Your friendly neighbourhood sh.it.head
Gamer, book and photography nerd, francophile // Gamer, geek des livres et de la photographie, francophile
That is what it seems like based on what I have read :/
I guess the best option in my case then is likely to add them as a non-admin user to my tailnet. The only concern I have is with the potential of one user deactivating the VPN connection unkowingly, which is probably where Funnel comes in as a better option, but I would prefer to avoid serving stuff on the web when possible. (It is specifically a FreshRSS instance for now)
Yes, there is two ways you can go about this. The way that you are thinking of (and the way that I would ideally like to go about this) is as listed on this help article. This is perfect for sharing a home server to some friends, and letting them access a given service without seeing any of your personal devices.
The other option is to have just one tailnet, but having multiple users as detailed here. Notably this can be a security regression (if you don't limit access on a per-user basis with ACLs), but is ideal for sharing access to your entire network with your spouse / older children within the context of self-hosting.
For example, I have a friend who has shared a minecraft server with me and that is an ideal example of sharing one node to a seperate tailnet. I am an admin of the server, and can manage the docker container for it + the backup sidecar and the SMB share, but that is where my access to his network structure ends.
This contrasts the situation with my partner for example, where we share a tailnet (with seperate user logins) to make things like gamestreaming just that much easier to setup. Hypothetically I can use ACLs to limit access to stuff like the Cockpit web-management portal, or block the SSH port, but I don't feel like I need to in my specific case.
Addendum: I also think sharing the device out strips it of its subnet routes + services, which is part of the problem I am running into where I do want it to strip subnet routing (my elderly parents DO NOT need access to my printer), but I ideally want to be able to still use tailscale serve + services + https certificates to be able to share my self-hosted RSS feed reader for them (ad-free, no AI slop, much better for my one parental figure with early-onset dementia).
Addendum 2: I highly recommend exploring tagging + ACLs if you are looking into personal usage / seperation of networks. It is just a much easier approach of seperating devices that are owned and operated by the same person. I would only explore multi-tailnet option when it is different users and you want to share a very limited scope of your network.
Selfhosted @lemmy.world Tailscale serve and sharing devices
- JumpDeleted
Permanently Deleted
I still think a syncthing client of some form is ideal. As someone else mentioned there is the option of using the Syncthing Tray devs experimental android build. To avoid issues with sync-conflicts / maintain high-availability access to the most recent file, I sync the databse to a raspberry pi with the encryption option selected (not that the pi is untrusted per se, but it is a device that doesn't need access to the file, it just serves the most recent changes to other devices since often my laptop / phone / desktop are not all on at the same time).
The more important metric to Canonical however is corporate / paying customer marketshare - I am guessing it hasn't suffered too much otherwise they would have backed down on some of their decisions regarding snaps.
One thing not mentioned, BTRFS supports transparent compression which hypothetically can increase the longevity of SSD media by reducing the amount of writes to the drive.
I say hypothetically because further information on use case (potential write amplification from CoW) could nullify those gains — but frankly, SSD write longevity has improved so much that it is not a huge issue at this point.
If you don't mind clarifying, what do you mean by DoD?
Frankly the best solution i have seen is always a combination of things. At least in the city I live in, people can take bikes on buses and trains, many people walk, and for trips that require trunk space (e.g furniture, DIY supplies etc) there is a Car sharing service that is cheaper than owning a car, or using ride share / taxi.
I don't think waymo is a better option than a combination of what's above, I think it can perhaps compliment it but it should not be the sole last-kilometre solution.
I would like to see waymo-like tech provide better public transit for the disabled. As of now, people in my city with disabilities can book special routes which are serviced by specialized buses/ taxis, and existing lines are all wheelchair accessible as well.
Self driving cars give the opportunity for those people to have even more freedom in booking, since as of now they can't do last minute booking for the custom routes. It wouldn't really create a traffic problem and massively would increase quality of life for those who are sadly disadvantages in society
I'm pretty sure they still are a defence contractor in the US, they also are generally the option for biking computers. There are competitors but Garmin has a chokehold on that sector, with other options just feeling worse.
So while their smart watches are more niche than Apple, Samsung etc, they still have found a solid niche in the smart tracking sector.
That makes a lot more sense! I'm sad now though I thought this was some cool homebrew-ery going on since the GameCube is just a power pc computer (I've seen people run Linux on the hardware before).
If you don't mind me asking, what exactly is the use of a floppy drive on a game cube?
3d printing an IO shield is a great idea! Retro console modding is a great use of a 3d printer.
Frankly I would like to not use Apple CarPlay / Android Auto — however, the built in software needs to actually usable and continuously updated.
I particularly want to see better non-touch input. Rotary dial + buttons à la Mazda, and much better voice input. I live in a multilingual region, and it consequently renders most in-built navigation voice commands useless, as it won't understand language switching. Even Google assitant has issues with this despite supporting multiple input languages, usually resulting in me saying the entire command in the same language as the address. (Or just giving up if the name and street are in two languages).
But with built in systems that only support one language at a time, I just can't say some of the addresses since I don't know how it wants me to mispronounce them in English.
I also have found media playback frustrating in any modern vehicle. This is likely a lot harder to solve, but the inability to switch playlists or change playback settings without my phone connected to Android Auto is frustrating when in vehicles without it.
I know this is very ranty and not that big of a deal, it's just frustrating seeing so little progress in the past decade on this front — and in some aspects like human interface design of vehicles, they have frankly regressed. If I look at the voice input systems on cars from 15-20 years ago there has been huge improvement, but even 10 years ago to now it doesn't feel that different. Maybe a few new commands, but the quality of recognition / utility of the system is lacking.
I find it's just a lot of effort to go through visual customization for very little benefit, I have spent more time creating rootless podman images for certain apps, custom scripts, keybinds etc.
What I mean to say is there's likely many people who customize functionality of their systems one way or another — without ever touching the visual side of things much beside maybe changing the font or turning dark mode on.
The foundation supports a bunch of other open source projects, after all there is a lot more to devices that run the Linux kernel then just the kernel.
Also, I found it a but funny that the foundation created the PDF using Adobe InDesign 19.4 (Windows), according to the metadata in file posted on their website. (original | archive of the PDF)
As of now I am currently using FreshRSS, although before I properly deploy this to other users in my family / friends I might give Tiny Tiny RSS (tt-rss) a shot as well. I don't think the differences will matter for end-users as the majority of mine will likely all be using it through the API via a mobile app (e.g NetNewsWire (ios & mac), FluentReader (desktop), CapyReader (android) etc. etc.)., however the main difference that will dictate which one I stick with is the filtering capabilities and the ease of setup of article-collection with readibility / mercury to remove extrenuous content / ads.
I am also quite interested in miniflux, although it is quite intentionally bare bones. It lacks a plugin api (a potential security improvement), and instead natively supports many of the things people would use plugins for (native youtube-nocookie embedding / invidious embedding, integrations with readlater services like instapaper and wallabag, etc., integrated article fetching and parsing with readibility [and can change user agent / cookies to bypass bot protections]). It also seems to have a bit better security stance (supporting modern web browser features like passkeys, content sanitization, sanitizing url parameters in share links automatically etc.).
Miniflux definitely feels like the best ratio of ootb functionality + security, but the UI of FreshRSS feels more natural if you envisage less techy users to use it (and in my case I see one person using the website over an app).