@flathub@forteller for *offline* LLM features, I’m not sure if that warrants flagging? Do we need to flag the types of algorithms used if they're used offline?
3 years ago I paid $6.99/month for Disney+, tax included. They offered an annual subscription for $69.99, so I switched to that.
2 years ago it increased to $79.99 plus tax, or $86.39.
Last year it increased to $109.99 plus tax, or $119.34.
This year it increased to $139.99 plus tax, or $151.89.
In the span of 3 years the cost has MORE THAN DOUBLED. Conveniently, they now offer an $7.99/mo plan—with ads. Disney knows they can make more off of you by selling ads.
I just saw that Mozilla is retiring Mozilla Location Services which provided an open, crowdsourced alternative to proprietary location services from Google and other vendors. 😰
This is used by GNOME and thus Fedora, Ubuntu, Endless OS, etc. to help provide OS-level location services without requiring every app to implement its own. And apparently a patent troll is part of the reason Mozilla stopped investing in it. 🙃
I’m frustrated by the Yuzu/Nintendo settlement; it sets a dangerous precedent:
>Developing or distributing software…that…functions only when cryptographic keys are integrated without authorization, violates the Digital Millennium Copyright Act…because the software is primarily designed for the purpose of circumventing technological measures.
Isn’t that how *every* emulator works? What does this mean for game preservation?
Have I just reached a point where I’m older and out of touch, or has tech naming just gotten completely out of hand lately? Just this week I tried to read a few articles and lost interest because of the intense and new jargon.
>Alphabet’s Gemma is like an open-weight Gemini, competing with Meta’s LLaMa, following their work on FLAN-T5, TensorFlow, BERT, and JAX; it does have higher confabulation rates, but is available on Kaggle and Hugging Face.
Several people have commented on an open source project feature request asking “what fire can they light under this issue” to “get it moving.”
Is it a dick move to flatly tell them it would cost approximately $100,000 USD to implement, considering I estimate it would take months of multiple people’s time with both design and engineering expertise required across at least three different open source projects?
Then there’s the issue of: given $100k appearing out of thin air, would *that* issue actually be the *most important* thing for those folks to work on?? And now you see the complexity of prioritizing massive open source projects… 🙃
Want to digitally de-clutter AND improve privacy? Be okay with deleting things!
Gmail (and other products) train you to just “archive” things when you’re done with them, but… why? If you don’t want or need it anymore, you don’t really need it sitting around as an ever-present data mine for Google or anyone else. Same with photos; Google Photos suggests you “archive” blurry photos and screenshots, but… you can just delete them!
Obviously save what’s important, but default to delete.
Pro tip: #OpenSource projects are developed by Genuine Real Humans! Before telling tens of thousands of people something “absolutely sucks,” you could, idk, talk to those humans!
I’m not saying you have to fix it yourself (though PRs *are* welcome…) or even report an issue! Just, like, talk to the people.
You’ll have a more interesting story to tell, at the very least. But you’ll also avoid sending undue abuse to the people behind the project based on your incendiary blog post/video title.
Hey @TTimo, volunteer for Flathub here. Do you happen to recall the concerns or have a link discussing them that I can share with the Flatpak and Flathub teams? It would be great to learn if/where the tech falls short to see if we could remedy it. :)
Remember to share positive posts, memes, etc. about open source things you use and like. A simple “I like this” or “thanks for working on this” can go a long way!
Happy people carry on using things quietly while negativity gets shared, memed, and shoved in front of folks who work on your favorite projects. One negative comment too often outweighs ten positive ones, because human brains are dumb.
If I want to make a nice map overlay (say, like you’d see when attending an event in an unfamiliar city…), are there good open editors that can like, plug into OpenStreetMap somehow? Ideally something you could load into GNOME Maps, but I’d be fine with a web embed or even just exporting as an image.
My current idea is to screenshot the area on a map and overlay in Inkscape, but there’s gotta be something better?
Instead, if you have to link to something that happened on Twitter for some reason, I recommend:
• Reproduce the content locally! If embedding on a blog, just… quote the text inline. Style it up nicely if you want, but the content is more important than the tracking-laden embed code that could break at any time
• Nitter.net; replace twitter.com with nitter.net in links; it works like an archive version that includes important context
Only include a Twitter link as proof/a primary source imho
Stop linking to Twitter; it is intentionally broken without creating an account and signing into it, even just to see a tweet, now.
• No thread context: Link to a tweet in a thread, and visitors cannot read the other tweets in the same thread, including the immediate parent in the thread. So context is entirely broken
• No replies: If you link to a tweet, visitors cannot see any replies to said tweet
• Aggressive prompts to create an account when trying to find more context/clicking anything
Building useful, usable, delightful products that respect privacy.:eos: Partner success at @EndlessOS Foundation:gnome: @gnome Foundation member:flathub: @flathub contributorPreviously: co-founder and CXO at elementary OS, UX architect at System76.Frequently posting about #OpenSource, specifically in #GNOME and #Flatpak realms. I also enjoy #StarWars, #LEGO, #3DPrinting, and #SmartHome.I have a background in UX architecture, open source, product design, & communication.