39
votes
If you could completely refresh something and rebuild it from the ground up, what would it be and why?
A lot of things we live with have significant technical debt because they were designed and implemented without modern knowledge and understanding.
Knowing what we know now, in the present moment, what would you be interested in fundamentally redesigning if you could?
This does NOT have to be technology related, by the way, though it certainly can (anyone want to talk about redesigning usernames and passwords -- please?). It can pretty much be anything: NASCAR races, art criticism, specific social norms, sunglasses, etc.
In your explanation, don't just share what you're interested in tearing down, but how you would rebuild it for the better. What improvements would your methods bring to the table?
This is pretty abstract — I'd rebuild the ideas and words of computing to be about resources and transformations of those resources, rather than about features.
At a low level this would solve a lot of philosophical problems I have with some software. Think of a word processor that only works with an internet connection; now imagine "internet" is a resource you give to the program, and all of a sudden it's a bug. Not just "why should the program need this?" but "why should I the user give the program this?"
It goes further and further down the technology stack. Files are resources to be accessed. Not just files, the very concept of a filesystem is a resource too. "… Then it opens a file. / The file might not be accessible. / Then there's nothing to do. / But what happens in that case?" Connections to other computers, processing time, visual space on a monitor, user input — all resources. Not "things I the developer need", but "things I the developer can use".
I've seen hints of this in Erlang (connections to agents are resources) and Sculpt/Genode (computer hardware and some other abstract resources) and cgroups (CPU, memory, IO resources) and cloud hosting (virtual servers are atomic resources). I think there's room to put this in software generally, and think of software abstractions simply as providing different kinds of resources. For example, flat memory → memory heaps → manually-managed objects → automatically-managed objects; network card → IP → TCP → TLS → web pages. Complicated web apps make use of memory and time, both of which are resources with budgets.
I know you can think of things this way right now, but it's a little unnatural. We think of software (computers) as having features, when in fact they transform resources.
I think this would travel up to less savvy users too. If you can base everyone's understanding not on "the computer can do [this]" but rather "if I give the computer [this] then it can do [this]", then that makes people more morally engaged with what they're providing to what. It's not weird to ask "wait, why am I giving the resource of my physical location to this social media app?" Computers become less monolithic: anyone can make active decisions and control what they grant to whom.
Thats a fascinating take.
Ok just for my clarification (I'm a tad thick) - I have this silly plan of making a RGB to SVG image converter, beyond just internally how can I communicate the resource/transformation ideal to the user in a way that not only makes sense but perhaps push this ideal?
Frankly I don't have any well-thought-out ideas for how to take this to the desktop/GUI. Most of my time with this is spent grumbling about the low levels. :-)
But here are some ideas! Take a look at these Handbrake screenshots. (Handbrake is a graphical video transcoder.) The input file/title/chapter selection is fairly good, except that once I select a file, it's not clear at a glance exactly what's inside of it — that is, what's in the resource I just provided. I'd expect to see clear title/chapter/track/whatever metadata right next to where I selected the input file. Those subresources are inputs too.
However, the "Destination" is totally disconnected from the metaphor. The user has to separately select a folder and a name for the output. "When you press 'Start', I the program will write into this directory." That's backwards. The program is offering you a new file, so the "Destination" section could simply look like a file you can drag into a regular folder. Dragging it into a folder (or doing something similar with a keyboard) is what should start transcoding, right? Like step 3 here, but within an application window.
Actually, as long as I'm brainstorming — after the chapter/time selection, suppose that the video, audio, and subtitle tracks from the input were in a list. The user will produce another file with tracks of its own, so let's put those tracks in a list too, and let's draw arrows showing the conversions between the tracks! That's where conversion parameters belong, on those arrows, between the file you gave and the file being offered.
Maybe that's somehow applicable to your application? I don't think you have to extend any desktop or visual metaphors too much. It's mostly at the file–application (mine–offered) boundary. More creative or productive applications like editors are supposed to take ownership over a file, which I don't think fits into this ideal as well.
Here's one way to look at it. I'm not sure if this matches @wirelyre's intent, but maybe something like this:
Overall, your tool would transform an image resource into a vector resource.
