Hacker News Reader

Comment by weinzierl | Hacker News Reader

The Internet Archive does great work and please support them if you can[1]. Still, they are a single point of failure[2] and if they should falter one day it will be a big loss. What we really need - in my opinion - is a second mainstay - an archive of the archive so to speak.

[1] https://archive.org/donate/

[2] Yes, I know about archive.today (formerly archive.is) but their extent is no way comparable to archive.org. Also, they don't archive pages from archive.org.

View on HN

Library Genesis serves as an unofficial alternative to Internet Archive' Open Library, with no lending controls and copyright restrictions (as they operate outside of US jurisdiction and make no attempt to adhere to copyright law, similar to SciHub).

To your point, the Internet Archive most definitely needs to operate in a distributed, global fashion to prevent a single country from causing a loss of cultural data because of political winds or profits being put first. Mirror the whole thing on multiple continents.

Isn't SciHub just a rich "client" for LibGen?

Implementation detail I gloss over for the audience.

Would you call Netflix a rich client for s3 though?

I think there is value in the destinction if it's only used as a storage backend, but actually gets the media through different sources.

No, I'd call Netflix a rich client for Netflix's OpenConnect appliance network.

To be clear, I have a deep appreciation for SciHub and LibGen, I just don't have the energy to dive in depth in this forum.

I thought the openconnect network was basically just a cache which tries to serve popular media directly from the providers?

They didn't have enough storage for their full library on the appliances previously. So that'd be like a caching proxy in front of the service

Maybe that changed with how tiny their library became over the years. (Edit:it didn't change. That's what it is.)

SciHub is a web service that downloads papers on demand. In addition to serving the paper up, they also send a backup to LibGen, making LibGen's scientific paper section an archive of SciHub.

Archive.is/Archive.today is currently embroiled in a petty battle against CloudFlare DNS users, which is, imho, contrary to the spirit of an "archive" that is supposed to serve as a backup for things. I would rather support essentially any other service than one that actively tampers with DNS resolution for petty reasons.

https://community.cloudflare.com/t/1-1-1-1-does-not-resolve-...

A "petty battle against CloudFlare DNS users"? No. They require a standardized DNS extension in order for their services to operate properly, and CloudFlare is waging a petty battle against the standard.

A "extension" is by definition something they do not require, since otherwise DNS clients written before that extension would not be able to interoperate with them. That's what makes it a "extension" rather than a unconsionable violation of backwards compatiblity.

And that particular extension exists solely as means for DNS proxies to violate the privacy of their users by leaking client identity data to upstream DNS servers. There are several reasons why Cloudflare is evil and needs to die (especially ReCaptcha and associated attacks on TOR), but archive.is is firmly in the wrong on this particular point.

Is this relevant to archive.org at all? It doesn't seem to be affiliated in any way.

There are plenty of digital preservation initiatives around the world, many (most?) inspired by the Internet Archive, including various national libraries.

The Internet Archive is a precious initiative and by far the most well known in its field, but it is not a "single point of failure".

Lots listed here: https://github.com/pirate/ArchiveBox/wiki/Web-Archiving-Comm...

My approach is to try and archive things locally, and then rehost them via ipfs/my own file servers.

I've also listed many other projects that do related things in the ArchiveBox wiki: https://github.com/pirate/ArchiveBox/wiki/Web-Archiving-Comm...

1. One Time 2. $500 3. Nothing in terms of payment is clickable. I disabled ublock origin so it's something else.

note: it works in Firefox, not Chrome. YMMV.

Agreed all around.

The sooner we can put this data/historical records atop decentralized solutions, the better.

Reply on news.ycombinator.com