@technocounselor Kinda sounds too good to be true, doesn't it? I can't quite tell if this has its own interface or whether the visual interaction takes place through the phone. If I can interact with them accessibly, I very much want to see what I can do with them. I'm hoping the open-source OS means the live OS is a bit more open to sideloading/tinkering as well. If these are accessible, the worst we can say is that there's potential for blindness apps to come to these glasses without first needing to beg Meta to allow it.
I'm very intrigued. Probably not $300 intrigued, but this is the first set of smart glasses I've seriously considered. I did a bit of searching to see if I could tie them to a home server for AI work, instead of defaulting to Open AI or something. I never found an answer.
Thanks for the article. This is indeed intriguing. From what I've been able to dig up (hard to know how much of it is promises vs actual capability), you can write your own apps which run in the cloud. But the key point is that they all run in the cloud. That might mean you could run it on your own server, but something truly local, even running on your phone, just isn't possible. That rules out something like on-device OCR, which is sad. Still, it's nice to know there might be an option that doesn't involve sending data to companies like Meta where you have absolutely no control over their use of your data.
Any #GameDev folks have any experience working with both #playdate and #pico8 (and/or #tic80 fantasy console)? Iβm curious if anyone has first-hand experience with trying to work on a project with shared #lua code that works in both environments but has each using its own resources for input handling, graphics, and sound.
Specifically looking for tips, best practices, learned experience, etc.
* CleanShare: Xposed module that removes Direct Share suggestions from Android's Share Sheet π‘οΈ * Middor: mirror apps on your device (e.g. for HUD) π‘οΈ * KashCal: Privacy-focused calendar with optional iCloud sync π‘οΈ * Auto Off Bluetooth: monitors Bluetooth connection status and automatically turns it off π‘οΈ
This is a repository of apps to be used with your F-Droid client. Applications in this repository are official binaries built by the original application developers, taken from their resp. repositories (mostly Github, GitLab, Codeberg).
If you haven't been able to see my posts for a while (I mean, you should consider that a blessing/feature, but I digress) it was because apparently my firewall was doing something strange with the Spamhaus DROP list and blocking connections it shouldn't have.
After receiving several reports that people couldn't connect to my services and I couldn't find them on the official list (didn't search my pf table, sadly) the problem went away after I disabled and re-enabled the list.
How did that fix it? Only the magic smoke inside the machine knows the answer
Really interesting how the Black Panther Party came out in opposition to the murder of Renee Good, and not a single βWhite Lives Matterβ or βAll Lives Matterβ or Gadsen Flag etc. person has shown up. Really interesting.
Wowy-wow! I love your version of I'm a Little Teapot! How would 1 go about making songs like that? Is it Suno, or are you just that good with instruments? *smile* In any case, thank you for sharing it. It's a bit of a bummer that it's only an instrumental, but that doesn't diminish from its overall awesomeness.
Okay, so rant of the day: I think it's great that people develop apps and such to make things work better, IE mastodon, bluesky, etc, but for the love of it all, can't there be a place to actully find them without having to use the someone who knows someone who know someone else so you might, maybe, they'll consider, giving you the link? I mean, I got the mastodon one today with no problem, (thankfully I knew the right people), but now I'm googling a telegram one five ways to Sunday and getting nothing. I get that we as people who are blind are a small population, but damn, that doesn't mean we should have to play virtual hide and seek every time something new comes out.
I actually echo this sentiment. There's probably a really cool damn app that someone made somewhere that I'm not using because I don't know about it's existence.
@Bri Including me, but now you say it, I do remember them saying that, although I think that was before the takeover, so wonder if it even still applies. I would prefer a different site though, too much on one site becomes really hard to deal with
so, in iOS 26, is there no way to see your browsing history in Safari? pre iOS 26 that option was in bookmarks, but no longer. I can still see it in Safari for macOS naturally, but not iOS.
Google: "We found that this gambling ad doesn't violate our policies against gambling ads, but you can personalise your ad experience to block the kinds of ads you don't want."
Me: "So, can I block gambling ads?"
Google: "No, we don't allow gambling ads so a toggle to block gambling ads doesn't make sense. What are you talking about?"
A student in their final semester who has been checking out a laptop from us every week their whole time here just came up to return, not renew, one because they could finally afford their own and they wrote up a testimonial to talk about the positive impact it had for them.
There is a lot, way too much, bullshit at my institution, but these little moments where we get to really see what the library can do for a person is what keeps me in the fight.