On a more granular level, if your tool does an analysis of the image for the purpose of doing image segmentation to start off with, that takes resources. You could do the image segmentation in a small amount of memory (a resource), but it might take longer, or you could use a larger amount of memory and have it happen faster. You could do it slowly on the CPU (a resource) or quickly on the GPU (another resource). Or maybe segment different parts of the image in parallel on multiple CPU cores (resources).
Next is the process of transforming the pixellated outlines of those segmented objects into curves. That leaves a lot of the same choices as above for which resources to use.
Finally, you have to put the output somewhere. Either write it a file (a resource), send it over the network (a resource), or display it on the screen (another resource).
Yes! That's exactly how I would describe what a developer could be thinking about.
I'd change how we teach languages in school in the UK. I assume its some left over classical education along with a fair bit of faux-"utility" added to the mix and sheer apathy has left it in place so long when it clearly doesnt work. In schools, this was true 20 years ago when i was there and is still true for my sister in recent years. You get a choice of French or Spanish because they are "useful", after a couple of years you get to add another language, usually German or the one you didn't pick, smart kids get to add whatever asian language is in vogue at the time. EDIT: In places where there are a large amount of non-english heritage pupils their heritage language is generally available. Eg, lots of Poles near me, so Polish is offered.
The problem is that no language is useful in the same way English is "generally" useful. Most dont go to France on holiday, none go to Germany. Kids arent exposed to foreign languages online unless you go searching for it. There no foreign language music on the radio or in TV/Movies. They have no exposure other than the 2 hours a week doing grammar or learning basic vocabulary. Anyone who has learned a language knows how soul destroying text Grammar drills can be.
I'd change the lessons to be 3 hours a week instead of 2 and i'd let kids experiment with languages. For the first year i'd teach them different aspects of languages, how various languages sound, what a case is and what writing systems exist. I'd expose them to a couple dozen languages. Then i'd let them choose their language. I'd use a company like Assimil to a produce a variety of books for them to use, lots of audio, a small amount of writing and other "fill in the blanks" for grammar. I'd expose them to as much media as possible in the language in the computer room. Let them watch YT or TV or read simple books or comics. in their language. I'd set up groups of penpals from other countries for them to interact with.
Probably unrealistic but its got to be better than the massive waste of time language lessons currently are for 99% of pupils.
Your description of language teaching in English schools is consistent with my experience, although at my school it was even more limited - just French and Spanish, although I think you could get a external tutor for an exam in Polish or Urdu if you were already more or less fluent.
I certainly wish we were taught language at a more abstract and linguistically rigorous level. I would've liked to have been taught in school what a phoneme is, how sounds are defined in terms of articulation, along with perhaps a little IPA. I never actually understood properly what the difference between a vowel and a consonant was until after I had finished my French GCSE. I also wish we were taught a little more grammar. I think older generations were left a bit traumatized by tedious rote learning of Latin case declensions and gerundives and thus there was a reaction away from the denser parts of grammar in the school syllabus, but I would certainly have appreciated actually being taught, for instance, how to identify a direct object.
I do however think that your description of a better system for language learning in schools is perhaps a little impractical. The self-guided approach you describe could plausibly work for the most driven and engaged students, but I am fairly sure it would leave most floundering. If you give a fourteen year old a computer and tell them to choose a language and learn it, I would not put good odds on them getting very far. In addition, it is essentially impossible to measure progress and assess aptitude across such a range of languages, and learning is substantially harder when one does not have direct access to a fluent speaker. Besides, there is something to be said for practicality, and French and Spanish definitely have more real world application than the sort of languages that one might pick on a whim.
A really big issue I see with language teaching in schools is that our schooling system essentially requires students who aspire to study for degrees in STEM fields - that is, a huge proportion of more able students - to focus exclusively on Maths and Science. I was a respectable, if not stellar languages student, and did achieve with much hard work an A* in my French GCSE. I would have loved to have continued those studies to a higher level, which would perhaps have allowed me to gain a genuinely useful command of the French language, but knowing that I wanted to study Computer Science at University, I was effectively forced into choosing the classic tetrad of Maths, Further Maths, Physics, and Chemistry. Therefore my French abilities have slowly atrophied since age 16 and are now depressingly modest. Surely there is a better way than this.
