I don't think anyone on Mastodon needs to hear the obvious criticism of this "Accessible Hygiene Helper" app (which I'm certain is 100% private and totally not streaming your camera feed to a server somewhere). But my favourite part about the AppleVis post is this line:

> I noticed that for visually impaired individuals, independently confirming personal cleanliness after using the restroom can be a frustrating and undignified challenge.

And in a reply to a comment rightfully asking how on earth they came to this conclusion, the developer says:

> Why did I build this? Because behind closed doors, many blind individuals face the hidden anxiety of not knowing if they are clean, and asking a family member for help feels undignified.

"Many blind people", you say? Where are these people? Can the real blind testers please stand up? (After they're done wiping, of course.) Because it seems like you saw one of the many repeat posts of this question on AskReddit, and instead of reading the comments or spending a moment to use your brain, you told Claude to materialize the answer in the form of a $4.99/month website. And now you're deep in sunk cost land and every time someone posts another criticism along the lines of "No-one asked for this", you tell yourself "Surely the next comment will be a thank you? Surely someone will sign up soon? Then I can afford my own toilet paper!"

Anyway, this is a weird timeline. I'd like to switch back to one of the ones where this app doesn't exist.

Authentication. Application. Application.

Tailscale for Mac is a true Apple experience--that is to say, it works if you use it exactly as intended and any deviation is a massive headache.
Specifically, there's no way to use the GUI to enter an authorization key, period. The process of using a custom Headscale server is also unnecessarily complex, requiring you to go into settings, find "Accounts" in the toolbar, and then activate the "Go Down" button next to "Add an Account." I ended up using the terminal (which you also have to allow in settings), but which worked instantly.

Looks like Microsoft has proven themselves to be as bad as the rest when it comes to cloud storage.
On family office plans, storage has always been 1 TB per person, which allows for up to 6 TB total. Today, they announced they'll be changing it to a shared 2 TB of storage, much like Dropbox. This comes out to roughly 333 GB per person, and it also means people on family plans will need to pay more attention to how much storage they've used, because now they could accidentally hog the entire storage pool.
OneDrive is built into Windows, and it offers to back up your documents, pictures, desktop, and even downloads. It would be laughably easy for family members to use too much storage without even knowing they're doing it, and this is going to create tons of user confusion.
They've also raised the price of Office 365 a few times and while it's still cheaper than Dropbox Family, it's not *much* cheaper anymore.
To Microsoft's credit, they still have a way better policy on shared folders, counting them toward the owner's storage but letting others sync them for free. Nothing about this change will make me go back to Dropbox. But two wrongs don't make a right, and at this point consumer cloud storage just seems universally terrible and I'm probably going to set up my own solution.
I also feel like Microsoft is using language that genuinely insults the intelligence of anyone reading it. This is not a positive change and it doesn't make storage more flexible--unless one person wants to use more than 1 TB, in which case the other five users on the plan will have less than 1 TB remaining to share amongst themselves. That's not a good scenario, and neither are any of the other ones resulting from this change.
Storage upgrades now go to the whole family too, so if someone's individually using 1.9 TB and you decide to purchase additional storage so you can--like--actually use OneDrive to store anything at all, the storage hog will be able to use that too. Microsoft knew exactly what they were doing here, because if you're in a big family and Uncle Todd is (perhaps unintentionally) using all the storage, what are you going to do? Walk him through clearing out his OneDrive, or just pay Microsoft a few more bucks to upgrade?
I write in Markdown and barely use Office now, and when I do, it's for work, where I have a separate office subscription provided by work. I'll keep my family subscription around for the other members who are relying on it, and because I think I can use discounts to get an early renewal and keep the better storage longer. But as soon as that stops being an option, Microsoft and their sugar-coated marketting slop can fuck right off a cliff.
https://support.microsoft.com/en-us/onedrive/changes-to-microsoft-365-storage#will-i-receive-any-additional-storage-during-the-transition-to-shared-storage

Today I learned CURL speaks a ton of protocols other than HTTP. From the output of `curl -V`:
Protocols: dict file ftp ftps gopher gophers http https imap imaps ipfs ipns ldap ldaps mqtt pop3 pop3s rtmp rtsp scp sftp smb smbs smtp smtps telnet tftp ws wss

I discovered this while I was trying to run a telnet command, finding that Windows didn't have telnet, and waiting a billion years for the feature to install. curl telnet://somewhere.local:23 worked and didn't require a feature install.

