On 04/11/15 22:57, Bob Ham wrote: > On Wed, 2015-11-04 at 07:49 -0500, Steve Dougherty wrote: >> On 11/04/2015 07:41 AM, Steve Dougherty wrote: >> ... >>> Did you see https://wiki.freenetproject.org/Research_challenges or the >>> section titled "Do we understand Freenet?" on toad's blog? > No I missed those. > >> Also this thread: >> https://emu.freenetproject.org/pipermail/devl/2014-March/037504.html > That's all helpful, thanks. However, I'm still not entirely sure which > are the "fundamental" problems Toad spoke of, which are security > concerns and which are wish list items. Also, there's lots of > information distributed in different places. It would be useful to have > an answer to my question, if only to consolidate the information. Okay, I will give some here, since people are being a bit more civil.
In brief: - Opennet is hopelessly insecure. (Not solvable except by using darknet, which may be intractable socially) -- MAST appears to be less of an issue since I dug into it a bit. However, it is still cheap to watch everyone on opennet. -- Tunnels are possible, but important to fix darknet first. Tunnels on darknet might make it less "invisible", but IMHO this doesn't matter, see below. -- General Sybil problem: *Every* resource is cheaper for an attacker than a lowest common denominator new user. Unless we can e.g. charge for entry to opennet. -- Key priority is to make darknet easy IMHO. - Darknet security: Pitch Black (believed to be solvable). - Load management: Performance. Major mechanism design (incentives) problems: The Patch is actively used in practice. Will look into this this year. -- IMHO performance matters. - Darknet doesn't scale. This *may* be fixable, there are some leads. Practically it's an irrelevance right now! - Forums don't scale. Maybe we can make them scale sufficiently. Microblogging-style may work better but we still need something like search. - Search problem in general: No usable search. Forums rely on something like search, but it doesn't scale. Major research problem. Lots of potential apps. - Mobile code is fundamentally unsafe on Freenet (although tunnels may help here), so need to write official, reviewed plugins for everything people take for granted on the WWW. A proper plugin API would help with this. Long term problems: - ANY p2p network can be blocked cheaply. - Social/political/moral/legal difficulties in getting people to use Freenet in general and darknet in particular. -- Freenet isn't worth much unless it has users. Florent is right about this. Performance and usability help here. -- In general I don't expect *any* significant collaboration with other projects, for this reason. - Everyone's mobile. Mobiles are designed to talk to The Cloud. Freenet can't do ad hoc, it needs uptime. Worse on darknet. Hardware router-boxes may be a solution. There have been discussions about radical long-term ideas such as high-latency stego transports (Sneakernet etc), long term requests, publish/subscribe etc, which may help with some of these problems... However, we have made reasonable progress in the last couple of years: - Data persistence is massively improved thanks to published work implemented summer 2014. - Lots of usability and reliability problems went away with db4o. Some more sources: - https://freenetproject.org/help.html#attack - My flog, search for "What are the larger problems?" (probably more, but needs digging, sorry) - https://wiki.freenetproject.org/Roadmap (admittedly outdated) - The bug tracker (hard to use though!) Also regarding rewrites, one of the things we lost with 0.7 was China, admittedly they were blocking us already, but we were big in China in 2003, hence one of the reports... Today that's naive, sadly... Most of this is on the research challenges page.
signature.asc
Description: OpenPGP digital signature
_______________________________________________ Devl mailing list Devl@freenetproject.org https://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl
