@aelaraji@aelaraji.com So what is it about @sorenpeter@darch.dk’s feed that’s screwed with your client? (Jenny?) 🤔 Kind of curious now 🤣
@aelaraji@aelaraji.com Yes, according to the spec we wrote for Archived Extension:
The second value of prev is a name relative to the base directory of the feed’s URL in url (more specifically, in the URL that the client used to retrieve the feed). In the example above, prev would evaluate to the full URL https://example.com/twtxt-2021-10-18.txt for HTTPS and gopher://example.com/0/twtxt-2021-10-18.txt for Gopher.
@aelaraji@aelaraji.com LOL 😂
--fetch-context
thingy: It can now ask Yarn pods for twt hashes.
@quark@ferengi.one It would also be possible to use the search engine here too I think 🤔 i.e: https://search.twtxt.net
--fetch-context
thingy: It can now ask Yarn pods for twt hashes.
@quark@ferengi.one Looks like that would work according to the patch I just read 👌
These then become useful in filters like what you see here:
It’s useful to know however that such feeds are actually marked as type=rss
(e.g: https://feeds.twtxt.net/slashdot/twtxt.txt), just as feeds like @tiktok@feeds.twtxt.net are marked as type=bot
@aelaraji@aelaraji.com Ahh that’s interesting! 🧐 One of my original goals when I started out building Yarn.social was to also be a source of news, blogs, and whatever else that could be roughly/easily translated into a Twtxt feed. I’m not sure if others do something similar, but that’s why I built feeds.twtxt.net, which consumes RSS/Atom and produces Twtxt feeds.
My only desire one day is to build a “Feed Builder” of sorts that allows one to say, for example, construct a Slashdot feed but without AI hype, or as another example, a BBC/ABC feed that’s a digest once or twice per day.
@bender@twtxt.net Ack 👌
@aelaraji@aelaraji.com Good man 🤣 I keep getting this bloody AI hype from various news feeds I subscribe to via Twtxt like Slashdot cough 🤦♂️
@bender@twtxt.net Hahahahaha 🤣 Me neither 🤮
@bender@twtxt.net you’re probably right, but by that argument, cryptocurrency should be innate too right? as we no longer hear about that gold awful hype, right? 🤔
@movq@www.uninformativ.de Thanks mate! 🤗
@movq@www.uninformativ.de Haha same and I’m sick of it! It’s ruining innovation in anything else 🤦♂️
--fetch-context
thingy: It can now ask Yarn pods for twt hashes.
@movq@www.uninformativ.de Nice one 👌
@aelaraji@aelaraji.com it’s definitely your social bubble. 🤣 you need to use twtxt more as your daily driver 🤣
That’s an interesting side effect to the new Discover feature that I added sometime ago that only displays one post per feed. That is when you’re not logged in and viewing my pod’s front page. You can pretty easily and roughly see what the monthly active view account is just by looking at the pager size. 🤔
Amazingly though it seems to be slightly better to VPN in. 🤔
But you know speedtest.net I believe is a bit of a liar and I’m quite sure they do something to make sure the speed test come up good even remote areas the real speed test my actual surfer infrastructure is quite piss poor 🤣
Even though we’re quite a ways from any suburban areas, even with the Internet access via cell towers this poor, using my pod is still very snappy. 👌
When will the AI hype die down?
@lyse@lyse.isobeef.org Thanks!
@falsifian@www.falsifian.org One of the nice things I think is that you can almost assuredly trust that the hash is a correct representation of the thread because it was computed via our content, addressing in the first place, so all you need to do yes copy it 👌
@falsifian@www.falsifian.org Yeah that’s why we made them short 😅
@falsifian@www.falsifian.org I think I wrote a very similar program and go myself actually and you’re right we do have to change the way we encode hashes.
@falsifian@www.falsifian.org All very good points 👌 by the way, how did you find two pieces of content that hash the same when taking the last N characters of the base32 and coded hash?
@off_grid_living@twtxt.net Aww thanks! 🤗
There are certainly improvements that can be made to this tool.🤞
@aelaraji@aelaraji.com Have you considered https://git.mills.io/yarnsocial/twtxt2html
@bender@twtxt.net Thanks! 🤗 – I know it will 🤣
I think so 😅 Thanks$!🙇♂️
@aelaraji@aelaraji.com Hah interesting 🤔
@xuu What’s the keyoxide thingy you wrote/built? 🤔 What’s your URI/profile? 🤔
@aelaraji@aelaraji.com Sounds like it would work 👌 Though I’ve not tried or invested anytime into proofs and claims type things so far 🤔
@aelaraji@aelaraji.com Nice write up!
@aelaraji@aelaraji.com how would that work exactly? Does that mean then that every user is required to have a cox side profile? Who maintains cox site? Is it centralized or decentralized can be relied upon?
Ford, the company can honestly go fuck themselves! No one ever asked or even thought to themselves:
Gee I wish my car would listen to my in-car conversations and serve me ads.
@slashdot@feeds.twtxt.net i’ll get fucked! The US patent office should ban this immediately.
@xuu True 😅 I guess it comes down to our risk appetite and the attack vectors we’re trying to solve for 🤔
@bender@twtxt.net yes I agree.
url
field in the feed to define the URL for hashing. It should have been the last encountered one. Then, assuming append-style feeds, you could override the old URL with a new one from a certain point on:
@bender@twtxt.net there is a certain simplicity to that. 😅
@xuu it’s not really strictly required if we’re just talking about identity though right? If we’re talking about encryption then yes I agree rotate and keys becomes very important if you want to have attributes like perfect forward secrecy.
@xuu that could work too, but that requires a random value, a set of keys and signature verification of the value, which I don’t really have a problem with.
@xuu yes I’m less concerned about solving the integrity part of the problem of whether we can trust that the content of a feed is actually written by certain author, however, that’s not to say that we shouldn’t think about also leveraging keys to be able to do that maybe it’s an optional feature?
What were the recommended mitigations?
IMO we just have to fix the identity problem and figure out how to detect or support edits.
@sorenpeter@darch.dk No, this is what I want to avoid. For many reasons I stated before, content addressing or hashing is far better here for threading in a decentralized way.