stock_toaster

I agree. It would be very interesting to see a forjego config-mode/fork/clone/alternative with AGit as the main (or _only_) contribution workflow!

Yes there are ways around it if you have patience, but what the heck!!

https://forgejo.org/docs/latest/user/agit-support/

So like Broadcom?

Once dns-persist-01 becomes available/usable[1], it should make dns validation even easier.

[1] https://letsencrypt.org/2026/02/18/dns-persist-01

And the really insidious thing about this, is the fundamental asymmetry of effort between creation and deletion. Creating a Wikipedia article can take hours, days, or longer, of effort. Tagging an article as AfD takes a few seconds. The actual deletion (once whatever discussion happens) probably takes even less time.

Not really relevant in this case (that the article talks about), but I don't think that it is so cut and dry as "someone spent time on this so we have to keep it". Consider AI spam, or a company (or government!) paying, or forcing!, people to write articles with whatever focus/leaning/slant they desire. It seems like a hard problem!

Maybe people forget how things were before wikipedia existed? Like many things run primarily by volunteers, it is messy and imperfect. It's arguably still pretty great, and I'm glad it is around.

Search for "ITO EN Mugicha Barley Tea".

There are other brands of course, but I enjoy this one and it is easy to get at my local asian foods market (as well as online).

Pepperidge Farms remembers...

I wish there was an EKS-like for Nomad!

Ah. I think I figured it out.

FOX is acquiring Roku in a cash-and-stock transaction valued at $160.00 per ROKU share. FOX will pay $96.00 in cash and 0.9693 shares of FOX Class A common stock for each Roku Class A and Class B share outstanding immediately prior to the effective time of the merger.

So not a straight $160/share cash, but $96/share cash and some FOX shares. Roku trading below 160 means the fox stock is not considered worth (160-96) by the market.

I've always wondered.. Why would a stock (roku in this case) trade below the acquisition price (currently showing around 141)? (160 in this case)

Does the market think this sale won't be approved by the current administration for some reason?

Maybe time to pursue alternative implementations[1].

[1] https://michael.stapelberg.ch/posts/2026-05-24-minimal-memor...

Seems under-reported to me (as far as PDX goes).

For reference: https://www.datacenterjournal.com/data-centers/oregon/portla...

Or gleam if you don’t fancy elixir.

Unfortunately a lot of fiction (sci-fi/fantasy) ebooks are effectively kindle exclusive these days (amazon publisher deals exclusivity), due to the near monopoly amazon has… and since they have locked things down even harder lately, it is much more difficult to export purchases to other readers.

I think so too, otherwise why wouldn't you put that (purported) increased capacity/output into improving your existing products or creating new ones, with the headcount that you already have?

I imagine layoffs are also very much "this quarter and next quarter" with regards to investor visibility.

While LLM Opex is "some future quarter" and very easy to co-mingle with other expenses.

My servers/VMs typically run either FreeBSD or Alpine. A Debian here or there where needed (proxmox, VPS that doesn't support Alpine, corp stuff, etc).

I've also got a couple of test systems running Chimera - going to wait until it hits stable before relying on it too much though. Experimenting a bit with AerynOS too.

all they are doing here is launching a container instance separate Linux VM, vs the typical shared VM instance

This (MicroVMs) is also kind of what apple's container[1] tools do.

[1] https://github.com/apple/container

I wonder if it can really be enforced.

At the very least maybe it would make the advertising (tv, college campuses, etc) of prediction markets illegal in Minnesota?

That alone seems like a good thing.

I think once url matching is added (which is now on their roadmap[1]), I'll try making the switch from my current password manager.

[1] https://proton.me/blog/pass-roadmap-spring-summer-2026

Also, discovery on the OpenAI side seems like it would be especially juicy.

I am dubious that this is anything other than a marketing/bluster/distraction tactic from OpenAI though.

running all performance-sensitive workloads in The Clown now

I really hope "The Clown" isn't just a typod "The Cloud".

If not, tell us more!

I'm inspired/tempted by this to rename my external "higher grunt offload" machine to "clowntown".

Suprises no one.

Well, it is maybe a surprise to at least _one_ person! /s

I think the issue is that Zig lost their biggest project, which was a posterboy project for real uses of Zig.

Bun, Ghostty, and TigerBeetle are 3 popular projects that I have heard about using zig.

I could totally imagine this kind of thing being added due to AI-slop security report overload, like curl was experiencing[1].

[1] https://daniel.haxx.se/blog/2025/07/14/death-by-a-thousand-s...

Strongly agree. And not only that, but time has _already_ shown the continued degradation of the github experience even with users ostensibly sticking around trying to "make it better".

  > My version of your post reads differently:  
  > "GitHub only gets better if people who give a shit stick around to make it better"
  > Walking away would be easy.

Especially with corporate owned software or SaaS ecosystems!

Sounds like you made the right choice with Heroku back in the day. I feel like this is Github's Heroku moment.

I've been using octarine[1] recently (after having used obsidian for quite a while), but I'm definitely going to try this out.

[1] https://octarine.app

Chimera uses mimalloc instead of musl’s mallocng.

https://chimera-linux.org/docs/configuration/musl

Whoah. How have I not run across this yet?!

Thanks for sharing.