Interesting project. Thanks for the link and I do appreciate it and could see some very good uses for that but it’s not quite what I meant.

Unfortunately as it notes it works as a companion for reverse proxies so it doesn’t solve the big hurdle there which is handling secure and working flow (specifically ingress) of Jellyfin traffic into a network as a turn-key solution. All this does is change the authorization mechanism but my users don’t have an issue with writing down passwords and emails. Still leaves the burden of:

• choosing and setting up the reverse proxy,
• certificates for that,
• paying for a domain so I can properly use certificates for encryption,
• making sure that works,
• chore of updating the reverse proxy, refreshing certs (and it breaking if we forget or the process fails), etc

Which is a hassle and a half for technically proficient users and the point that most other people would give up.

By contrast with Plex how many steps are there?

1. Install (going to skip media library setup as Jellyfin requires that too so it’s assumed)
2. Set up any port settings, open any relevant ports on firewall, enable remote access in setting with a tickbox
3. Set up users
4. Done, it now works and doesn’t need to be touched. It will handle connecting clients directly to the server. Users just need to install Plex client, login to their account and they have access.

By contrast this still requires the hoster set up a reverse proxy (major hassle if done securely with certificates as well as an expense for a domain which works out to probably $5 a year), to then have their users point their jellyfin at a domain-name (possibly a hard to remember one as majesticstuffbox[.]xyz is a lot cheaper than the dot com/org/net equivalents or a shorter domain that’s more to the point), auth and so on. It’s many, many, many more steps and software and configurations and chances for the hosting party to mess something up.

My point was I and many others would rather take the $5 we’d spend a year on a domain name and pay it for this kind of turn-key solution for ourselves and our users even if provided by a third party but that were Jellyfin to integrate this as an option it could provide some revenue for them and get the kinds of people who don’t want to mess with reverse proxies and certificates into their ecosystem and off Plex.

Jellyfin needs to partner with someone people can pay a very low and reasonable and/or one-time fee to enable remote streaming without the fuss of setting up either dangerous port-forwarding or the complexity of reverse proxies (paying for a domain-name, the set-up itself including certificates, keeping it updated for security purposes).

And no a VPN is not a solution, the difficulty for non-technical users in setting up a VPN (if it’s even possible, on smart-tvs it’s almost always not, and I don’t think devices like AppleTV and other streaming boxes often support them) is too high and it’s an unwanted annoyance even for technical users.

I’m not talking about streaming video’s through someone else’s servers or using their bandwidth. I’m talking about the connection phase of clients and servers where Plex acts like an enhanced dynamic DNS service with authentication. They have an agent on the local media server which sends to the remote web service of the third party the IP address, the port configured for use, the account or server name, etc. When a client tries to connect they go to this remote web service with the servername/username info, the web service authenticates them then gives them the current IP address and any other information necessary. It then sends some data to the local Jellyfin server about the connecting client to enable that connection and then the local media Jellyfin server and the client talk directly and stream directly.

Importantly the cost of running this authentication and IP address tracking scheme would be minimal per Jellyfin server. You could charge $5/year for up to 20 unique remote clients and come out ahead with a slight profit which could be put back into Jellyfin development and things like their own hosting costs for code, etc. Even better if they offer lifetime for this at $60-$80 they’d get a decent chunk of cash up-front to use for development (with reasonable use restrictions per account so someone hosting stuff in Hetzner or whatever and serving 300 people with 400 devices will need to pay more because they’re clearly doing this for profit and can afford to throw some more money at Jellyfin).

Until Jellyfin offers something that JUST WORKS like that it’s not going to be a replacement for Plex, whatever other improvements they offer to users it’s still a burden for the server runner to set up remote streaming in a way that isn’t either incredibly dangerous (port forwarding) OR either involves paying money to third parties AND/OR the trouble of running your own reverse proxy and/or involves walking users through complicated set-up process for each device that you have to repeat if you change anything major like your domain name when using a VPN.

Yeah GIMP is more than a decade behind Photoshop and a lot of other software in many respects.

It’s frustrating. Basic things like content-aware fill for small spaces, not even AI generating huge things for large missing pieces but removing some text over a person’s cheek or plaid shirt, something in total 100x100 pixels big or so. Just doesn’t exist. You can clone stuff but it’s not aware of things like the gradient of a shadow that it should match or a highlight or other basic things so you’re left doing extensive work using layers and then cleaning it up to be visually acceptable using multiple tools over 10 minutes of time whereas Photoshop does it with one tool in an instant.

Incredible. This is one of those hard to believe moments.

It’s been 21 years since the release of GIMP 2.0.

It’s been more than 10 years since work on a majorly overhauled GIMP 3.0 was announced and initiated.

And it’s been 7 years since the last major release (2.10).

I can’t wait for the non-destructive text effects. After all these years of dealing with the fact applying drop shadows meant the text couldn’t be edited, at last it’s no longer an issue.

People got flack for saying Proton is the CIA, Proton is NSA, Proton is a joint five-eyes country intelligence operation despite the convenient timing of their formation and lots of other things.

Maybe they’re not, maybe their CEO is just acting this way.

But consider for a moment if they were. IF they were then all of this would make more sense. The CIA/NSA/etc have a vested interest in discrediting and attacking Chinese technology they have no ability to spy or gather data through. The CIA/NSA could also for example see a point to throwing in publicly with Trump as part of a larger agreed upon push with the tech companies towards reactionary politics, towards what many call fascism or fascism-ish.

My mind is not made up. It’s kind of unknowable. I think they’re suspicious enough to be wary of trusting them but there’s no smoking gun, yet there wasn’t a smoking gun that CryptoAG was a CIA cut-out until some unauthorized leaks nearly a half century after they gained control and use of it. We know they have an interest in subverting encryption, in going fishing among “interesting” targets who might seek to use privacy-conscious services and among dissidents outside the west they may wish to vet and recruit.

True privacy advocates should not be throwing in with the agenda of any regime or bloc, especially those who so trample human and privacy rights as that of the US and co. They should be roundly suspicious of all power.

Sure. Sure. They’ve been close or getting closer for 10 years now.

I’ll believe it when it actually releases and not a moment sooner. Otherwise I would be the opposite of shocked if July 2025 rolls around and it’s still not out but still “close”. As I would be if December 2025 rolls around and “there are only a few more issues, very soon!” is the statement. It’s become a joke at this point and likely will remain the butt of jokes and rightfully so for years, perhaps decades to come in the open source and graphics design communities.