If they can be hidden, like in Windows, I dont really mind if every app has one. I’ll hide the ones I dont care about.

The app dock isnt visible by default, so thats a partitial solution, but I’d prefer to be able to access it directly without opening a menu or overview first.

Steam provides a quick launch menu, thats a good enough reason for me.

Technical issues are worth resolving, but I want the UI to remain the same as the windows one. Its a pattern that works just fine.

Does seem worse IMO. There is nothing wrong with the windows tray, should just copy that and call it done.

Actually, (and I wasn’t aware of this until you mentioned it, so thank you), it does support serverless connections:

https://github.com/rustdesk/rustdesk/wiki/FAQ#i-only-use-rustdesk-for-a-few-devices-on-my-local-network-with-no-internet-connectivity-can-i-still-use-rustdesk-with-direct-ip-access

So I think between cloud server, self hosted server and direct IP, OP should be covered.

RustDesk (rustdesk.com) is open source, and similar to TeamViewer, and has paid plans, including a paid self hosted option.

Direct link to skinners comment: https://github.com/systemd/systemd/issues/15004#issuecomment-2264687287

Is it actual server hardware? I’ve seen some very weird things with real servers that take ages to reboot (I was assuming it was self checking or something). Are you sure its hung, and not just very slow to shutdown/reboot?

Is there any serial/monitor output before the hang?

Up-to-date runtimes definitely makes sense, that is where docker shines.

Gitlab is obviously a bit overkill, but maybe you could just create some systemd timers and some scripts to auto-pull, build and deploy?

The script would boil down to:

cd src
git pull
docker compose down
docker compose up --build

Your welcome to steal whatever you can from the repo I linked before.

Its definitely not a lightweight solution. Is the pi dedicated to the application? If so, is it even worth involving docker?

Yeah, probably, but thats not very common is it? Normally you’d just let the docker daemon handle the start/stop etc?

Afaik, systemd has nothing to do with docker, except to start the docker daemon.

I think I have done almost exactly what you want to do, I use gitlab CI to build and deploy my application:

https://github.com/cameroncros/discordshim_rs/blob/main/.gitlab-ci.yml

Gitlab is relatively heavy, so I dont know how it will go on a raspi (I run on a Intel nuc). You can run gitlab on a separate machine, and the CI runner on your Pi.

You really got dogpiled hard there, as if being corrected half a dozen times wasn’t enough, they had to downvote you to oblivion as well :(

Old kernels still run on it, it just won’t get new versions.

The reason Linux (and BSD) is perfect for old machines is that the source code is available, so anyone can maintain it if they want to. It doesn’t mean the core Linux team has to maintain it for them though.

Just to lightly temper your expectations, the OCR isnt perfect, and you may need to add your own tags/text, but its still an awesome system.

At least for paperless, one of the selling points is OCR plus text search. Do you can dump in all your receipts as photos, and then 3 years later, search “lawnmower” and find the receipt for it. (I dont know if this applies for this software, but its very nice in paperless)

Is this a fork of paperless?

https://docs.paperless-ngx.com/

I didn’t say you can’t have a GPU, but to me, its wasteful. I keep my jellyfin server off when not in use, and use WoL to start it when its needed.

I have played with local LLMs, and the models I used were unimpressive, but without knowing what the OP has in mind, we cant know how much power it will use. If it just spins up the GPU once a day for 20 minutes, probably okay, you won’t even notice it. But anyone like me who doesn’t already have a GPU in their lab will probably notice it quite clearly on their power bill.

A megacorps server farm is huge, but its also amortised over millions of users, they probably don’t need 1-1 GPU to customers, so the efficiency isnt necessarily bad. (Although at the moment, given megacorps are tripping over themselves to throw compute at LLM training, this may not be true)

Idle is low power, not zero power. And it won’t be idle when its scraping and parsing the sites, so depending on how much scraping its doing, it could be significant non-idle energy usage.

Yeah, absolutely. And running a GPU 24/7 to occasionally search is just a waste of power. I’m not convinced that google and bings AI search makes financial sense either, Google dropped live search (where the results updated as you typed realtime) because it was too expensive, how does LLM search end up cheaper than live search?!

Edit: This is the live search thing: https://searchengineland.com/test-google-updating-search-results-as-you-type-49116 ~~Annoyingly hard to find, and I can’t find the articles on its cancellation, but from memory it was related to expense. ~~

Edit2: Google Instant Search, and the death was blamed on mobile, and wanting to unify the mobile/desktop experience. I do vaguely remember expense being an unofficial/rumored reason, but I can’t back that up.

I personally have zero interest in AI search, if you mean LLM. The fact that it can make stuff up, also means it can miss stuff as well. Neither are acceptable for a search engine.

If you mean some kind of deterministic algorithm for indexing and searching, then maybe.

Also, attempting to crawl sites locally sounds like a great way to get banned from those sites for looking like a bot.