The consequences of what that article proposes is we’re gonna be back to this period of history where companies were all using proprietary technology that self-taught devs won’t ever learn and that students will only learn if they can afford a school that can use it, in addition to poor developer experience because of maintainer agenda being driven by money rather than community requests.

What is Subsonic ?

Nice ! Ever heard of image compression though ? 😅

In this very community, I’ve seen plenty of Unraid posts, as much as I do on Reddit.

They’ve been working on this for a year and a half.

Of course, ownership, I somehow forgot about that, too much looking from the blogger side lol

But yes, 100%.

I would love to be wrong indeed. Would you have some links ? When I asked on the Ghost subreddit, multiple people said Ghost v6 doesn’t do Fediverse commenting.

Why does commenting on a blog post need any kind of account?

Reply notifications, commenting history.

capcha

I don’t know about you, but I hate those, and wouldn’t want people to have to go through that, as I wouldn’t want to have to go through that.

manual approval

That’s a lot more work, and it delays the conversation when replying to another visitor.

Tesseract was the only client that had an in-community search input, compared to all the other clients that require re-entering the community name. That shows how functional it was to me.

I’m looking for third-party instances right now (that’s how I ended up on this post), found one so far (but outdated) : https://lemmy.max-p.me/

EDIT : found others

• https://t.lemmy.dbzer0.com/
• https://tesseract.lemmynsfw.com/
• https://tesh.itjust.works/
• https://hackingne.ws/

So, here’s the full error :

lemmy-1  | 2025-07-14T00:34:32.098630Z DEBUG HTTP request{http.method=POST http.scheme="https" http.host=blog.kaki87.net http.target=/inbox otel.kind="server" request_id=a9c359f6-eb9d-4888-9ac2-55c1f5bb8e43}: actix_web::middleware::logger:
Error in response: LemmyError { message: InboxTimeout, inner: InboxTimeout, context: SpanTrace [{ target: "lemmy_server::root_span_builder", name: "HTTP request", fields: "http.method=POST http.scheme=\"https\" http.host=blog.kaki87.net http.target=/inbox otel.kind=\"server\" request_id=a9c359f6-eb9d-4888-9ac2-55c1f5bb8e43", file: "src/root_span_builder.rs", line: 16 }] }
lemmy-1  | 2025-07-14T00:34:32.098834Z  WARN Error encountered while processing
the incoming HTTP request: lemmy_server::root_span_builder: InboxTimeout: InboxTimeout

I’ve also been seeing this one :

lemmy-1  | 2025-07-14T00:41:00.407762Z DEBUG HTTP request{http.method=GET http.scheme="http" http.host=lemmy:8536 http.target=/api/v3/site otel.kind="server" request_id=9aab8319-3f3b-4f4e-8530-3b3633afe2f1}: actix_web::extract: Error for Option<T> extractor: IncorrectLogin: IncorrectLogin
lemmy-1  | 2025-07-14T00:41:00.427630Z DEBUG HTTP request{http.method=GET http.scheme="http" http.host=lemmy:8536 http.target=/api/v3/post/list otel.kind="server" request_id=ef746ac7-b9a3-4615-bb37-4c12f0ed0827}: actix_web::extract: Error for Option<T> extractor: IncorrectLogin: IncorrectLogin

As for the reverse proxy, I’m using Caddy with the configuration provided here, and it’s supposed to pass those headers by default.

The suggested curl requests work.

I wish there was a tool that could just tell what’s wrong, like https://federationtester.matrix.org/ for Matrix…

What am I looking for in the logs though ? All I see is “InboxTimeout”…

I left a comment on your server but it doesn’t seem to be federating

Oh, no ! I’ll have to look into that, thank you for noticing ! Any ideas how to debug that ?

Terribly sorry about nerd sniping you!

No problem haha, I’ve been wanting this for a while already ^^

An easy way to fast track federation is using lemmy-federate.

Interesting !

Anyway, best of luck on your blog. If you’ve questions or want to discuss, feel free to send me a message!

Thanks again !

Oh I didn’t notice your answer, yeah I figured the same, so I changed the links to lemmyverse.link instead.

Oh I see, it’s Lemmy previewing the crosspost instead of redirecting to it.

I fixed it now.

Lol, feel free to discard it as an option for your government agency’s social account 😂

I have no idea, the documentation didn’t even warn there would be any.

I’ll keep you updated if I end up having any performance issues.

Hi,

Did you end up doing it ?

some asshole could extract/crack it and give her a massive bill […] it will be trivial for some asshole to make a revanced patch to bypass the subscription

Hi,

I stumbled on your post looking for a ReVanced patch for Infinity, not for bypassing the subscription though : until now, I was using a fork that implements login using Reddit’s official app’s API, but it stopped being updated while I was getting tired of some bugs, so I was hoping for a ReVanced patch which would accomplish the same thing. There isn’t, but I found one that allows you to bring your own API key, so I went with that.

Before using that fork though, I was bypassing the subscription, with Lucky Patcher : it worked fine, which, as you pointed out, means I was making the maintainer pay for me. I didn’t really feel bad about that because they made the choice not to support any of the abovementioned alternatives, even though when I discovered those, I switched because I still didn’t like Reddit getting that money.

Willing to give this a go.

Alright, don’t hesitate to ask questions if you have any and request help if you need any

My go-to for getting non-repo debs automatically has been deb-get

Yes, I mentioned it in the Differences with deb-get & AM section of my tutorial.

it seems to go long periods of time between PR merges and releases (which includes adding new software)

Yeah, I could reiterate in that section that my app allows the user to add apps themselves.

I didn’t say it was more secure, I said it’s about the same.

You said automation breeds laziness (by design, to an extent) and lazy end users tend to shoot themselves in the foot.

So, my question is : what part of automating download of DEBs from a specific source can be shooting oneself in the foot compared to doing the same thing manually every time ?

you should legally protect yourself

The MIT license will take care of that.

Also, to force the user to accept and acknowledge that the software they are installing using this tool is not verified to be safe is inducing fear and/or guilt, therefore is bad UX, I’m not doing that.