On the latest production release for Intel Macs, I can no longer connect to two external displays at the same time. I have to plug them in one at a time (sometimes multiple times before they register), and reconnect them after sleep. A very annoying regression.
The hidden "Detect Displays" button in Display settings does nothing.
This bug is one of many papercuts in Apple software.
Here’s another: Hide My Email’s “Forward To” feature is broken. You can’t change which email address messages get forwarded to. I have nearly 500 Hide My Email addresses and want to change the forwarding address, but can’t because the feature is broken.
I migrated family email hosting from Google Workspace to Apple iCloud+ Custom Email Domains a few years ago and it's been smooth sailing. Few people realize that Apple has been a high-volume email host since long before Google Mail. This is the guide I used: https://blah.cloud/miscellaneous/migrating-google-workspaces...
I also used Hide My Email to de-Google. I changed the email on basically every online service I use from my Gmail to a unique Hide My Email address for each service.
At one point, when I had fewer Hide My Email addresses, I was able to change my Forward To address. But now when I try to change it, it simply doesn't work.
I see the same screen as in your screenshot, but selecting a different Forward To address and submitting does nothing. There's no error message; the Forward To address just doesn't change.
I've been trying to update my Forward To address for months on multiple versions of macOS and iOS, and it doesn't work. I think the bug is related to how many Hide My Email addresses I have, since I was able to update it when I had fewer.
Another annoying thing, even if updating the Forward To address did work, is that you can only have one Forward To address for all your Hide My Email addresses. You can't choose separate Forward To addresses for different Hide My Email addresses.
I for one stopped holding my breath after witnessing a pretty major issue introduced in a macOS beta carry into the stable release for the dozenth time.
Doesn't meshtastic operate in the unlicensed bands? Would that mean they're blocked? The article says they're only blocking devices that need FCC certification and since meshtastic is unlicensed it doesn't need FCC certification as far as I know.
The band it's operating in doesn't change whether or not it needs FCC certification. wifi operating on 2.4ghz is an unlicensed band, but it still needs certification for instance.
IIRC, Meshtastic has an is_licensed / “Ham Mode” setting that changes the software's regulatory assumptions and power limits so it can be used.
From that Wayback page, it looks like only one version of macOS is certified at any one time: the latest version. Since macOS Golden Gate was only recently released, perhaps it's a process blip and we'll see Golden Gate listed soon.
Interesting - when I visit the original page from this post and choose "Filter by Standard" I can see Tahoe under UNIX 03, but only AIX is listed under UNIX V7.
This suggests that different parts of the website aren't necessarily in sync with other parts, information wise...
Having long-term backups like this I wonder if bundling a known-working Blu-ray reader or even a combination writer is a good idea. Hypothetically a reader would be hard to come by. With a reader I would either need a adapter to make it work.
why build here? why should we grow our gardens in this factory farm?
if this muse gadget seems so cool, why not make your own?
don’t outsource your curiosity to people like this, they do not care about you, your family, or your livelihood. they want to control you; this is their toy that they made for you to become fascinated by
First, the muse gadget is just an API and SDK. It’s a fun toy for silly projects.
The software world is spoiled in that making something new is just text in a file, we’ve proliferated open source freeware, and with AI we can readily clone entire products in a weekend.
Sometimes though, people like to use commercial products for various reasons (reliability, cost, network effect, etc). If you like to tinker, this seems like a neat way for some without yak shaving an AI assistant along the way. We don’t need to reinvent every wheel for everything.
I personally think “embodied” assistants are a really fun idea, and this is just an API and SDK to build on and it’s super accessible to basically everyone in the world. If you were inclined to use Muse already, why not tinker with this. I’m sure the hardware would be easy enough to point to an API running on your own server later if you wanted.
Meta should have called it Abuse Gadgets instead of Muse Gadgets.
After all, React is a Graphical Abuser Interface Toolkit, so it would be on brand. React is the Virtual DOM, and you're the Actual SUB.
The firmware is Apache 2.0, so yes, you can point it at your own server. But read the token terms before you plant anything: "not a supported product and not a developer platform... may change, stop working, or be withdrawn at any time, with or without notice," and you "may not enable functionality that Meta has disabled." Max 50 devices, none sold or advertised.
That's not a platform, it's a petting zoo with a revocable leash. You're not a hacker, you're a sharecropper. Meta's word for it is "gadgets." The honest word is "tenants."
If you'd finished reading the whole sentence you'd see it went on to say "...so it can reach compatible devices you already own, or anything you build with a local HTTP API."
It's not about using cloud AI. It's using wifi to enable the local AI to do more stuff.
And when my smart TV asked to be connected to the internet it too told me all the fun things I'd get out of it. Of course it didn't tell you about all the spying
Indeed and Ai keychain voice assistants can double as a home Ai smart speaker and also conversing with it via earbuds when out and about.
I was a bit shocked last week to see Zuckerberg announce the Muse Charm keychain as that's what I suggested and blogged about in August -- portable Ai speakers are best suited on keychains.
You just have to fine tune an LLM like Qwen on some synthetic data to do so. There was even someone that had a model that was exactly like Typesafe and published their work a year before Jev (but wasn't marketed as heavily since it was academic).
This doesn't make sense at all. It is a bad statistical analysis of consumer spending. [1] A few things from the thread on why by Levy on why it feels wrong .
First, why does it feel wrong?
1. the top 10% get about 50% of pre-tax income
2. they get 30 to 37% of disposable income after tax&transfers (depending on source)
3. and we know the rich have a higher savings rate, so consume a lesser share of their income (US, China)
May be not top 10% but top 0.1% are good at avoiding tax using various loopholes. Like borrowing against stocks (and other appreciating assets) till death to avoid paying CGT and debt is then repaid by heirs which do not pay CGT AFAIK. So they can spend more than income stats would suggest.
Sort of interesting how Python has got rid of the GIL. Serialization of data securities is still something relevant though and I’m not sure if you should use pickle still?
Pickle is likely still handy, as long as you created the data that you're deserializing. Zope's object database was a big pickle file, as I recall. But yeah, not a good idea if you're cracking open user-supplied files in any way.
Hahaha. Well, I didn't run into that issue, so it mostly "just worked" in the sense of where we had used it. I did mess around a bit with the Plone CMS that runs on top of Zope, and found it horribly slow and resource-hungry for the time. Someone must like it, it's still getting releases in 2026.
I was more traumatized by the Java-based product from Open Market (acquired from Future Tense back in the day), and all the horribleness that that entailed. Starting from XML as a template language, then seemed to get only worse. Nothing like having multiple content areas all rendered as 500 Server Error pages within one page, when it all goes sideways.
Pickle is insecure by design, but definitely convenient if you can trust the input.
If you need something safe then you really need to identify your requirements and your threat model first anyway; I wouldn't blindly reach for a one-size-fits-all solution in any programming language.
(Although I would do a quick check to see JSON or TOML is sufficient and practical.)
reply