Email or username:

Password:

Forgot your password?
Sarah Jamie Lewis's posts Post Back to profile
Sarah Jamie Lewis

Thank you everyone who commented on my thread yesterday regarding decentralized search.

It became clear to me this morning that in order to have a useful discussion on the topic I needed to gather my notes into a more useful format. So I spent this morning writing them up.

Notes on Decentralized Search: sarahjamielewis.com/decentrali (pdf / draft)

This is very much a draft that captures where I am in my thinking regarding the problem - less so about the prototypes I've been tinkering with.

6 March at 20:55 | Open on mastodon.social
4 comments
Sarah Jamie Lewis

These are very high level notes, and it is worth noting that the document itself is subject to updates - I will likely augment it as the conversation progresses, and I am very open to collaboration and critiques on missing aspects.

Also, as noted above, I did not include any of my more concrete ideas for solutions to some of the challenges outlined (I'm putting something together for that too)

However, this now allows far more room to set out the context for a discussion than mastodon permits.

6 March at 21:05 | Open on mastodon.social
Jamey Sharp

@sarahjamielewis Fantastic starting point, thank you for writing this up!

Regarding proof that someone indexed an authentic HTTP response, I had thoughts on that a while ago, though I'm not sure my idea is practical for this (or any) use: jamey.thesharps.us/2020/06/13/

I love that you cover query privacy and the desire to do ranking locally. I mentioned yesterday that I'm interested in using data from my local history and bookmarks during ranking and wouldn't want to send those elsewhere, and it sounds like you're on the same page for other reasons.

@sarahjamielewis Fantastic starting point, thank you for writing this up!

Regarding proof that someone indexed an authentic HTTP response, I had thoughts on that a while ago, though I'm not sure my idea is practical for this (or any) use: jamey.thesharps.us/2020/06/13/

6 March at 21:48 | Open on toot.cat
Gary Blosser

@sarahjamielewis Thanks for sharing, I learned quite a bit!

6 March at 21:30 | Open on infosec.exchange
Rob Landley

@sarahjamielewis I am reminded of the old quote "I haven't got a solution but I certainly admire the problem".

I remember how searching gnutella was both crazy slow and unreliable, with two consecutive searches producing different results.

I remember when the coinbros started complaining that having the entire ledger locally was way too big and computationally expensive, but relying on servers to say who had what was untrustworthy...

Lots of existing work in this space. Few if any solutions.

6 March at 21:32 | Open on mstdn.jp
Go Up