> -----Original Message----- > From: arne_bab at web.de [mailto:arne_bab at web.de] > Sent: Thursday, April 23, 2009 11:21 AM > To: devl at freenetproject.org > Cc: xor > Subject: Re: [freenet-dev] Non-convergent encryption kills > easy filesharing > > Am Donnerstag 23 April 2009 09:25:15 schrieb xor: > > > -----Original Message----- > > > From: devl-bounces at freenetproject.org > > > [mailto:devl-bounces at freenetproject.org] On Behalf Of Arne > > > Babenhauserheide > > > Sent: Thursday, April 23, 2009 9:14 AM > > > To: devl at freenetproject.org > > > Subject: Re: [freenet-dev] Non-convergent encryption kills easy > > > filesharing > > > > > > Am Mittwoch 22 April 2009 18:26:05 schrieb Matthew Toseland: > > > > block [16:39:31] <toad_> duplicating the top block can be done > > > > with SSKs very easily [16:39:40] <toad_> but with CHKs > it requires > > > > much longer URIs [16:39:43] <toad_> is that a problem? > > > > [16:40:04] <p0s> how much longer? > > > > [16:40:10] <toad_> CHK@<routing key>,<decrypt key>,<extra> -> > > > > CHK@<routing key 1>,<routing key 2>,<routing key 3>,<decrypt > > > > key>,<extra> [16:40:29] <toad_> i.e. at least twice as long > > > > > > I stopped reading at about the middle, so it might be > that I missed > > > something, but why do you care if CHKs become longer? > > > > > > They are already so long that noone types them (I hope), and I > > > really don't care if the key I copy-paste has two times > the length. > > > > If they are exchanged via Freetalk it will bloat the messages. > > Further, I was told to implement CHK messages in Freetalk because > > messages could not fit into SSK. So now Freetalk uses SSK message > > lists to publish CHK URI of the messages... If the CHK URI become > > insanely long then the message lists will be bloated very much and > > maybe only five messages or so will fit into a SSK message > list. Sucks. > > So why don't you take the approach of filesystems and add an > option where the message list SSK can contain a link to a CHK > which contains all message CHK URIs?
1. Too complex. 2. The problem is only shifted to another place. The problem with message lists is that they must fit as many message references into a single SSK block as possible to gurantee high network efficiency. You cannot tell me to solve the space problem by INCREASING the used space by adding an extra CHK block. > > (that's what ext2 does with inodes to support almost > arbitrarily large files, though each inodes is only 4k in size) > > If the efficiency of CHKs should rise by more than a factor > of 2 when using the large URI instead of non-republishable > URIs, freetalk will even profit from this (ping time wise). > > But I don't know the implementation details of freetalk, so I > don't know if this is a possible option for you. > > Best wishes, > Arne > -- > -- Ein W?rfel System: http://1w6.org - einfach saubere > (Rollenspiel-) Regeln. > -- Infinite Hands: http://infinite-hands.draketo.de - singing > a part of the history of free software. > -- My stuff: http://draketo.de - stories, songs, poems, > programs and stuff :) > > -- PGP/GnuPG: http://draketo.de/inhalt/ich/pubkey.txt
