[jedieaston]

If you have public GitHub repos, people crawl them all of the time. I've seen at least three top-of-HN articles that say something like: "I scripted through 10 billion LOC on GitHub, here's a bunch of passwords", or something to that effect. Just set your repo to private if you don't want it in the vault, and if it's open source anyway, who cares?

[beokop]

Then why have the option to delete public repos at all? Public tweets? If we had the means, should the web as a whole be append-only?

There’s a big difference between “public” and “stored in a glacier forever”.

[mstade]

> If we had the means, should the web as a whole be append-only?

You probably meant this as a rhetorical question, but I'd argue that yes, (for public available data at least) it probably should be. It'd enable solutions to a lot of problems we have with the current web, not least archival and broken links.

[ocdtrekkie]

Honestly, I would argue "stored in a glacier forever" is drastically more private than anything you've ever set to "public". The number of people who can actually get to and read that data is incredibly minimal.

[beokop]

There’s no reason to believe they won’t have an online copy of the 2020 snapshot too. Isn’t that kind of the point? For future generations to be able to use it?

[ocdtrekkie]

The online copy is... GitHub itself. Which is current and up-to-date, and you can continue to remove your data from. The 2020 snapshot is useful historically... like a time capsule. There's no reason to invest the resources in keeping it online. And if it was online, it would have major regulatory problems, such as GDPR.

So there's a lot of reason to believe the 2020 snapshot won't also be online.

[huac]

ha, I am very curious if GDPR does apply to the offline snapshot. it will literally be on EU soil!

[afiori]

It will likely apply, still I imagine that they can simply pretend it does not.

My intuition is that archiving data for long term historical use is different from datamining a [meta]data to maximize invasion of privacy. Also there is a difference in accessibility, stored inside a glacier very few people are going to actually read it.

I believe that if mass complaints from all over the EU emerged it would be a different story. But this does not look like the activities the GDPR was created for

[ocdtrekkie]

This makes me wonder: How does GDPR apply to books? Essentially what's happening here, is that GitHub is printing off a "paper copy" and putting it in a box somewhere.

You can't exactly GDPR request deletion of your information from a printed book, so I'm curious how GDPR applies to such physical archival mediums.

[ascorbic]

Norway isn't in the EU

[saagarjha]

> Then why have the option to delete public repos at all?

So you can feel better?

[afiori]

There is an important choice in deleting a public repo, even it has been archived elsewhere. At the very least you are no longer claiming that it fits your criteria of public portfolio.

[tcd]

I feel the term "deleting" is misleading.

It's more akin to "hiding" - to me, deleted means unrecoverable, by anyone, at any time.

Even your OS doesn't "delete" files, until the actual sector on the drive is overwritten (depending on media used, of course).

[hadlock]

> If we had the means, should the web as a whole be append-only?

Yes, probably

[krapp]

>If we had the means, should the web as a whole be append-only?

I'm going to go against the HN zeitgeist and say no.

If I have the right to publish something to the web, I should also have the right to edit and delete it if I so choose.

[celeritascelery]

You do have the right to delete it. And others have the right to make a copy before you do.

[krapp]

That's fine, but that's not the same as the web being "append-only."

[eitland]

It is a saying that

"What happens on the Internet stays on the Internet. Forever."