froztbyte @ froztbyte @awful.systems

I'm @froztbyte more or less everywhere that matters

---

Read more

Posts

47

Comments

3,299

Joined

2 yr. ago

Xanadu’s micropayment-oriented transclusion-and-royalty system is impossible to correctly implement, due to a mismatch between information theory and copyright; given the ability to copy text, copyright is provably absurd

it kept being funny to me that even while xanadu had already shown the problems with content control the entirety of the NFT craze just went on as if it was full greenfields novel problem

The details lie in the devil, for sure…you’d want the price [of making a change to a document] low enough (zero?) not to incur Trivial Inconvenience penalties for prosocial things like building wikis, yet high enough to make the David Gerards of the world think twice.

some of these people just really don't know their history very well, do they

on a total tangent:

while xanadu's commercial-aspiration history is intimately tied up in why it never got much further, I do occasionally daydream about if we had, and if we could've combined it with more-modern signing and sourcing: daydream in the respect of "CA and cert chains, but for transcluded content", esp in the face of all the fucking content mills used to push disinfo etc. not sure this would work ootb either, mind you, it's got its own set of vulnerabilities and problems that you'd need to work through (and ofc you can't solve social problems purely in the technical domain)

has there been any meaningful advancement or neat new research in agoric computing? haven't really looked into it in a while, and the various blockchain nonsense took so much air out of the room for so long I haven't had to spoons to look

(separately I know there's also been some developments in remote trusted compute, but afaict that's also still quite early days)

much of the lore of the early/earlier internet being built is also full of some extremely, extremely unhinged stuff. I've had some first-hand in-the-trenches accounts from people I've known active from the early-mid 90s to middle 00s and holy shit there are some batshit things happening in places. often think of it when I see the kinds of shit thiel/musk/etc are all up to (a lot of it boils down to "they're big mad that they have to even consider other people and can't just do whatever they like")

I really don't buy the "billing mistake" line - they've been doing the same thing to many other community-org slacks. I've seen with my own eyes the mail that was sent to the ZA tech slack

will keep the offer in mind when I have the spoons and round tuits for it :)

I’m a bit split on this one

on the one hand, the post as first posted had a lot of “victimised” language (“omg slack is extorting us”) and frankly that felt like bait - esp as many, many volunteer-type orgs that have had similar slack setups have been taking a hammer for months now (as I posted before, a local ZA tech setup was one, and more recently that big k8s one too). there’s enough precedent here that expecting slack to have behaved otherwise (even “honourably”) seems to me to have been almost foolish

on the other, slack 100% only took action once this did hit hype and enough eyeballs, and only reacted since it was an embarrassment

but…yeah. slack hasn’t been a good option for public use for literally years now :|

it would be incredible if yud’s one of the types to lose his focus around sauce 3, would do a real kicker to the shine of his grift

someday(tm) I’ll get around to looking into getting a season pack from the states to here (which possibly might be distinctly non-trivial, and if it is I’ll have bother trying to figure out the logistics of it, which ugh)

yeah, image is compressed to hell. happen to still have the source pdf @yimyam@piefed.social?

is the consensus solution on how to subvert the acausal robot god

dunno if you've yet gotten to look at the most recent yud emanation[0][1][2], but there's a whole "and if the robot god gets too uppity just boop it on the nose" bit in there

[0] - I mean the all-caps "YOU'RE ALL GONNA DIE" book that came out recently

[1] - yes I know "emanation" is a terrible wordchoice, no I won't change it

[2] - it's on libgen feel free to steal it, fuck giving that clown any more money he's got enough grift dollars already

heard people reached those by just deleting tweets by hand.

yeah, the various backend interactions tied to web controls are extremely low-count limited

you could probably do it by smacking together a userscript (or whatever the fuck is the these-days version of greasemonkey/tampermonkey/??? to use) with a moderately simple algorithm.. open a window, click execute, leave it going by itself for however long it takes to get through everything. it doesn't have to do everything in minutes

I also heard blocklists put a high strain on the twitter so not going to look into removing that

probably the feed compute stuff only has this computational expense incurred for any displayed feeds (pruning off calculating stuff for long-enough-inactive users is one of the cheapest easy gains in that type of content feed), so this might not matter much. don't have enough insight into real ops there to know one way or the other tho

it's kinda hilarious how close "steelmanning" (as practiced by some) already is to this, but probably not far enough to be usable for that purpose on its own

thanks for linking this, was fun to watch

hadn't seen that saltman clip (been real busy running around pretty afk the last few weeks), but it's a work of art. despite grokking the dynamics, it continues to be astounding just how vast the gulf between fact and market vibes are

and as usual, Collier does a fantastic job ripping the whole idea a new one in a most comprehensive manner

from when I last looked into this: twitter 100% has[0] (unstated) web API ratelimits for various subservices[1], but getting direct API creds became a "give us your actual phone number" thing even before felon took it over...

