Hacker News Reader

Comment by mlinksva | Hacker News Reader

> Scihub Test

Did you just coin that? Well done. Quick web and HN search, nearest I could find is https://news.ycombinator.com/item?id=27805134

View on HN

Yes. :)

It seems to be easier and less prone to flame wars than asking:

* how easy is it for a newcomer to discover content on the privacy overlay?

* what metadata gets leaked when someone searches, downloads, or hosts content?

* how does the privacy overlay perform compared to the regular old web/internet?

* what's the user experience like (e.g., do the devs even care whether non-technical people use their software or not)

* what happens when some mildly interested party attacks the network

* probably other bullets I'm forgetting

It's way easier to just know if Sci-hub is there. Because if it is then a) they've solved the most important problems of a privacy overlay and b) they are probably actively being attacked so nobody has to speculate about their defenses.

solid points.

  Scihub test :-)

Note our scientific progress towards "The Global Brain" https://github.com/Tribler/tribler/issues/3615#issuecomment-...

Specifically designed to improving science

-Tribler team

So: currently failing the Sci-hub test, but already planning work in the direction of passing it. That sounds promising to me.

But what about the issue of filtering out malicious relays? How do you achieve that without a centralized directory server?

Malicious relays have the option to give you bogus content or real content. The block-level hashes of Bittorrent prevent anything bad getting through. So just avoid relays which are not giving you good blocks. At another level, if relaying as honest, but curious (e.g. spying) you need to randomly select several relays and use them all in a long relay path. Pioneered by Tor team, results in excessive bandwidth usage.

Reply on news.ycombinator.com