I'd consider Youtube to be one of the top easiest sites to get one's data (or someone else's!) out of. Why do you think it's not?

Do you mean downloading using unsupported 3rd party tools? Because those have to be regularly updated to ensure they even work and could some day stop working entirely if Google decided to care about blocking such tools.

In comparison, GitHub offers a simple clone button (among ways) which gives you everything.

Youtube-dl?

I mean I get your argument but op isn’t wrong. YouTube is reallly easy to rip content from.

And the point of that argument is that youtube-dl only works because Google doesn't care, and the need to patch it regularly is only caused by the incidental changes in the service. When Google starts caring for some reason, it'll be a difficult battle.

I haven’t updated mine in ages and it still works...

I've had to update several times in the last couple years due to some YT links not working. Updating always helps.

I have to update every time I use it.

it could be easy to rip, but exact word "rip" is indicating that you are going against service provider's SLA. Meanwhile GitHub is providing tools (git) to get, share content which fits in their SLA.

No, I mean Google.com/takeout

That only helps you backup your own data. Github (or rather Git) facilitates distributed backup.

I just did a complete takeout - it doesn't download videos, only history, comments, playlists and subscriptions (in various formats: json, html, etc).

Then you did something wrong or didn't pick the right export options. I do a download of mine every 3 months, and it works fine.

1. It's not legal (see ToS).

2. Google can block it any moment, or inject ads.

3. You don't get the original data, but a transcoded version. And/or a watermarked version.

In contrast, GitHub is intended to download data from, so above points will never apply.

1 and 3 are incorrect. I'm talking about Google.com/takeout not YouTube-dl.

You said:

> to get one's data (or someone else's!) out

Google/takeout does not apply to someone else's data.

> easily

So they added a button to dump all issues and PRs without messing with the API?

We’re programmers... working with APIs is what most of us do day to day, they’re not holding it hostage behind complicated and metered usage APIs.

GitHub API is decidedly metered. https://developer.github.com/v3/rate_limit/

When I say metered, I meant "pay per query", not rate limits meant to keep the service online.