![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
timeasmymeasureBefore...
...and after.
Gonna try and talk through what's happening here.
( Read more... )Customizing takes a lot longer than we expected. But also it turns out we had a lot more ability to decide what we wanted than we feared. And it turns out we do like customizing it a bit.
A couple weeks ago, in a fit of frustration about not understanding object-oriented programming, we read several articles from the August 1981 issue of Byte Magazine explaining Smalltalk. One of those, "Is the Smalltalk-80 System for Children?" by Adele Goldberg and Joan Ross, contained this passage:
Contrary to the idea that a computer is exciting because the programmer can create something from seemingly nothing, our users were shown that a computer is exciting because it can be a vast storehouse of already existing ideas (models) that can be retrieved and modified for the user's personal needs. Programming could be viewed and enjoyed as an evolutionary rather than a revolutionary act. The frustration of long hours of writing linear streams of code and then hoping to see some aspect of that code execute was replaced by incremental development. Emphasis was placed on learning how to make effective use of existing system components (objects in the Smalltalk sense). Much of the teaching we did was to show users how to search for and read the descriptions of the many useful components we and others (and even new users) continued to add to the system.
Using resources within an already-powerful system is a highly effective and rewarding form of programming, and understanding that it is programming allows users to program more effectively within these systems.
We bring this up because, when you create a formatted document using a modern word processor, this is what you are doing: you are simultaneously creating data and creating instructions for how that data should be processed. (That's why it's called a word processor: it is doing work to convert the words into a format where they can be comfortably accessed.) When the Accessibility Awareness disabled.social account talks about using paragraph styles, it is because you, by using these library functions, allow other document interpreters – like screen readers and keyboard shortcut tools – to execute their own versions of these library functions and therefore process the data more usefully.
Obviously, word processing programming languages are special-purpose, like the bespoke forks of Lua implemented in fantasy consoles and game engines, or the Personal Home Page tool that some folks use for managing websites. It is a vast storehouse of already-existing components, designed to support its specific purpose.
…and we are dwelling on this thought today because we know many users of word-processing tools who do not avail themselves of these existing functions. And because we think this is, in part, because using these functions comfortably requires a degree of philosophical understanding of computers – an ability to learn a computer's language and translate your intentions into it.
As sighted writers, we are well familiar with the idea of marking off sections by having some big text at the top … but the idea that you can directly communicate, below the level of visibility, that a specific block of text is a section heading? And that the bigness of the text should happen not because you chose it, but because you chose to invisibly designate a line of text as heading?
That's a programming mindset.
And you have to learn that.
Okay, you might or might not know the constructed language Toki Pona, but we're certainly still learning, so I think we can construct a good metaphor.
( A hypothetical conversation about food. )Folks who learned HTML in the past fifteen years, feel free to chime in with more hot tips. Because we need them. Please.
(Link to Indiepocalypse HTML tips thread, because which a number of folks did, in fact, chime in.)
For reasons, we wanted to play with color quantization and dithering, and we found out about command line tool ImageMagick's color quantization/dither tools. Then we came up with a very silly idea: what if we took a given starting image and, using our limited knowledge of dither algorithms and complete lack of knowledge of image size optimization, tried to make the coolest versions we could under 32 KiB? Setting the limit to be filesize and not raster resolution means that every algorithm is going to create some compromise between size and detail … so maybe Floyd-Steinberg produces a clearer result at the same resolution, but how much smaller does it have to be to fit in the size limit?
We finally finished this freaking project, and it's such a big one that it's too big for a Dreamwidth post - click through to check it out.
Reading the Sandia National Laboratories nuclear waste warning messages report - specifically the Team A apprendix, the one with the famous passage. We made a post about that passage a while ago but now we're thinking about the task they were set.
Like, any D&D player will tell you that a big warning sign with a clear map pointing to the exact place you're being warned about is enticing as all getout. But what are you even supposed to do?
Like, that's the thing, right? Imagine you, the reader of this post, were supposed to make it so some rando looking for treasure a thousand years from now - as far from now as "Hwæt! wē Gār-Dena in gēar-dagum/þēod-cyninga þrym gefrūnon" is from now - would see the notice-board you put up and go, "Oh, I shouldn't dig that up"? How do you make it seem like a bad idea to try and get at something which clearly was made hard to reach? Even knowing it was a trash heap wouldn't help - archaeologists love trash heaps, love seeing what people had and threw away! It says here in this warning message that there's machine tools in that pit! We could use some machine tools! So what do you do?
And what do you do if you have to get that message across for ten times a thousand years?