Discovered last week that Bobiverse book 6 is out. So far I'm enjoying it.
I do not enjoy the fact that it's exclusively released as an audiobook. I quite like the narrator, but if you can't or don't want to read it that way, you'll be waiting until January 2027 for the ebook edition, which (as I understand it) will also be a Kindle exclusive, forever.
Amazon is a cancer upon the world of books.

Is it just me, or is the "Show recent tabs when pressing alt+tab" setting in Windows literally a setting for keyboard users who can't be bothered learning the difference between alt+tab and ctrl+tab?
You can even ctrl+numbers to switch to specific tabs within the current window, or tell Edge to open new tabs in a new window instead. We didn't need a setting for this.

Yeah, as an Espeak update goes, the Stream firmware is underwhelming to say the least. Unless I'm missing something, there's no way to use Espeak as the menu voice, which means you still get an insanely laggy interface. The more responsive synthesizer is just sitting there, barely out of reach. Also, every single change you make to voice settings seems to require a full reboot. Want to change which voice you're using to read content? Reboot. Want to change which Espeak variant you're using as your non-menu voice? Reboot. Want to add a second Espeak voice in another language? Second reboot.
Also, I tried to use Espeak to read the user guide so I could get an idea of how responsive it is. I'll give them the benefit of the doubt and call this a bug rather than a feature, but Espeak is insanely unresponsive. If I press play or navigate through the book, there's a four or five second lag before Espeak speaks.
I'm officially done hoping for anything useful from Humanware. Quality assurance is just entirely nonexistent.

I have multiple sets of Airpods, and way too many devices to use with them. It's a very first-world problem to have, but I'm constantly trying to figure out which device is paired to what, and occasionally Windows devices just refuse to connect to a pair of AirPods that's paired to something else.
So I started playing with a line of tech products called "airplane bluetooth adapters". These are essentially small dongles that plug into a 3.5mm headphone jack using an attached cable, and broadcast the audio to a connected Bluetooth device, such as AirPods or Shokz.
I guess they're called airplane bluetooth adapters because apparently the most common scenario where people need a headphone jack is now an airplane's entertainment system? But it can plug into anything. I'm now taking one with me in my tech bag so I can plug it into an ATM if I need to.
Some of them have volume controls, which is good if your AirPods don't. Some of them can also act as Bluetooth receivers if you flip a switch. Some of them support the aptX adaptive or low-latency protocols, though sadly the AirPods don't. Most of them pair to multiple devices simultaneously so you can share audio to both at once, which is useful if you own multiple interchangeable devices or you want to have the modern equivalent of a headphone splitter. Of those, some have separate buttons for pairing and connecting to the two devices, which is pretty useful.
All of them are tiny devices, ranging from the size of a small USB drive to the size of a small SD card reader, with a tiny attached cable.
I also found USB audio devices that pair to Bluetooth devices, though the one I got from Ugreen has terrible power management on Windows and needs to be kept alive to avoid choppy screen reader speech.
The end result is that I can now have a couple of dedicated devices paired to my AirPods, and just plug them into whatever device I'm using. It gives me more flexibility and avoids the pairing song and dance when I first start. I just plug in USB or a 3.5mm cable and go. The 3.5mm devices are battery powered, but most of them have an all-day battery.
I'll link the transmitter I'm using right now with my laptop at the end of this post, because the Ugreen brand has a lot of similarly-named items that aren't as user-friendly. This one has only three controls: Two device pairing buttons and an on/off switch. Hold a device pairing button for five seconds to look for the nearest discoverable audio device and pair to it. This usually takes only a few seconds. With subsequent uses, quickly press one of the device buttons to tell the transmitter to look for that device. I haven't been using these long enough to tell you with certainty that the 3.9-star review is unjustified, but it's working well for me when I do use it, and I really like the dedicated buttons. It's also only $20.
The one I still need to test--but which has good reviews and seems to do everything--is the Twelve South AirFly Pro 2. It has transmit and receive mode, dedicated device buttons, volume control built in, and support for the low-latency aptX protocols. There is also a single-device version which only acts as a transmitter, called the AirFly SE. I'm not certain how proactively these devices handle pairing with previously-connected Airpods after they've been paired with something else, but they seem to be marketted specifically to AirPods users. It's also smaller and lighter than any of the other ones I've seen. However, the AirFly Pro 2 is $60 USD and the AirFly SE is $35. I do see used versions of both going for half-price right now.
This is adjacent to my exploration of Bluetooth transmitters and receivers for universally getting low-latency audio between my mixer and my headphones. That's probably a separate post if anyone wants me to make it, and the landscape is still shifting around. The short version is that all of this might become obsolete if manufacturors start supporting LE audio, but that's happening slowly.
Meanwhile, here's the 20-dollar Ugreen one on US Amazon. (changing the link to amazon.ca also works.) Ugreen is also the brand with the USB-C audio devices.
https://www.amazon.com/dp/B0FG7JFDN8

Also, Apple has done exactly nothing optimise the Braille autocorrection for Braille itself. For instance, try typing "osyk". Apple will autocorrect it to "paul", because it's using qwerty mapping, and o is next to p, s is next to a, y is next to u, and k is next to l. But "osyk" looks nothing like "paul" in Braille, and there is no way anyone would ever type "osyk" when they meant to type "paul". So they basically bolted the feature on with no changes from the way it works for qwerty typers.
Again: Excellent fucking job, Apple.

I just love how there's no autocorrect bubble sound in Braille Screen Input and also apparently no way to cancel the autocorrect, so you can just infinitely type one word and your phone will infinitely correct it to another. Truly amazing job there, Apple. And I bet that's in the 27.0 release too.

I installed iOS 27.2 beta last night, so I've done very minimal testing, but so far there's more good than bad. My phone feels a little faster overall but VoiceOver navigation feels very slightly more laggy. Evermusic Pro is crashing on launch. Paperback seems to lose TTS if I leave it stopped in the background for too long. I would probably hate both of these things if i weren't currently in the process of transitioning my media playback to Android.
Eloquence has fewer pauses overall, but sometimes it pauses between sentences on a large block of text, and I can't find a pattern to it. Also, VoiceOver seems to have a problem with Unicode apostrophes when the punctuation level is set to something other than "some", including my custom settings which are based on the "some" level. I never told my phone to pronounce apostrophes, but it started doing so after the update installed. Setting the punctuation to "none" just caused VoiceOver to break up the word instead, so "don't" became "don t". Not sure what that's about.
The new image recognition works on the 15 pro max, and I'm enjoying that a lot. Coincidentally, the communications coordinator at work made a "water cooler chat" on Teams this morning, which quickly turned into a furious exchange of pet photos, which made for a very good playground to test out the feature. I appreciate being able to get a one-off image description without having to close a pop-up afterward. Every other equivalent interface would be much more clunky when quickly scrolling through multiple images.
As a sidenote, it's interesting to see nearly every internet-based recognition app taking musical cues from Seeing AI and looping a little jingle while the image processes. Be My Eyes copied it, Meta had a musical progress indicator for a while, and Aira's ringback tone is suspiciously similar. Android and iOS use loops that seem understated in comparison, but it's still a deliberate choice. Apple could have used the progress click from Safari, but someone clearly tested the competing third-party apps and thought, "We're making an auditory animation because that's what people expect to hear."
All that aside, I'm not going back to my backup, yet.

Struggled to renew my Audible membership for the last 20 minutes. Eventually realized there was a confirmation prompt in an aria-hidden dialog.
I don't know why web developers have a fixation with using this attribute without bothering to test whether it's even necessary, but this is the second one in a month. The last one was American Airlines.

Is anyone here using a youTUbe downloader that works locally and isn't terminal-based? I've looked at several of the existing open-source projects and a lot of them are partially or completely inaccessible. Some of them even use WX, but the author seems to have forgotten what button labels are. I'm looking for something anyone could use, and yt-dlp is not really that.

According to my inbox, you can now buy the new BT Braille 20 for the very reasonable cost of five Orbit readers.

What are people using on mobile for easy notetaking?
On iOS, I use Drafts because it opens to a text field and doesn't make you think about where you want to put the note before you write it. Its interface feels like it would be a bit overwhelming for the average person though.
I also still don't have anything that syncs between multiple operating systems. Long ago I used an app called Droptext which presented a list of text files in my Dropbox, and that was about as close to sync as I've ever gotten.