English spelling/orthography is a pretty obvious one. There's no reason to spell words based on how they were pronounced 300+ years ago or whenever. I don't have much to say in terms of implementation, though. I guess I'd just want it more phonetically consistent with whatever dialect best matches the written standard? I bet you could get rid of a few letters as well.
idk, I think it's nice to have that history with a language, words feel like they've been around, and have real meaning behind them.
If everything was sanitised to be just phonetic or whatever other metric you'd use, it'd get pretty close to new speak from 1984, not a great look. I say this as a pretty heavy dyslexic that struggled all through school (and continue to) with spelling and reading, it would be oh so much easier if were phonetic, at least to my accent.
I say, let language be this evolving body that reflects the past but also has the capacity to change as well, new words come and go. Totally rambling; I wonder if, with widespread spell checking, we've disrupted this cycle of language evolution. Sure new words can be brought in, but it doesn't leave room for spellings to change, losing a few letters for instance. It'd be interesting if google or someone has some statistic on the most common misspellings of words and a decision somewhere can be made if they should be made legitimate.
I disagree strongly that simplifying spelling is akin to New Speak. The point of New Speak in 1984 was to limit the vocabulary of the citizens. The idea being that if people couldn't even formulate concrete thoughts about what the party was doing wrong, because they lacked the words to express those feelings, then they would be more compliant overall.
Changing spelling doesn't affect that at all, quite the opposite. English learners frequently struggle with spelling and pronunciation, and making that process easier could potentially improve literacy amongst the types of people who struggle with the English language. That includes not just foreign learners, but children who might otherwise have more interest in reading were it not confusing for them, and adults who managed to get through high school without achieving significant literacy. Also worth mentioning that reforming English spelling to be phonetic would be a boon to AI, which has a hard time understanding the pronunciation as it is, made all the more difficult by inconsistent rules.
Controlling the spelling of words is not a revolutionary idea either. Plenty of countries and languages have undergone reform to better reflect modern pronunciation, and some have even done so multiple times. To be frank, I think it's likely that without the advent of word processing sometime around the 80's, that there would have been a greater push in the English language to modernize spelling. For the simple fact that a modern world without spell-checking in a language as convoluted as English would be a nightmare, a nightmare whose simple remedy would be to mandate that words are spelled more or less how you think they should be.
The other goal is to make all citizens speak the same language, leave no room for accent or interpretation, no individuality. That speaks to my next point, I only alluded to it before, but I'll expand here:
Easier for whom and from what country? What pronunciation would you pick as the "standard" English? Would it be as spoken in; Australia, Canada, South Africa, India or The Caribbean, and so on, not to mention regional or racial lines you can draw. Each have their way of pronouncing the same words and could not in any world be distilled down to one phonetic spelling. You'd have to invalidate all but the one you choose to pick as your standard, and thus, we approach newspeak uniformity.
But ok, lets hand wave away this problem and say that English just is phonetic, how long can it stay that way? You'd either police it's learning and keep everyone's pronunciation inline (1984-ish), or let people speak as they like, quickly arriving at the state we're in now; not phonetic and following some arbitrary set of rules. At least now the rules have meaning, follow a history.
Either way, I think it's futile trying make it phonetic -- English is too global, use by too many.
I think schools failing at teaching English is more complicated then, "it's hard, so I give up" for children and adults alike. Maybe deemphasizing the importance of pedantic spell checking could make it easier on students, idk. Besides there are far more complex languages than English that get taught just fine.
Yes reforming everyone's pronunciation to be the same would make things easier for an AI to understand us, but again, why kill such a large part of human expression, for an AI? Nope, they are built as a tool for humans, not the other way around.
As for rules? Unless you have an insight I don't, it makes no difference to an AI what letter combination represents a sound, it's still got to learn it and adding extra rules is not an issue for a machine who's one advantage over humans is the ability to access many stored things.
I totally agree, I'd not have written this, were it not for my spellchecker -- it gives me the confidence to express myself online when I'd fail a spelling test given to a highschooler.