so I just decided to tombstone my account by making it private, updating bio, and never logging in again

not willing to give them what they want for API access. might at some point go write some web automation to recurringly click a delete button? idunno

[0] - ....well, 4 years ago, "had". probably maybe still does, on whatever parts of the haproxy or whatever config didn't get absolutely fucking destroyed in felon's mania to rebrand it to "x" overnight (a process which failed hilariously badly for weeks and I still think fondly of to laugh at)

[1] - when going through the "your interests" list (hidden deep in settings), if you unticked too many boxes too quickly you'd hit a webserver-enforced ratelimit on request limits and then half the webapp would get a bit fucky for an hour. ratelimit was something like 30/min with a 1/m type token-bucket refresh. quite the shitshow

ruby's had this problem for ~2 decades now. like, the "rockstar dev" archetype literally became big directly because of ruby's popularity and perception at the time

I haven't been active in/near the ruby space for a number of years now so I can't speak to the modern details well at all, but I wouldn't be too surprised to learn that the various branches of it haven't really learned how to deal. I will say that I have seen some improvement over that period, but... yeah

just yesterday I saw this toot and now I know why

(I mean, they probably should’ve bounced the guy a decade ago, but definitely even more time for it now)

The protocol is built for a future in which AI agents routinely shop for products on customers’ behalf

as I was ranting in dm earlier elsewhere, the part about this that especially fucks me off is how much of this is not just simply unnecessary but also strictly worse than what we already used to have!

~15yo ago the entire bloody internet was awash in APIs and accessible interactions! hell, it's the whole reason shit like Yahoo Pipes and IFTTT became a thing!

(and then after that ~everyone made fucking fences to wall their gardens because they want to Capture Users! to this day I still don't know if it could've gone any other way under how capitalism operates, but fuck it sucks.)

meanwhile so many people (both those who've come up Touching Computers, as well as casual users, in the last 10~15y or so (who I typically refer to as the Cloud Generation) typically don't even have a conception of doing it any other way but The Billable Platform Way. I have long suspected that this won't hold out (it's a truism that at some threshold people will start asking "wait why am I paying for this?") and I am heartened by seeing some indicators of this starting to happen, but...... fuck. there's been so much damage from years of this shit

I still stay hopeful for change (esp. because this current way can't hold), but I also grimace about what's coming in the near future (because I know that a fair number of these platforms will be cognizant of the same problem)

this is one of those things that's, in a narrative sense, a great way to tell a story, while being completely untethered from fact/reality. and that's fine! stories have no obligation to be based in fact!

to put a very mild armchair analysis about it forward: it's playing on the definition of the conceptual "smart" computer, as it relates to human experience. there's been a couple of other things in recent history that I can think of that hit similar or related notes (M3GAN, the whole "omg the AI tricked us (and then the different species with a different neurotype and capability noticed it!)" arc in ST:DIS, the last few Mission Impossible films, etc). it's one of those ways in which art and stories tend to express "grappling with $x to make sense of it"

The idea that a smart computer will be worse at math (which makes sense from a storytelling perspective as a writer, because smart AI who also can do math super well is gonna be hard to write)

personally speaking, one of the ways about it that I find most jarring is when the fantastical vastly outweighs anything else purely for narrative reasons - so much so that it's a 4th-wallbreak for me ito what the story means to convey. I reflect on this somewhat regularly, as it's a rather cursed rabbithole that instances repeatedly: "is it my knowledge of this domain that's spoiling my enjoyment of this thing, or is the story simply badly written?" is the question that comes up, and it's surprisingly varied and complicated in its answering

on the whole I think it's often good/best to keep in mind that scifi is often an exploration and a pressure valve, but that it's also worth keeping an eye on how much it's a pressure valve. too much of the latter, and something(tm) is up

hot off the heels of months of “agentic! it can do things for you!” llm hype, they have to make special APIs for the chatbots, I guess because otherwise they make too many whoopsies?

guess that’s got tonight’s watching sorted :)

this post extremely quickly goes many places I didn't come close to expecting

impressive.