It's easy to crack jokes about "this place should be shunned and left uninhabited" but there's something very compelling about looking at a manifestly impossible task and asking, "okay, but if we had to, what do we do?" It's a kind of bravery, to consider a horrible situation honestly and forthrightly, and do your best to manage it.
We don't have a solution for nuclear waste. But maybe we can try something. And maybe we can learn something from trying, even if we fail.
And maybe there's something comforting about spending time in a world where we do try, where we don't throw up our hands and say "it would mean a smaller yacht for Jeff Bezos so there's nothing we can do". A kind of grounded science fiction, not about starships but about dirt, and rock, and salt, and public health.
A suggestion we saw in the report that never really got turned into a joke was the visitor's center. The place where people visiting this very expensive public works site could go look at some placards, talk to museum staff, learn about the society that made nuclear waste and the society that tried to protect its children from it. Maybe watch a looping video with a CGI representation of the site and its design. Buy a hat with a picture of the berms embroidered on it.
Or not. Who knows what a visitor center would look like after ten thousand years?
There's a "no more turing tests" tag now.
This might be kinda obvious but we'll just stick it here anyway. Also, shoutout to acorn_squash, who made a perceptive comment about disability and neurodivergence in the last post.
Packbats' policy on AI content generation, one swear
Any machine built by scraping massive amounts of training data off the Internet without permission can get fucked.
Any machine designed to divert income from artists, writers, musicians, and other creators to companies, likewise.
That is to say: Midjourney, Stable Diffusion, DALL-E, Adobe Firefly, ChatGPT, Bard, and so on - all of them can go straight to hell.
If we share or boost something that uses these tools, please let us know. We will never make anything with any of those.
The thing is, "AI", in its modern deceptive usage, mostly means "neural net". And that makes things a little difficult for us as people who get hinky about the literal meaning of words, because the campaigners we agree with are saying things like "generative AI" and "no AI" and making Human Made logos for humans to put on human art that isn't made out of theft and exploitation ... and yeah, we support that, but that's not what we would say?
Like, we keep pointing to Adam Neely's collaboration with Dadabots. This was a neural net acting as a generative AI creating art ... that is based on two hours of bass playing by a musician who volunteered to have his work used this way, and that creates music that no bass player would ever play at a gig. It's not stealing anyone's creations and it's not stealing anyone's jobs. And it's not stealing credit - for as long as it existed, it had Adam Neely's face plastered on the thumbnail, it was explicitly made with his permission, and had a link to his video about it in the description. And, well, we said what we have to say about a hyperfocus on humanity.
So, like, the AI bubble is toxic but we don't have a pithy way to say it. Except maybe to call them plagiarism machines.
Anyway I don't think we made a post here on Dreamwidth about that shit, so let us say again that these plagiarism machines deserve to be destroyed. And as great as nuance is, it just lends emphasis to that conclusion.
the question "is this human" has expired - it has become a torment of con artists and cash grabbers
the one, perpetually remaking their machines to deceive us
the other, perpetually demanding that we become harder to fake so they can keep selling us to advertisers
the doppelgangers yell, "I am human!"
the marketers yell, "Prove you are human!"
we all tuck ourselves into corners trying to find respite from the noise
why? why are we being yelled at? why for humanity are we being yelled at?
Alan Turing started with "can machines think?" and went from there to "can machines make humans think they are humans?" but listen to his imaginary machines
his machines talk about poetry, arithmetic, chess, Mr. Pickwick and winter's days
his machines speak honestly to their interrogators, try to understand, try to help
our phone makes a noise after half an hour, because we asked it to in a way it understood
our friends speak honestly, because they want to us to know that they miss clubbing, they like umbreon, they think of a sunrise as a sun, rising
why should we put the people who traumatized us in the category with the friends who help us survive? why should we put the machines that exploit us in the category with the machines that help us survive? why is this the split we are asked to make?
why "are you human" instead of "are you good"?
why not "are you good"?
can you tell if something is good to you over a teletype connection? are you good to it?
(edit: followup)
(thread)
okay, no, actually, I think it's actively harmful when a page is intended to display static information - text and images you can't edit - and requires javascript to do it
any website building tool which functions that way is actively doing harm
no we will not explain
yes, we hate these things because they're rubbish
but we hate them more because they're born out of indifference and callousness at best, and active hostility at worst
Yesterday, we posted a fediverse checkbox poll on indiepocalypse.social asking, "Are .md Markdown files plaintext?". The poll received 49 votes, with eight respondents additionally replying to discuss details of the question, and the final percentages including 71% yes, 31% no, and 16% I don't know, with substantial overlap between all three of these.
(Also, 12% spiderman. Because Mastodon injokes.)
I think what makes it an interesting question is the term "plain text" (which apparently has a space, apologies). What is, and what is not, plain text? Let's run through some possibilities.
