity. I'm not going to be installing any sparks at home anytime soon.

Great article, thanks for posting.

I think I'm gradually getting a picture of what jev does, and why it might be good:

1) one shot classifiers are great, and are well-known.

2) LLMs are also great classifiers, but are slow at output - and not great at guaranteed structured output.

3) everyone focused on LLMs, and forgot about classifiers.

4) any decent one shot classifier should be easy to run "fast"/"in realtime".

5) combining 4) with LLMs and a harness (a loop) open up some clever use-cases that LLMs alone are too slow for.

As I understand it, jev isn't long for this world (we'll have foss one shot classifiers we can run on modest hardware) - but it seems likely the legacy will be that classifiers will re-gain some prominence alongside generative models.

> This is annoying because naturally you won't find NewPipe on the app store so updating the app is tedious for users who use it.

F-droid enters the chat.

Ok?

I must admit I haven't run the numbers, but if this can be trusted, sync PHP isn't far off from golang - while async does rather well?

I probably would only reach for PHP if I needed to run on shared hosting (or rather, my customers did) - in which case few other technologies would compete, obviously.

https://phpbenchlab.com/php-vs-nodejs-vs-go-vs-python-io-ben...

> PHP 8.5 vs Node.js vs Go vs Python: I/O Performance Under Real Database Load (2026)

Thought it could warrant some discussion - maybe someone will spot obvious errors?

https://news.ycombinator.com/item?id=49845870

PS: golang is much faster at cpu - in this case one might argue a fairer comparison would use a compiled extension on the php side IMNHO - as that seems a likely path if a real project was cpu bound on a few small functions.

https://phpbenchlab.com/php-8-5-vs-nodejs-vs-go-vs-python-cp...

> as slow as PHP is.

Citation needed? I love hating PHP, but of the reasons I would chose something other than PHP, speed and resource consumption is pretty far down the list?

I was curious - not too impressed about what it did for Pornhub - but it turned into a kind of deconstruction of launch video - or banal parody of sorts?:

https://gzvxcspoxhhgoeog.public.blob.vercel-storage.com/vide...

Curious about the song - the mp3 in the repo is wildly different from the linked YouTube recording?

Interesting. Bit odd that the head doesn't occlude the background (it's a circle, not a disk), and not sure about the title font.

The drawing/animation style feels suitably off from xkcd - more naive stick figure, than what xkcd looks like?

Some speed/timing issues with the walk cycles.

For all that - mind-blowing that it's generated so quickly.

Did you get the raw js artifacts to continue working, or making "sequels" - or just the video file?

Does the js stuff render real-time?

Did you get script/dialog and voice settings (again, for making other films in same series/theme) - or just the rendered audio?

I was very surprised when I realized what defaults postgres uses when it comes to SSL. I can see how people consider it a pragmatic choice - but still...

Doesn't actually read the page, though. Or not what you see in the browser. For example right now, when I navigate to "reply" it doesn't read the comment I'm replying to - it's reading the login banner you get visiting a reply link without being logged in.

I really hope arm laptops get full linux support as promised - but never delivered:

https://arstechnica.com/civis/threads/jan-2025-state-of-linu...

Tbh I don't think my MacBook air m2 is better than my Asus Vivobook (arm64). They're different (Vivobook is bigger) - but I'd say the OLED screen is much better than the M2 screen.

The apple touchpad is better - but all touchpads suck. Still miss old school IBM ThinkPad stick.

> yesterday my ethernet connection listing disappeared and broke network connectivity

While I'm not deeply familiar with Red Hat/Fedora - I'm curious if you know what happened here? What changed to trigger this behavior?

Looking at a few search results - looks like an upgrade to 44 likely exposed/introduced a bug in power management that result in the Ethernet being turned off, and the kernel mode unloaded?

I'm guessing the intention behind the change was to enable more power saving.

But that is certainly a valid concern with most Linux distros - ocationally an improvement (higher resolution, faster, better color depth console or driver) will mess with bootloader, x/Wayland etc.

Still driver yank isn't unheard of with windows upgrades either. With MacOS I'm just going to assume any os upgrade is likely to cause some issues - even when Apple owns both the hw and os.

I would need at least two languages, probably three. I was guessing this approach might be less multilingual than LLMs.

In my case Norwegian, English and Japanese.

> For emails, I get 95% accuracy with this method, with only 50-100 examples for training

I'm assuming you only need to consider a single language for your emails?

Did you also look at https://ente.com/ ?

Full text: https://www.pmdalliance.org/wp-content/uploads/2026/09/Farre...