I get it - Google sucks for a lot of reasons. Unfortunately, they own the largest video sharing platform, and it’s difficult to avoid. Many people opt to use and share links to 3rd party web interfaces that greatly de-enshittify the experience (Piped/Invidious), and I’m glad for that and that those projects exist.
That said, when sharing a YT video, please just share the canonical YouTube link rather than a link to a random Piped/Invidious instance and let people handle using a 3rd party interface themselves.
Why?
Most People Who Care Probably Already Have a Mechanism in Place
People who want to use 3rd party YT frontends probably already have a mechanism in place to deal with that: integrated Lemmy client support, browser plugin to automatically redirect YT links to their preferred instance, mobile apps that handle YT links, annoying bot, etc.
You’re Forcing Someone to Use Their Non-Preferred Instance
With YouTube having a relatively small number of domains, it’s infinitely easier to detect YouTube links and automatically/transparently re-write them to a Piped/Invidious instance of the user’s choice than the other way around.
It’s much more difficult to do the opposite and account for all the random Inv/Piped instances in the wild, and there’s no way to really identify them by URL alone (aside from a big list which is difficult to keep up-to-date or be all-inclusive).
The Invidious/Piped server you’re linking to may work well for you, but could be on the other side of the planet for someone else. It may also be unreliable, slow, overloaded, or otherwise sub-optimal for sharing links with a wide audience.
Combined, this makes it much more difficult for people to use a local or preferred Invidious/Piped instance while also contributing to a degraded experience.
Boulevard of Broken Dreams Links
Invidious/Piped are in a constant cat and mouse game against Google. In between Google making a change to break Invidious/Piped and those projects implementing and deploying workarounds, we end up with a lot of non-functional links that need to be re-written to another instance or back to YT. That’s not even accounting for Invidious/Piped instances that shut down/go permanently offline. Again, it’s infinitely easier to re-write a YT link to another Inv/Piped instance than detect every possible Inv/Piped link and redirect those.
Conclusion
So, while people’s desire to de-Google is laudable, please be aware that it can also be counterproductive. Sharing the canonical YT link allows the link to avoid dying due to numerous circumstances while also making it much easier for Lemmy clients, browser plugins, etc to use the user’s preferred instance to avoid a degraded experience.
This is generally good practice, as Lemmy is a link aggregator, so it needs the canonical link. If you are using an archive link, then drop it into the post, which will cover all the bases and give people options.
Hard agree, if you are trying to actively convince me to click on your link using an alternative front end for YouTube is the greatest way to prevent me from clicking on it in the first place. I’ve had enough experience with them not working that I just don’t bother.
That’s not to say I’m going to click on an actual YouTube link either, but you have a significantly higher chance of me clicking on a YouTube link then one of those alternate front ends that I can’t guarantee safety or stability of.
Just give me the original source and if I wanted to go elsewhere I can go elsewhere.
Agreed, especially on the “broken links” part
When using the share function to get a link, delete the ? and everything after it to get rid of them tracking that share.
Eg https://youtu.be/rDbgzqR4trE
Instead of https://youtu.be/rDbgzqR4trE?si= …
In general, solid advice. Though there are some extra params such as
start,t, andendwhich can be useful if you want to reference a specific part of a video. Sadly, though, those params are no longer honored when viewing a video embed and only now seem to work on YT proper.URLCheck (available in F-Droid) is a nice tool for inspecting URLs and removing tracking and other nonsense. It includes a lot of nice features such as:
- Url Cleaner: Uses the ClearURLs catalog to remove referral and useless parameters from the url. It also allows for common offline url redirections. Built-in catalog from https://docs.clearurls.xyz/latest/specs/rules/
- Unshortener: Uses https://unshorten.me/ to unshorten urls remotely.
- Queries remover: Displays the decoded individual url queries, which you can remove or check.
It’s handy for checking/editing links before pasting them into posts/comments, and also for checking links before you open them.
Also for a short you can just change the url from /short/(video) to /v/(video) and get the normal UI
Change it to
/watch?v=to have the normal standard YouTube URL. Since comments have no such detection, feel free to use the short URL there.
Especially when random piped links break on a regular basis or the instance shuts down and makes older posts way more annoying to browse.
As always, if you want to use those frontends I recommend LibRedirect: https://libredirect.github.io/ . But yes, post the genuine links for everybody else.
At least the video ID is still valid in those cases. It takes a few seconds to combine it back into a youtube.com URL but it works.
It’s even easier than that - You can actually just replace the Piped hostname with “youtube.com” or another Piped/Invidious hostname and use the rest of the link as-is. Same for Invidious. They all use the same API, at least for the basic watch and embed functions.
The only problem is doing it programmatically since you have to know the link is to a Piped/Inviodious instance, and it’s difficult to do by URL alone (regexing the path and params gives too many false positives).
Even better use Farside
Agree. I use NewPipe and csn easily open YouTube linkscin my instance, but cannot if they are posted on random instance. I would have to manually copy the video code, which I definitely won’t do.
Totally agree.
I don’t even bother clicking on video links on random instances. Most are broken, and the rest are incredibly slow.
Also: By using the plain YouTube link (without “?si=” etc. parameters), you can see crossposts and duplicates.
Didn’t a recent lemmy update start stripping those out automatically? I seem to remember reading smt like this
Didn’t a recent lemmy update start stripping those out automatically?
Maybe. My home server is still on 0.19.3 because reasons: https://lemmy.world/post/23471887
Personally, I think that the new features outweigh the regressions but I’m not going to jump instances all the time.
For a second, I thought Canonical had made a YouTube front-end.
It could also be helpful to explicitly provide the exact title and creator.
That’s the dream lol.
I’m 50/50 on whether I’ll click a random, nondescript YT link someone drops in the comments (at least with posts, a title is required and it’ll often fetch the video metadata). My Lemmy frontend will preview YT links without leaving the app, so it’s now more like 80% I’ll at least give it a glance lol.
Maybe even a note why this is worth clicking on, and not just a lazy teen spamming /all with anything they come across on a school night.
That may just be my preference 🤔
Yep, agreed.
If I can copy the link, then good. If I can’t copy, I’ll pass. YT is not getting my click.
Also are you really “de-googling” if you’re just using a different frontend?
Maybe some can strip out the ads or something but I’d guess that’s always a cat and mouse game.
No ads, no “privacy” popups, no “pleeeeeease log in” popups… I dunno… Makes the platform usable for me…