So, yeah. I think the moral of the story is that it's not obvious what "plain text" means, but hey ... it was fun asking.
1.
okay so the big problems with the modern wave of generative AI are that:
...but an additional problem is that, because they are intended to be used without warning, they force people to try and find intention, worldview, meaning, all the things we expect from our fellow writers and artists, in material which contains none of that.
and that's just abusing our assumption of good faith.
2.
like, okay
imagine a friend mentioned something bad that happened to them, and we replied "oh no!"
this is as nearly an automatic and thoughtless response as we can think of ... but consider what it would mean to our friend
at the very least, they can infer that:
if they had said the same thing to ChatGPT, and ChatGPT had said, "oh no!", would it have anything like the same effect? ChatGPT doesn't know them and won't remember them, it only understands that "oh no!" is a thing that is said frequently in its corpus in this kind of context. it is simply and utterly hollow.
Did we not write this post? We can't find a copy of this post.
Okay, so, the trick is this: each time you study from your stack of paper flashcards, you follow these steps.
What this means is that, as long as you're taking more cards off the top than you put back on the top, you always go through the cards that you need to learn, and you slowly but consistently cycle through the cards you know. That means you're reviewing all your old cards while learning your new ones, while only studying as many cards at a time as you're comfortable with.
It's not the scientific system that the scientific science-based flashcard programs use, but we hate all those programs because our memory is a sieve and those things don't care if we want to study more or not. This lets us decide what we've learned and what we haven't, and it's dead simple.
We have been struck with a sudden inspiration.
In the past - for example, while writing and editing our guide "naming yourself in toki pona" - we've thought of nouns for speakers as a static thing: if someone's head noun is 'kulupu', then you call them 'kulupu', end of story. And it's important that we understand and conform to the nomenclature of the subject we're discussing.
...but in the style of lipu pu, "Toki Pona: The Language of Good", the first official Toki Pona book, words can be more fluid that that. jan Sonja might say "meli li lili" to talk about a woman being little, even though that meli is definitely also a jan. Provided there is no misgendering or other deliberate misidentifying going on, there doesn't seem to be anything hostile about that.
So, you know how, in resources like pronouns.page, there will be lists of words that can be marked with hearts or thumbs-up or thumbs-down or other such indicators of mood?
What if tokiponists did that for words about them? Like, if we add this to our profile:
sina toki e mi la, mi olin e nimi ni:
- kulupu
- poki
- tonsi
...nimi ni li pona:
- ijo
- jan
- meli
- nasa
- tomo
...nimi ni li musi:
- kijetesantakalu
...nimi ni li ike:
- mani
- mije
- pakala (mi ken pakala! taso, kon mi li kon pakala ala. nimi "pakala Pakapa" li ike a!)
- sewi
...would that be clearly understood? Would people know that they can say statements like "meli ni li kulupu Pakapa" and be polite and accurate?
I think it would be good if this is something they could check as needed. Obviously, most speakers would simply grab "kulupu" or "poki" off the top and use one of those, which works fantastically and requires very little memorization ... but I think it's good to have, and good to have thought about in case people ask, "mi wile ala wile kepeken nimi 'meli' lon sina?"
wile. mi meli tonsi. mi tonsi meli. pona, pona.
Right off the bat, we're just going to link to Tamiko Nimuras's review on DiscoverNikkei.org, which is better informed than our opinion.
We were looking at the books available through OverDrive, a library digital lending service, when we saw Displacement by Kiku Hughes. Specifically, we recognized the art style - Kiku Hughes' guest comic about asexuality for Oh Joy Sex Toy was part of our journey to realizing we were ace - and we remembered thinking her comics were very good.
And Displacement was certainly good.
This is a comic about the internment of Japanese citizens in the United States of America during World War 2. It's a time travel story, but it is not about time travel, it's about history, and ignorance of it, and tangibilities of it, and aftereffects of it. And explicitly places this eighty-years-past atrocity in the context of more modern atrocities, and connects resistance to that to resistance to these.
It's historical fiction, first. And it uses the format well. It shows us insides of camps, in tangible details: smells, temperatures, censorship, propaganda, and resistance.
At a formative age, we read a 1981 science fiction novella by Vernor Vinge titled True Names, about a population of hyper-hackers doing operations around the world and maintaining the secrecy of their identity to protect them from being controlled by others through elementary magic spells computer exploits.
...come to think, it used "the Internet as a 3D immersive space" as a narrative device eleven years before the publication of Neal Stephenson's Snow Crash made "the Metaverse" a meme. Kinda funny.
