@element My team has been working on this for ages, and we finally got it done!.
Now, you can be invited to an encrypted room and read the history, if the room settings allow it.
Other encrypted messengers can't do this.
If you're interested in the security implications, there are details in the spec proposal: https://github.com/matrix-org/matrix-spec-proposals/pull/4268
@quad oh yeah, I love manually copying and pasting an ID placed somewhere in a line of plaintext.
A simple link would only improve this email.
The optimal approach would probably be something like Gemini does with webpages, semantic markdown with links to be interpreted by a client however it wants. But we're talking about email, it will use HTML 1 until the heat death is the universe.
Google claim that their upcoming laptop is the "perfect partner to your Android phone" because you can "access files from your phone as if they live on your laptop" - but did you know that this is already possible, no special laptop model needed?
KDE Connect is completely free and open source software (FOSS) to link your devices together, supported by the @kde charity. It works on almost all phone models, and it does more than just allow you to access your phone's storage. For example, with KDE Connect your phone immediately becomes a remote control for music or videos playing on your laptop or desktop PC!
Don't get drawn in by the slick marketing - you don't need AI to make your life easier with technology, and most of the features large software companies such as Google are promoting have already been thoughtfully implemented by dedicated volunteers in the FOSS community.
@infobeautiful this alt text is garbage, worse than nothing.
@jonny it is transformative i.e. it is not just directly redistributing copyrighted material. So if something is free to read but copyrighted, you can read it and learn from it, AI can read it and learn from it, everyone's happy. If it's not free to read, both you and AI are pirating it if you read it and learn from it.
Access and distribution are different things.
@bremner better proposal, switch to base12 for all of math. Graft an additional finger onto every child's hands if needed. Beats not having the hour nor day dividing neatly even into 3 parts.
W newsroomie Onetu doszło do wypadku. Jeden z redaktorów spadł z krzesła. Koledzy natychmiast dzwonią po pomoc.
Operator: 112, słucham?
Kolega: Stało się coś niewiarygodnego. Mój kolega właśnie wykonał ruch, który zmienił wszystko.
Operator: Proszę pana, spokojnie. Gdzie państwo są i czy ta osoba jest przytomna?
Kolega: To miejsce, które codziennie mijają tysiące ludzi, nieświadomych czyhającego tam niebezpieczeństwa. Co do jego stanu... lekarze będą w szoku. Czy to koniec pewnej ery?
1/2
@VD15 the state of input emulation? Another casualty of the Wayland switch, or was it always bad?
Excited to announce that the @EUCommission has updated it's follow buttons on the website footer!
What's that first platform there? Could that be #Mastodon?
And where did the link to #X go?
All the posts and comments here on Mastodon calling for this, trust me we read them!
@shibao I hereby declare you an ascended centrist. Warm welcome, and happy grilling.
@PixelRobot @foone it's a nice way to cope, but what they actually want is something that "just works" without having to navigate a complex politics network of what instance to choose so that they don't end up 1) rekt by owner's incompetence (queer.af lmfao) 2) in a "nazi" instance with unpleasant company and blocked by everyone else or 3) in an oversensitive instance that considers everyone else "nazi", unable to talk to their friends.
All 3 happened to my friends when they tried to join the fediverse. No wonder they picked Threads, it's cancer, but effortless cancer.
Why are AI people so monumentally *bad* at copyright?
I'm looking for ethical/copyright-safe training data sets. Common Corpus sells itself as that... but then I go read the paper and they include CC BY-SA scientific papers and GPL stuff from GitHub, and then in models trained on that dataset they proudly state:
> Only trained on open data under a permissible [sic?] license [...] By design, all Pleias model are unable to output copyrighted content.
Um, no?? CC BY-SA is not public domain, it's a copyright license. You can't train on CC BY-SA content and then claim your model is any more copyright-safe than whatever Google and Meta are releasing. It just means you're violating the copyright of people releasing content under open licenses only.
@nextdoormat @lukem looks scarier in writing, all the times I've seen people struggle trying to pronounce "sz" like "s-z" instead of making their life easier and checking it's actually just a variant of "sh"...
What’s worse, the reason the slides aren’t shareable is because they’re built with a web-based presentation tool.
Read that again.
Software developer, open-source enthusiast, wannabe software architect. I like learning and comparing different technologies. Also general STEM nerd.