And like i said before. ytmous is still resting, resting, resting, and finally wake up again

ytmous nightly (Desktop view)

Another progress. Almost complete.

So feel free to compare the beginning of ytmous-nightly with the current ytmous

There are two types of an open source project.

One that allow you to use the code, And another one that let you see the code.

However, There's also the one that let you to modify the code, But you need to publish your change.

Server owner were advised to fetch updates & reinstall ytsr with our fork to prevent indefinetely dump files

The ytsr package that ytmous use to search videos did not get any update for 9 months, which is now making error about short's video author being unable to fetch.

The issue is already open, however no response from maintainer. So then i create a fork that fixes this issue: https://github.com/Yonle/node-ytsr, and opened a pull request to the upstream.

So yeah, no choice other than forking then using my fork.

Before Twitter was taken over by Musk, Mastodon (or the fediverse) is still a quiet place with random strangers on it.

It was a good place to share you thoughts at the time.

You actually do not need to ask how AI could read how an code works.

Because basically you are asking the AI to read an computer languange, that the AI already understand

It's almost the same as asking your friend that knows physics about how Newton's 1st Law works

The impatient - 3.0.4
Release: 3.0.4
- Backend: Fix backend Range bug
- search.ejs: Do not show channel metadata if there's no data about it.
- Backend: Use my fork of ytsr - watch.ejs: Hide author avatar in related videos section if unavailable
- Backend: NO_CACHE support
- Backend: Proxying M3U8 and HLS chunks support introduced

Signed-off-by: Yonle <yonle@lecturify.net>

Did you know that somebody did commanding crawler bot to crawl ytmous.yonle.repl.co ?

Even i added tag to not crawl or even robots.txt, The bot did not listen anyways.

It's crawled so bad that even the old ytmous mirror is getting crawled

(Don't tell me that this is the reason why Heroku shut down it's free plan?)

Pros:
- The crawler could help me find bugs

Cons:
- My instance resource or even bandwidth could cost a lot than what i expect

There seems to be a memory leak in package side, Could be in ytdl, ytsr, ytpl, or even miniget

I am suspecting miniget for some reasons

miniget is not the one that has problem.

Tested with 200 requests to 15 MB of files already (15 * 200 = 3000 MB traffic).

Now i am suspecting ytdl-core

As i expected. ytdl-core

But as far as i could get, this seems to be node issue itself. As i saw multiple issues that pointing to node v16