Well I agreed, until you got to the convoluted part, yes English has many words and spellings to remember, but it's stimple in terms of; tens, gender, possessive, all these other rules that make French, German or Russian, (and other no European languages I'm sure, It's just I know too little about them) more structured and logical, but certainly not simpler to learn or use.
We're talking about reforming spelling though, not enforcing some standard pronunciation. Spelling doesn't have to do with how you say a word, just how you write it. As long as there is correspondence between the sound and how it is written, reform will be useful to you no matter your dialect. The problem is that under current English orthography there is not correspondence, and so the same sound can be written ough, ow, oh, o, etc. If we were to always represent this sound as "o", this would be useful regardless if you pronounce the sound as "eh", "ah", "ih", etc. Do you see what I mean? I suppose this wasn't really made clear in the parent comment, though.
To me these two contradict each other. Your pronunciation, accent or dialect, will change how a word sounds to you, ie. what letters you think should makeup its spelling.
You mention the "o" sound, the examples you list don't sound the same to me, I hear a distinct difference between ough (as in "wrought"), and o (as in "onomatopoeia"), for instance. I realise that depends on my pronunciation, and that's my point. How would you reform the spelling of "wrought", without making an assumption about it's intended pronunciation?
I was thinking borough.
The vowel in wrought would be a different grapheme, such as ah or oh.
Ok, let's use borough in an English vs American accent, just focusing on the ough sound, I hear an American accent rhyming it with furrow, or burrow, but not the British one, when you standardise do you use the same letter combination (grapheme?) for furrow as for borough? Wouldn't that decision exclude one side of the atlantic?
Assuming I'm not going to get monkey pawed into an apocalyptic nightmare, I'd rework the concept of Global Economics to separate the concepts of resources, wealth, and achievement outside of the idea of "currency" and aim for an end goal of sustainable fulfilment and satisfaction for now and the future. That's probably going to fail because humans be greedy like that, but I suppose it would be nice to have it on paper.
There have been attempts at splitting the term "wealth" into "wealth/illth". Obviously, it didn't quite catch on back then, but perhaps we could try it again?
https://en.wikipedia.org/wiki/Illth
I've got about 9,000 bigger ideas, but I'll start small:
My house. We love our house, it's a great layout in a great neighborhood (semi-urban suburbs, before they went insane cul-de-sac. I want to burn it to the ground and fix everything wrong with it:
There's tons more that needs fixed, improved, or tweaked if I had the time and/or money. I'm lucky enough to have escaped renting. Not wealthy enough to bring this house to modern glory. I'd basically have to tear the whole thing down and rebuild, which would cost approximately double the price of the house, and maybe increase the 'value' by 20%.
The first thing that came to mind with the original prompt was something very similar so I'm gonna piggyback on your post. I want to completely rebuild my shop. It's a carport, so right off the bat it's not an ideal shop space because it lacks four walls. It was used by my stepdad as a woodworking shop prior to him actually getting a shop. I've kind of inherited the permanent fixtures along with a good deal of old clutter.
I used it as a glass shop for a while, but the lack of walls means lots of dust, and dust + glass + heat doesn't work too well. Everything is always covered in at least a few layers of dust. So I want to completely clear it out, reroof it (a 15 year roof that's gone 30 years, and it's definitely not leak-free), and redesign my work stations from the ground up. I want to be able to store everything in dust-free conditions, minimal horizontal surfaces (they're a magnet for clutter) that aren't contained, and a bit of modularity. Glass isn't the only thing I want to work on, so it'll be super useful to have the ability to convert workspaces depending on what I'm working on. And most importantly, space to do things. It's so confined right now that I can barely move around in there without bumping into stuff.
I don't care if it takes me the rest of lockdown. I can't afford to rent shop space in this area, it's just too expensive. This is the only solution I can think of to actually be able to make things again.
The cost of switching is not limited to tech debt. Look at the Dvorak keyboard or Esperanto, both objectively better designed than what they each intended to replace but have not been widely adopted after decades of constant use by a small group of people.
You're probably already aware of this, but Ido is the result of an early-20th century effort to address these perceived shortcomings of Esperanto. Although originally concieved (and subsequently rejected) as an reform, it sprung off on its own, albeit with a speaker population of < 1% of Esperanto.