Tonalic adapts real studio performances from world-class musicians to your songβs chords, tempo and groove. No AI, no loops, no MIDI. Just living, musical performances powered by Celemonyβs Tonalic Engine and always royalty-free.
Pretty cool. I have no idea what's going on under the hood, but I assume it morphs the notes and tempo of different recorded audio loops to adopt to your project?
I'm honestly not sure, they weren't super clear, but said it doesn't use AI. That's something I find a little hard to believe, but I haven't tried it yet. No idea if it's accessible, I really hope it is but knowing them I doubt it.
π€ Nvidia released Music Flamingo model for music Q&A for genres, instrumentation, tempo, key, chords, lyric transcription, cultural context, etc. I wonder whether they are laying the groundwork to train their own music generation model. Music Flamingo could be an easy way to produce a huge labeled dataset. #LLM #AI #MusicGenmusicflamingo-nv-umd.github.ioβ¦
Schwuler "Tagesthemen"-Chef geht zu Axel Springer Als ich die Γberschrift gelesen habe, dachte ich zuerst es geht um den Herrn, dessen Namen man nicht nennen darf, weil man dann gleich mit Anzeigen ΓΌberzogen wird queer.de/detail.php?article_idβ¦ #ARD #ΓRR #Tagesschau #Tagesthemen #gay #queer #lgbtQ
So this is really bad. I know. But y'all I feel better these days than I have in years. Why? Because LLM's not only make me feel like I'm doing things and making things, but they actually do make things. Slowly yes, with lots of answering questions and guiding it, but it really helps.
So now I have Emacs with Emacspeak on my Android phone, working on an MUD client, got accessible Lemuroid so I can get TalkBack descriptions of games, about every thing I could ever want besides some Braille stuff. And when I get tired of all the Emacs stuff, I'll try and rangle BRLTTY into accepting Contracted Braille input and try and get a Braille Access-like mode going. And then see if I can get BRLTTY to work as well in Termux as it does in Linux. Like no joke, LLM's have opened this up to being a possibility for me.
Now I get it. These are just nice-to-have things. But shoot y'all, it's freaking fun!
Hm. How did you get a mud client working on android? that one I'm super interested in. Not so much emacspeak, but a linux shell that wasn't utter garbage would be nice.
@fireborn TalkBack will speak (usually) only new incoming text, terminal window is explorable by touch, line by line, you can hit volume keys for control or FN, with state announced, and focus tracking sometimes works. Enough for reading info files and nano.
I disagree that these are "nice to have things". They're important, we only don't think so because we, blind people in general, have been in such a serious mess and have been living in a constant battle to get the basics. These things are actually very important, and I say that as one who has never used Emacs and wouldn't know a MUD client if it came up and said hello. Because there has been so little done for blind users, and because every step has been like pulling teeth, and because of the constant low quality we're subjected to, we think if we have the basics, that's enough. It isn't. Tech is supposed to let us do things we couldn't do before. The opinions of others on how valuable those things are is of no importance or interest. Of course, I include my own opinion in that. If you find these things improve your life, then they are absolutely vital and there's no reason whatsoever to think they're unimportant/frivolous. This is one of the advantages of LLMs. I don't see many, but I do see this, the fact that they will get some work done without huge expense when that work simply wouldn't be done otherwise is a wonderful surprise. You feel better because you're rational. The life of a blind person consists, very often, in smashing one's head against walls. Now that smashing is both less painful and actually achieving results. Also, you're not using your head on the wall. Isn't feeling great about that the most reasonable thing anyone can do?
In November 2026, a historic milestone awaits humanity's Voyager 1. The legendary probe will reach the "one light-day" mark, meaning it will be so distant that communication requires a 48-hour round trip!
Audio descriptions are necessary for making videos accessible. They narrate the crucial visual elements that would be necessary for understanding the plot without the ability to see the screen. They describe non-verbal cues like gestures, facial expressions, or eye contact.
Spotify subscribers -- prepare for a price hike, as the company has announced another increase for users in the U.S., Estonia, and Latvia driving the cost notably above Apple Music. appleinsider.com/articles/26/0β¦
Spotify subscribers β prepare for a price hike, as the company has announced another increase for users in the U.S., Estonia, and Latvia driving the cost notably above Apple Music.
The latest episode of 99% Invisible is pretty good, but had a moment near the end that I found super valuable. (30: minute mark)
They discussed the value of being senior and reaching out to younger folk in the field to give them support. Going from "I don't know how to do this" to "My way is how to do this" to finally "There's a million ways to do this, just do it better than I did" is one of the most rewarding things. You never know the kind of impact that you will make, particularly if you were successful.
