I used KODI for a while but switched to “just” minidlna/ReadyMedia for a lighter setup. I know that VLC can resume videos even over DLNA so I’m wondering if it can abstracted away, e.g having a VLC script that checked, maybe based on filename, if the opening file has already been played from another location and if so, resume from there.

I see that in ~/.config/vlc/vlc-qt-interface.conf there is a [RecentsMRL] section with list (unfortunately local only I believe) and times (in seconds I imagine). One could imagine to scp that file to the NAS using e.g inotify (after writes) then merge it (not replace) on that section only, then a shortcut that before starting VLC get the merged file and does the same. That should allow for resuming across devices.

Genuinely curious, what pieces do you suggest we can keep from LLM/GenAI/etc?

scientific accuracy is anathema to AI marketing

Even though I agree in this context “hallucination” is actually the scientific term. It might be poorly chosen but in LLM circles if you use the term hallucination, the vast majority of people, will understand precisely what you mean, namely not an error in programming, or a bad dataset, but rather that the language model worked well, generating sentences that are syntactically correct, that are roughly thematically coherent, and yet are factually incorrect.

So I obviously don’t want to support marketing BS, in AI or elsewhere, but here sadly it matches the scientific naming.

PS: FWIW I believed I made a similar critic few months, or maybe even years, ago. IMHO what’s more important is arguably questioning the value of LLMs themselves, but then it might not be as evident for many people who are benefiting from the current buzz.

FWIW you can fix it locally and eventually use a pull request.

I don’t know find well enough to comment on feasibility but I bet it’d be interesting to try, and maybe realize it’s not trivial.

It’s called having standards.

OK provocation aside yes, you actually have to stand for what you believe in. For some people it means not going to a meat restaurants, for others, like me, it means not accepting a WhatsApp chat or a Google Drive share. You also do that but because it’s either so ingrained or socially accepted you do not even notice anymore. Your standards are definitely not mine but if neither of us do push back, then we as a society go backward IMHO (even knowing my standards are not yours, assuming at least some of us do think and act based on new knowledge rather than random beliefs). So… yes it means my circle of acquaintances is not the most inclusive but I do accept boundaries and if it means someone is toxic according to my perspective, they are out, simple.

PS: you actually have no idea what my social life is. You literally can not judge if it’s “richer” or “poorer” than anyone else.

Easy, I don’t talk to such people. They can have my email or phone number if truly necessary. Yes, same for family or work, just not using Meta products for communication. Surprisingly enough people do understand.

I said Google Glass was fake. I thought everything about it was true except the display. I had never encountered this kind of optics before so when they announced it I claimed it was not possible to ship that then. I was wrong.

I don’t think that’s a realistic comparison. Deflation is just about money whereas technology has actual usage and each iteration is more powerful. Money is exactly the same now and ever, a mean to exchange value.

Not as much as on reddit but still more and more. Just like with Mastodon over X it’s not a perfect replacement (sadly, maybe) but still I feel more resilient. If somehow alternatives are too toxic or unusable I find confident I could invest more time participating here in few different ways. Reddit/X not dying is a double edge sword.

Or you stay in your bubble, not seeing any algorithmic suggestion but only discussing with people you trust. Not saying it’s better but definitely one possible way to use it.

If only the BBC does it then sure, it’s pointless. If the BBC does it and you and I consider it, it might change things a bit. If we do and others do, including large websites, or author guilds starting legal actions in the US, then it does change things radically to the point of rendering OpenAI LLMs basically useless or practically unusable. IMHO this isn’t an action against LLMs in general, not e.g against researchers from public institutions building datasets and publishing research results, but rather against OpenAI the for-profit company that has exclusive right with the for-profit behemoth Microsoft which a champion of entrenchment.

Any example? Curious to see how it compare to e.g NixOS.

shocked Pikachu face

Did customers really forgot the ebook 1984 event or assume they’d “just” get “better”?

Honestly kind of deserved, don’t buy from Amazon! Wondering why? Read Chokepoint capitalism but TLDR their business model is monopoly and monopsony. They’re terrible.

If the goodwill they garner from that makes APPL go up because it matches the privacy expectation they are branding themselves with, they might be making even more money anyway.

Genuinely no idea how Linux gaming could be better. I’ve been playing on desktop and Steam Deck for years, both “flat” games and VR games and it just works. Sure I don’t try literally everything but with ProtonDB I’m confident it will work, or not, and decide accordingly. Obviously not all games work on Linux but definitely more quality games that I have time for. For me it just works, I spend at least 99% of my time gaming on Linux actually gaming, in fact I can’t even remember when is the last time I tinkered. I don’t even have problems with GPU drivers despite tinkering with containers with machine learning. I’m not trying to say nobody has problems or dismiss problems people do have, just sharing my experience.

Not only it works very often but one can even check https://www.protondb.com before buying to make sure it does work. It also works for VR games. I recently tried a brand new game, supposedly “Windows only”, and it worked without any tinkering. I then updated ProtonDB to clarify so that others could play too. It’s simple I didn’t boot on Windows to play for years now. I’m also traveling today and instead of bringing a laptop I bring my SteamDeck to play, to work I’ll also bring a BT keyboard.

TL;DR: it works, even with VR, and ProtonDB can help to identify problems

Makes me curious, can you have a Windows image leading you to work without any interaction? (e.g no activation, mounted data partition, etc)

Surprising to me so I must do some things right :

• dedicated /home partition
• OS on SSD, new OS on fast USB stick
• backup on another physical disk of important data (usually a subset of /home )
• other partition for OS testing
• other working device for instructions and search online (mobile phone is usually enough)
• documented setup for complex tools, e.g /home/Prototypes where you might have container setups, e.g docker-compose.yml

Usually if you have this in place its a matter of hour, at most. Sure in 1h you will not have ALL the apps you need perfectly configured but, for me at least, enough to feel at “home” again. It’s usually about having ~/.bashrc or ~/.tridactilrc in place but if you do have /home on another partition, it’s basically “free”.

… but if you do it starts here https://www.linuxfromscratch.org/lfs/view/stable/prologue/foreword.html

I had a lot of fun few years ago.