News for hackers. Stuff that matters.
Re:IBM Releases Fully Homomorphic Encryption Toolkit for mac... by plerpin on Friday June 05, 2020 @09:46PM (#23436071)
I'm getting a kick out of these performance numbers. It's a common refrain on HN condemning $BIG_COMPANY for not using homomorphic encryption for $CLOUD_SERVICE.

With these performance numbers in hand, I'm finding those posts retroactively hilarious.

Re:IBM Releases Fully Homomorphic Encryption Toolkit for mac... by bawolff on Friday June 05, 2020 @10:36PM (#23436323)
Do people really condemn companies for not using FHE? Anyone who knows anything even remotely about FHE knows that its not practical in real applications yet, and won't be for a while.
Re:IBM Releases Fully Homomorphic Encryption Toolkit for mac... by andrewflnr on Saturday June 06, 2020 @12:16AM (#23436796)
I can't think of any such instances. I suspect GP is thinking of something else.
Re:IBM Releases Fully Homomorphic Encryption Toolkit for mac... by cakoose on Friday June 05, 2020 @10:26PM (#23436281)
Is it really that common? I often see HN posts calling for end-to-end encryption, but not homomorphic encryption.

(But even "just" end-to-end encryption is more difficult than many realize.)

Re:IBM Releases Fully Homomorphic Encryption Toolkit for mac... by Andrew_nenakhov on Saturday June 06, 2020 @05:09AM (#23437763)
Most people asking for mandatory e2ee everywhere to protect their privace do so only because they have heard from someone that it is a good idea.

Most often they do not understand the tradeoffs that come with mandatory e2ee (difficult information transfer, easy loss of data, no server-side search, etc), and are not ready to deal with them.

Re:IBM Releases Fully Homomorphic Encryption Toolkit for mac... by remcob on Friday June 05, 2020 @10:06PM (#23436176)
You should compare them to what the performance numbers where 5 and 10 years ago. Homomorphic crypto systems have become many orders of magnitude faster quite rapidly. While they are still no where near non-homomorphic solutions (and likely won't ever be), they have only recently become practical for applications where performance doesn't matter and privacy is valuable (in the sense that people pay a premium for it, which is often not the case).