Going back to your original idea, you could just use public keys and distribute the work of encrypting and decrypting to others, so that almost every message you sent or received you would help spread the message by encrypting it again and sending it to 2 others who didn't receive it yet, then you would be able to spread the message with a small O(log(n)) delay and an O(1) extra CPU cost per message per user for a group with n members.
But with your latter idea (the asymmetrically encrypted shared secret with the symmetric encryption for subsequent messages), the (powerful) central server could help distribute the message without having to trust it and without inducing the O(log(n)) delay or the extra send-to-a-friend encryption costs. On Wed, Sep 23, 2015 at 1:45 AM, David Beckwith <[email protected]> wrote: > Oh I get it! That's a great idea. Thank you. > > On Wed, Sep 23, 2015 at 1:22 AM, Rogier Mulders > <[email protected]> wrote: >> Several hundred people in a chat room does not sound like much privacy :) >> >> Just asymmetric encrypt some randomness with the pub key of every >> member and symmetric encrypt decrypt their messages. >> >> On 22-09-15 0:22, David Beckwith wrote: >>> Hi Roger, >>> >>> If there are several hundred people in the chat room it's gong to >>> take a few minutes to send a single message, plus eat up a lot of >>> CPU/battery. >>> >>> On Mon, Sep 21, 2015 at 10:37 AM, Rogier Mulders >>> <[email protected]> wrote: >>>> Just use all puplic keys of people in the chat to encrypt a message. >>>> Then all that have the corrospondig private key can read it >>>> >>>> On 17-09-15 9:26, David Beckwith wrote: >>>>> Hi, >>>>> >>>>> I'm just a lay user of the library and just joined the list, so >>>>> please forgive me if I say something stupid. But, I was just >>>>> wondering if there was a way to generate a shared secret with a >>>>> corresponding public key generated from the public keys of 3 or more >>>>> users. I think something to this effect: >>>>> >>>>> http://crypto.stackexchange.com/questions/1025/can-one-generalize-the-diffie-hellman-key-exchange-to-three-or-more-parties >>>>> >>>>> I would like to implement encrypted group chat without having to >>>>> trust a central server. I thought maybe a group public key + group >>>>> shared secret could be generated from N public keys, so that users >>>>> could encrypt their message just once with the group public key to >>>>> post to the group instead of having to encrypt N times for a group >>>>> with N members. Does OpenPGP.js support this use case? >>>>> >>>>> Thanks a lot! >>>>> _______________________________________________ >>>>> >>>>> http://openpgpjs.org >>>>> Subscribe/unsubscribe: http://list.openpgpjs.org >>>> _______________________________________________ >>>> >>>> http://openpgpjs.org >>>> Subscribe/unsubscribe: http://list.openpgpjs.org >>> _______________________________________________ >>> >>> http://openpgpjs.org >>> Subscribe/unsubscribe: http://list.openpgpjs.org >> >> _______________________________________________ >> >> http://openpgpjs.org >> Subscribe/unsubscribe: http://list.openpgpjs.org _______________________________________________ http://openpgpjs.org Subscribe/unsubscribe: http://list.openpgpjs.org