I have access to both the second and third-gen Victor Streams. I thought I might like using one of them as a media player, but I have enough gripes with the shared interface that I lost most of my interest in doing that. However, the second-gen stream boots from a micro SD card hidden under the battery, and is clearly Linux-based, which gives it some hacking potential. I don't think the same can be said of the third-gen, although I wouldn't be surprised if there's some hidden developer mode that gives full root access or something.
Overall, I really find the third-gen to be an underwhelming upgrade, and the entire interface feels more laggy because they replaced all the recorded voice prompts with Acapella TTS, which continues to be terrible for efficiency of any kind. It has wi-fi, but no way to wirelessly send files to it. It has Bluetooth, but no Braille display support, presumably because Humanware sells displays that have some of the same reading functionality built in. It has USB C, but seemingly no support for any accessories at all, not even storage devices, which the Stream 2 did support. They added a standby mode because somehow despite being a decade newer, the Stream 3 takes an entire minute to boot compared to the ten or so seconds of the second-gen model.
Anyway, if anyone knows of something fun that us puny mortals can do with this overpriced thing, I'd be curious. For now, the only reason I'm using it over the 2 is Bluetooth, and I can solve that with a $30 dongle.

On a Windows computer, try to give your external drive a custom label like "External".
Then, maybe you realize you typed "Extenral" instead.
If you go rename it again and you just swap those letters instead of re-typing it, the rename will fail, because Windows puts the (E:) into the rename dialog but doesn't have any logic for removing it from user input.
It's kind of like the way Notepad saves your file as "HelloWorld.py.txt", then doesn't show you the file extension by default.
Sometimes I think I might have been born on a different planet from the people who come up with these defaults.

Spent about an hour trying to figure out why I couldn't log into Windows over SSH using an authorized key. I got to the point of manually running the server and watching the debug output, and still got nothing useful. Supposedly Windows was looking at my authorized_keys file, concluding that my key wasn't in it, and failing.
The problem? I used a Powershell command to grab the public key off my webserver, and that command re-wrote the file as UTF-16, which OpenSSH doesn't like.
Conclusion: When the problem isn't DNS, the problem is Powershell.

There's a cafe just down the street that serves all your favourite dishes and every hot and cold drink you could imagine. Every day, you walk in and they give you your usual order. By the end of the first month you know all of their names and they all know yours. You're there so often that they give you an account you can deposit into, so you don't have to keep paying for your second and third drinks. The table in the corner by the wall outlet is basically yours.
And then you spend a weekend in the next city over, and while you're there, you get a haircut.
When Tony doesn't recognize you Monday morning, you assume he's joking and laugh it off. "Yeah, those sandwiches are making me put on some pounds. I had to lose a pound of hair to keep up. Can I have my usual? I should have $30 in my account here."
But Tony still doesn't recognize you. Even when you give your name. Even when you recite your regular order. "Look, dude. If this is some attempt to keep my money--"
"No, no, of course not!" says Tony. "We just want to make sure someone else isn't trying to order a drink under our favourite customer's name."
"So you think someone with my exact face decided to put on my clothes, come into your cafe, and pretend to be me so he could get a five-dollar latte?"
"Well, it's theoretically possible, isn't it?"
"Okay. Sure. Thanks for keeping my coffee secure. Now how do I prove I'm me?"
"Well, can you give me your customer ID?"
"No? What's that?"
"It's written on every receipt we give you."
"You mean the receipts you offer me and then throw away when I say 'no, thanks'?"
"Yes, those ones. Do you have your customer ID?"
"Right. You're being very difficult here, Tony. Tell me what I can do."
"We can mail you a new customer ID. Do you still live at 1927 Newcastle Road in Seattle?"
"No. Obviously no. We are in Portland. And how did you even know I lived there?"
"Not important. What's the name of your oldest cousin?"
"Patricia. But how did you--"
"Great, you're in. Do you want your usual again? How was your weekend?"
"It was ... Fine, I suppose. And yes please."
You turn to sit at your usual table, but find that someone else is already there, so instead you collapse into the closest chair you can find and manage to log into work only two minutes late.
At 10:30, you're finally hungry for more than a tiny croissant, so you call out to Tony: "Hey, could I get a sandwich?"
He looks up from his phone and gives you an odd stare. "I think I know who you are, but you're sitting somewhere unusual. Can I just get your customer ID from you?"
"I still don't have it. Just like I didn't have it two hours ago. I don't mean to be rude, but is something wrong with you?"
"No, I'm perfectly fine, thanks for asking. But since you just used a security question two hours ago, I'm going to need you to sort through this box of coffee cups and give me all the medium ones that don't have a lid."
"Okay, this is insane. Can I just pay for my coffee with my credit card?"
"Not while your name is flagged as possible fraud."
Slowly, methodically, you pack your bag and walk out, never to return.
#MicroFiction