Show HN: A personal YouTube frontend based on yt-dlp (opens in new tab)

Agreed. Youtube downloaders are essential for backup purposes and for getting clips to put in your own videos as fair use. But people turning them in to fully user facing ad free frontends are driving the crackdown on the tools so we will end up with no way at all to download videos..

Would be nice if Youtube just let premium users download the actual video files. What I find interesting is how so many of the Chinese social media platforms just let you download videos while western tech companies pretty much universally block it.

I'd say it's less people's fault and more Google's for driving people to want something like it.

so, essentially, what you are saying is that yt-dlp should have never been open-sourced/published and ever posted on HN (so that not even you would have found out about it)?

I don't think they can ever kill it. Something else will rise. There is too much demand for it.

do you feel the same about ad blockers?

My take is: its either there with all of its features and popularity or its not. The argument that it will be taken down if its more popular seems to me fundamenally wrong.

I mean, the root of the problem is that there is essentially only one "Youtube" that isn't a public service. Not sure if you make this better by leaning into it or not.

the desire is already there. they've testing DRM for videos as we speak. this cat and mouse game will never end until google creates some anti-cheat with kernel permissions to attest anti-tamper

Not sure it was ever youtubes desire to shut it down. Why would they, as there are a multitude of reasons why someone would want a video off a platform. It was the RIAA's, since there the ones who sent the takedown.

people are idiots... and trying to become famous by using the lowest hanging fruit, hence killing it in the process.

gatekeeping is not the way.

If they shut down yt-dlp for good, a lot of power users and creators would find the YouTube platform useless for themselves and abandon it en masse for its nearest competitor. A tool like yt-dlp is very much required if you want to engage professionally with that kind of community. Even something as trivial as making a well-produced "video reaction" relies on it.

Yes, YT has good monetization, but it still pays peanuts to the average creator. So the competitive threat is very real - superstars alone wouldn't be enough to make for a really compelling platform.

I view Discovery as a social problem where the content you want is almost always clustered between a relatively small number of creators, regions, etc.

Technically it then becomes less of an indexing everything problem and more of a find a few cornerstone creators, say Khan academy, and occasionally branching out.

So to answer your question I don’t thing the cost/benefit for automating discovery is much better then spending 20 minutes and finding enough cornerstones to fill you for 100+ hours of content. Or similarly finding a social group like an rss feed, say in ios development it would be fatbobman, and sourcing it from there.

Time to source content isn’t the bottleneck worthy of software solutions, yet for monetization reasons discovery is the vice grip of social media and made to be the most important thing.

If you were to have something local build you an algorithm, what signal would you want it to consume and how far from the median would you want it to deviate? Would you want it to use signal from online socials?

I've been using a third party app to watch the videos and the official app to discover content.

Instead of just clicking the video I click share and watch on the unofficial with no ads.

I looked into this as well since I find the YouTube algorithm terrible, but couldn’t find any API for exploration. Which makes sense they want to control what you watch and hence monetize. In a perfect world you could just pick an open source recommendation algorithm from a marketplace and YouTube would just be a wrapper around s3 buckets and some index.

I am almost a month into having a Perplexity subscription and I am not sure I can not have a deep research subscription at this point.

I have found youtube videos this month that I don't know how I would have found otherwise that were just part of the sources for what deep research came back with.

It has really created the opposite problem for me is I have so much good information I don't even know what to do with it right now. I am probably taking a month off to just sort through what I found this past month.

I've been using Unhook[0] for years that it's almost a jumpscare for me to see a recommended video or the Youtube homepage. Your social circles and natural serendipity should be plenty for finding new creators. And in general, avoiding algorithmic feeds will help with ADHD and mindless scrolling.

[0] https://unhook.app/

Check out the Vinegar extension if you use Safari. Same old YouTube but all the videos are replaced with HTML5 <video>s.

I use a Firefox profile to watch specific videos while logged-out just for the focused recommendations.

I've also noticed that I getting more recommendations for small creators with little to no views/subs when I'm browsing from a smaller, developing country.

I readily follow youtube links offered on HN discussions. If anything, I could use more of these.