Identify the folk that will move the folk forward and do what you can to support them.
Come see Roman as he hosts a conversation with The Memory Palaceβs Nate DiMeo on Thursday, January 22, 2026 (7pm PT) at KQED in San Francisco. Get your tickets! Michael Bierut is one of the best known designers in America.
βοΈ Der Deutsche Presserat hat sein schΓ€rfstes Schwert gezogen: FΓΌr ihre falsche βSkandalβ-Story ΓΌber angebliche GeheimvertrΓ€ge zwischen EU-Kommission und NGOs erteilte er der WELT eine RΓΌge. Diese muss sie nun auf der Seite des Artikels verΓΆffentlichen β was bisher nicht geschieht.
π Bei dem Artikel handele es sich um eine βgravierende IrrefΓΌhrung der Leserschaft und einen schweren VerstoΓ gegen die journalistische Sorgfaltspflichtβ, urteilt der Presserat. Das ist ein Γ€uΓerst wichtiges Signal, auch fΓΌr alle anderen Medien!
π£ DIE Welt muss die RΓΌge jetzt dort verΓΆffentlichen, wo auch die gerΓΌgten Artikel stehen. Aber bisher hat sie das nicht getan, auch unsere Emails beantwortet sie nicht. Damit darf sie nicht durchkommen. FΓΌr ihre Verletzung des Pressekodex sollte sie zumindest gerade stehen.
π Helft uns, den Post zu verbreiten und Druck zu machen, damit DIE WELT die RΓΌge auch ΓΆffentlich macht!
Der Deutsche Presserat hat die Springer-Zeitung DIE WELT fΓΌr ihre Berichterstattung ΓΌber Nichtregierungsorganisationen (NGOs) gerΓΌgt. Damit reagierte er auf eine Beschwerde von uns.
Announcing The Linux Command Line Seventh Internet Edition and Third Print Edition
Good News Everyone!
I have just released the Seventh Internet Edition of my book. Along with the Internet Edition release we're just a few days from the general availability of the new print edition which is based on the aforementioned Internet Edition.
I read chapters of an earlier edition to decide whether I could recommend it to aspiring members of the Linux community. I've been enthusiastically recommending it since then. I also notice that this edition is a tagged PDF file, which will make it more accessible to readers who use screen readers (e.g., people who are blind or who have low vision). You can find out whether a PDF file is tagged by running the pdfinfo command.
Installing Gajim on macOS is now only a single click away. This release brings a simplified macOS setup, easier sharing for files ποΈ and support for link previews π.
Thank you for all your contributions β€οΈ Let us know what you think!
Managed to catch a gnome-shell message "cannot reopen /dev/dri/card0: device busy" when plugging a USB-C dongle for an external monitor. So I logged out, and restarted my session *after* plugging in the dongle - surely it won't need to fuck with the device then? - and it started working.
(This is a dongle for an external monitor that I had never managed to make work. Apparently all I needed was to plug it in before starting my session.)
(Why was GDM happy with plugging in the middle...?)
So all I can do is experiment. Iβm refining the settings for Ι, Ι, and Ι. For Ι, Iβll set cf1 to 530, cf2 to 1350, and cf3 at 2400. For Ι, I'll use cf1=500, cf2=1350, and cf3 at 1700, with pf settings to match. For Ι, I'll use cf3 at 1600, with similar pf settings. I'll keep 'r' for the tap/trill, and we'll see where it comes out when I get to it. Probably nowhere. But this is how a speech engine like this comes together. :D (LOL I asked GPT to convert the sounds into their IPA notations on this post)
Fun fact! There isn't a fixed number for IPA representations because the International Phonetic Alphabet (IPA) is a system with about 107 core symbols for sounds, plus numerous diacritics (around 31) and suprasegmental marks (around 17), allowing for countless combinations to represent sounds across all human languages, creating unique "strings" for each word or utterance, and what NV Speech Player needs to do is map these in languages to how they actually sound out by the voice in terms of frequencies. So yes, quite a task. That's why it may not come together for awhile and I'm focusing on responding to Jira tickets at work :D
queer.de/detail.php?article_idβ¦ #ARD #ΓRR #Tagesschau #Tagesthemen #gay #queer #lgbtQ
macOS is now only a single click away. This release brings a simplified macOS setup, easier sharing for files ποΈ and support for link previews π.
modulux
in reply to ben • • •