Anyway, in retrospect, that novella really encouraged us to think about security of identity in a substantial way. (At least, once we started caring at all.) This is a story in which an action which restricts your identity to three million possible candidates is considered high-risk. The concept of what could expose personal information becomes very broad, when you're thinking about that level.
And then you get into the stories about people identifying physical locations by nothing but the background scenery, or incidental environmental details. Like that time Tom Scott challenged people to identify where he and Matt Parker watched an eclipse from, and they got it within feet.
On that level of investigation - the, to be frank, "ten thousand people decided to stalk you" level - we are shit out of luck. If we wanted to be secret at that level, we would have to burn the name "Packbat" altogether and start over. But ... increasingly, we are very careful about landmarks in our photographs, we don't talk about trash pickup schedules or the weather, and we do not name streets. We will say we are in the mid-Atlantic region - we will even say where within on occasion - but we do not want to be specific enough to phone book.
Because it's 2023, there have been a lot of harassment campaigns that jump from Internet to city streets, and we don't want to make it easy.
There's an interesting arc you can have with studying Toki Pona.
At first, you're learning words and phrases. That "toki" can mean "speech", "communication", "stories", and suchlike. That "pona" can mean "good", "acceptable", "approved of", and similar. That "jan" can mean "people", "characters", "humans", that kind of thing. That "jan pona" can mean "friend" - a person or people you approve of, that you like and endorse. That "insa" can mean "inside", "center", "stomach", and so forth. That "toki insa" can mean "inner monologue" or "thoughts". And learning all these things helps you understand what people are talking about.
And then you start unlearning them. Sort of.
Like, the thing about toki pona is that a lot of its strength is being not specific, is being contextual, is being personal. And you can translate the English word "think" with the toki pona phrase "toki insa", but there's a lot of things that a communication can be within - a house, a community, a back room, standard usage ... a lot of things. And "think" does a lot of work - I think that origami is delightful, but when I say that, I am saying that to me origami is delightful: musi pi lipu sitelen li pona tawa mi.
This isn't actually about Toki Pona - this is about Kyle Kallgren's analysis of the movie "Network", and the cover of "Land of Confusion" it ends with, and the idea of reading classic rock songs as saying something. Saying "things aren't okay, we're being lied to, and we need to stop the damage". Saying "justice for those American Indians who fight against poverty and police violence". Saying "the classist and racist status quo isn't actually a natural state, we can do something about it". Saying "The USA gave us a terrible life, sent us off to kill people who looked different than us, and left us with nothing when we got back, if we even made it back". All these words that got flattened into "angry" or "comforting" or "exciting" or whatever else, they were saying something.
People learn to ignore what a communication means.
Probably because if any of these people were allowed to be understood, we might not be okay with letting the rich get richer as everyone else kills each other.
Howard Beale was saying something. He was saying that you have to get mad because the other option is depression and not caring, and if we have any hope of not taking it any more, not letting all the evil be wrought upon us, we have to care.
And on a personal level, to paraphrase jan Sonja about Toki Pona: if a friend is a "jan pona", a good person? A bad friend is a contradiction in terms.
Words mean things. And caring about what words convey can mean caring about a lot else, too.
Another last-minute one, so, you're getting something short and mostly inconsequential.
In our post about our plans for hand calculations of π, we mentioned casting out nines and elevens. If you're not familiar with these methods, you might theoretically be interested in what they are. If you are familiar, you might not have given any thought to a pretty fundamental question for our application: how do you do it on a fraction?
This math probably won't be terribly readable, but we still want to give it a go.
( A quick summary of what casting out nines even is... ) ( ...and the bit about how to do it on decimals. )"The Salt Merchant and His Ass" is not a particularly famous Aesop fable, but it's a funny story and we recently translated it into Toki Pona.
By our count, the English translation we were working from (George Fyler Townsend's, basically), was about 169 words long.
By our count, our Toki Pona translation was 114 words long.
Now, some of that is editing. The Fyler English translation was a little florid, and we could have stood to be a bit more expansive when writing ours. But the obvious thing to expect when going from a language with a typical average vocabulary of twenty thousand words to a language with only about a hundred and twenty would be, y'know, the same ideas taking many times more words to express, and that's not what we're seeing. Toki Pona is really remarkably good at expressing the essentials of these old stories, despite having far, far fewer tools to do it.
I think that's really neat.
Edit 2023-08-03: We poked around a little more looking at other texts, and typically our Toki Pona renderings are longer than the English versions ... by somewhere between 5% and 30%. (For example, the IPA edition of "The North Wind and the Sun" has 113 words and our off-the-cuff Toki Pona telling would have about 144 if we didn't run out of space.) And all of these are written with the Sonja Lang's basic 120 words - we didn't even add "kin".