But otherwise I agree with your concern. Video recommendations on youtube was far from perfect (very repetitive in my experience), but was uncovering useful stuff.

good question. I don't think I have a definitive answer but I'll try:

- pure luck. sometimes I discover a channel/creator/blog by pure accident, I'm an avid rss reader and HN adept so content comes to me naturally, so to speak.

- following a feed (be it a website's rss feed, reddit/YouTube) sometimes made me discover related feeds, simply because someone wrote about a cool project a peer made and links their YouTube/github/blog

What you're describing is a piracy platform. That makes it pretty tricky to get off the ground, with regards to funding and outreach.

What I'd like is essentially a user-controlled caching layer for everything. When you view a webpage or video or something you are fully downloading all of that data, you might as well optimistically write it to a local cache. Then a browser extension could be made that says "save this version" which tells the caching layer to add a tag to all of the assets that were downloaded in this page view. It would create a tag that means all of those assets aren't garbage collected from your local cache and you retain your copy forever.

Super-charging this idea with IPFS is even better. Essentially a collective Internet Archive will be created with every version of every page someone has decided they are interested in, for whatever reason.

This kind of thing would be perfectly feasible with the web as it was designed, which was designed with caching in mind.

But, of course, big corporations like Google will fight hard to stop such a thing happening because they don't want you in control. They want to be in control. They hate peer to peer technologies because they can't control them.

Write a script to call yt-dlp command with url in clipboard on ipfs server

I have one too, it's honestly a very fun area to program around, and I'm not going to be surprised if this thread is full of me-toos.

Mine is specifically meant to help get videos onto plex in exactly the way we want - with particular emphasis on playlists, taking the numbering and putting it in plex format, and transcoding any codecs (detected via ffprobe) i know certain shitty players (smart TVs) will have issues with. Along with putting it in the right spot on the filesystem with the right permissions and user+group set so it serves correctly over samba too (for management from windows / via GUI).

Interesting! How do you stream it to your phone? I imagine its on the local network?

I don't see it. Has it been removed?

Maybe mine are mg/emacs.

copied it from someone else, can't remember who :)

Along with so many shorts. So many. Going from Smarttube back to the official app and it just plain sucks.

They are constantly testing pushing other things into the subscription box.

What I want is it to only show me videos. Now, it also shows shorts, and also now “community posts” which are frequently just self-promotion and useless polls that drive engagement. I’ve started unsubscribing from anyone that uses those features too much. I want videos not “check out my twitch channel” and “want more merch? Check out my merch! Also this is a poll so that you will click it”

One channel I follow got some new “comments from the community” kind of feature, and suddenly posts from anyone on YouTube were showing up in my sub box because they also subscribed to the same creator. All of the posts were image posts that were blatantly rule breaking spam, or comments like “why is this a feature”. None of them were from anyone I intentionally followed. Literally just random internet comments as a huge section in my sub-box. I instantly unsubscribed.

YouTube REALLY wants to shove other content into the “subscription box” because as-is it lets you avoid all the algorithmic clickbait.

On android, you can even force the app to open up to that page (long press the icon and you can place a shortcut to subscriptions).

Monetizing this would put YT-DLP in danger of having legal action taken against it, or at least being shut down.

What is “safe app”? Too generic to be googleable.

I use h264ify plugin and didn't see performance issues for playback. The UI depends on test group you go into, but only the first load is really terrible.

I never have yt issues in FF. Do you have addons that are yt related?

Thanks for the feedback! Some of the new additions are already wip :)

tried to find a solution for your 3rd point, would love to get your feedback

I remember a project from some 20 years ago that acted as a proxy and kept a local copy of every single page you visited. I don't remember any details other that you could access that app and search through the history based on time, urls, page titles and content separately.

I believe this is one area where current AI could really shine.

For instance I have a large collection of links about the stuff I care or one I use as one-line answers to different questions (e.g. a friend is taking part in a hackathon and needs a color palette to display some statistical data - in my collection I exactly that along with 20 page long explanation on why these particular colors were chosen if one wish to know).

I keep them in a long markdown file I can somehow navigate by using tags, hierarchy and short descriptions but it gets clunky. Having youtube links doesn't help.

Would be nice to have a tool that would be able to get transcription, distil it to a short summary and maybe you could even ask direct questions about the contents.

I have one for categorizing subscriptions/channels. I've been running it for 3y maybe more. SQLite too. No history integration. I have not opened sourced it because the code is thrown together and there are some subscriptions that the YouTube API doesn't return. I'm not certain what the commonality between them is, either. Possibly country of origin.

thanks!

Done

will do, thanks for the suggestion

Why disable youtube history? I have it enabled, and my subscription page works fine.

To answer my own question: "To use YouTube.js in the browser, you must proxy requests through your own server"

hey scosman, let me give it a try: I don't really know if it could work, but I have a feeling it won't work mainly due to CORS (regardless if the app is served from http or file protocol) in the browser

There's Piped but that keeps running into "IOS player response is not valid" error. (I don't know if my Invidious instance works either, I shut it down because of errors.)

a lot for sure.

but the more the better, right?

It should work out of the box by just changing the server port of the llm service you’re contacting

LMStudio 1234 ollama 11434

currently into lmstudio, but had it working before with ollama. it's compatible with both, since it uses the standard /chat/completions endpoint.

haha who cares :) the missed opportunity would have been if I kept it for myself instead of releasing it

Indeed, those with ridiculously slow broadband networks won't ever get 4k content again.

DRM was.. and still is dumb... as it collectively punishes paying customers. While ContentID is sometimes abused by brazen scammers, it is a better solution given the majority of content is still served off the YT platform. =3

How many active users do you have?

thanks!

Just recently I stumbled upon these options of yt-dlp, but haven't had the chance to dig deeper (sorry in advance for the formatting):

  SponsorBlock Options:
    Make chapter entries for, or remove various segments (sponsor, introductions, etc.) from downloaded YouTube videos using the SponsorBlock API (https://sponsor.ajay.app)

    --sponsorblock-mark CATS                                         SponsorBlock categories to create chapters for, separated by commas. Available categories are sponsor, intro, outro,
                                                                     selfpromo, preview, filler, interaction, music_offtopic, poi_highlight, chapter, all and default (=all). You can prefix the
                                                                     category with a "-" to exclude it. See [1] for descriptions of the categories. E.g. --sponsorblock-mark all,-preview [1]
                                                                     https://wiki.sponsor.ajay.app/w/Segment_Categories
    --sponsorblock-remove CATS                                       SponsorBlock categories to be removed from the video file, separated by commas. If a category is present in both mark and
                                                                     remove, remove takes precedence. The syntax and available categories are the same as for --sponsorblock-mark except that
                                                                     "default" refers to "all,-filler" and poi_highlight, chapter are not available
    --sponsorblock-chapter-title TEMPLATE                            An output template for the title of the SponsorBlock chapters created by --sponsorblock-mark. The only available fields are
                                                                     start_time, end_time, category, categories, name, category_names. Defaults to "[SponsorBlock]: %(category_names)l"
    --no-sponsorblock                                                Disable both --sponsorblock-mark and --sponsorblock-remove
    --sponsorblock-api URL                                           SponsorBlock API location, defaults to https://sponsor.ajay.app

awesome, thanks!

the project currently supports cookies (never use your own though, of your google profile), just place them in cookies.txt in the root of the project. but it didn't seem to work well on my server, on a residential IP it works well

Didn’t invidious stop working?

Absolutely

does it work on iOS?

Awesome! Have some things cooking for the future, hope to see your feedback if you can

Thanks! It is definitely not the cleanest code I've written but I'm slowly making it cleaner and ready for OSS contributions. Learned a ton along the way too, which makes this all worth it nevertheless.

I'll use the common excuse: I jotted this project down for myself without the thought of publishing it ^^

eventually

if you don't get the sarcasm, I can't help you.

love you too

thanks, and totally agree on the enshittification of the web (and not only).

this is the very reason why I wanted to dig deeper into my-yt and trying to build a custom solution for my needs

This already kinda exists in the TV sticks like Google TV in that all recently watched / continue watching is a mix of different services that open the correct app when launched. Also has the "Live" tab which shows a combined TV guide of all app's that share it (pluto, tubi, prime, etc)

Of course this still locks you into the end app for playback but the concepts are there.