-ojto-i, off the top of my head)-o, with feminine forms created by replacing the ending with-ino, and masculine ones with-ulo)Wikipedia's comparison of Esperanto and Ido is pretty informative for those who want to know more.
(disclaimer: what I know about Ido comes from various websites and articles - I've never studied the language, although it seems fun)
Could it be, a Conlang Critic viewer in the wild? I know very little about conlangs, but for some reason have been binging Jan Misali's videos for weeks now.
I actually don't disagree with you. Esperanto is not perfect. I think Ido shows the imperfections pretty clearly. And I have no doubt that one could design a conlang better suited to the task of being an international auxiliary language.
My point here is simply that Esperanto was designed from the ground up for a specific purpose, and at the time of its invention it was better suited toward that purpose than any other natural language. Again, I think there's room to debate the specifics, but I'm simply saying that designing something from the ground up doesn't mean it'll be adopted readily even if it is better by some standards.
Those are actually good examples of the type of answers I'm looking for! I'm not really interested in what could be feasibly implemented and adopted but more about hypothetical solutions. Dvorak being a better keyboard layout than Qwerty is exactly the type of "start fresh" solution I'm interested in hearing about.
If you want something really off the wall: race. We all know that race is a social construct. At the same time, race and identity are intertwined such that trying to deconstruct it may invalidate the lived experiences of many oppressed peoples. I strongly believe that race was invented to justify the inhuman treatment of certain peoples.
What if we had the chance to do it all over?
The research on the Dvorak layout seems somewhat inconclusive? Anecdotally I used it for 2 years and I never improved on my QWERTY speed. I didn't really notice any ergonomic benefits either.
I've used it for 3ish years. I still can type qwerty fluently and, to me, the ergonomic difference is night and day. I have carpal tunnel, so maybe my case is special. Like you, my typing speed has remained constant (though slightly higher in Dvorak).
I thought there were objective measures like finger travel that show how much less effort it takes to type on Dvorak or Colemak?
I, too, had horrible repetitive strain from using QWERTY and switched to Dvorak 30 years ago. I've been using it since and my pain has never returned. I am a much faster typist in Dvorak than I was in QWERTY, but I forced myself to learn to touch-type Dvorak. I could only kind-of touch type in QWERTY. (You pretty much have to touch type if you use anything other than the standard layout your keyboard comes with because most keyboards don't have removable keycaps. I only recently (in the last 3 years) got an old IBM Model M with removable keycaps, but since I already touch type in Dvorak, I haven't bothered to switch them.)
I believe you are right. I was mostly talking about it in terms of speed. Maybe I just don't type enough for Dvorak to make much difference.
When I first switched to Dvorak, I consciously noticed how much more comfortable my hands felt after a long typing session compared to QWERTY.
Education in Brazil.
If we're magically rebuilding things from the ground up, lets swing for the moon. I'd love to see teaching the most venerated of all professions. Teachers have the highest pay, choosing to become one is a highly respected and honored choice, and the position of Head Teacher is on par with being chosen as Pope.
Yeah, that might be problematic in itself :P
Just look at how medical doctors are treated in Brazil. Bad things happen when a class of people believe they’re gods...
What bad things are happening with medical doctors?
Medical doctors in Brazil are treated like gods and sometimes believe they are. They make lots of money, way more than the rest of the population, but never cease to complain and strike. A previous government brought Cuban doctors to try to solve the scarcity of doctors in some regions, and local doctors violently protested because fuck the population. Brazilian medical doctors are spoiled, extremely well paid and frequently rich. I don’t have much sympathy for them.
Very interesting. Thanks for sharing!
Care to elaborate? Politics and rhetoric should be pretty uncontroversial and logic seems fair (although I'm not exactly sure what that implies, and I definitely don't know what Aristotle has to do with it) but... poetics? What's that for?
The Poetics, by Aristotle, is a study of the tragedy art form—how each of its elements must be composed and organized in order to elicit the maximum cathartical effect in the audience. This is the most influential treatise on how to tell a story in the history of humankind. To understand the Poetics is to understand the very foundation of western fiction in all media. Education is not only concerned with the real world, but also with the narrative arts. Doing so without at least a general introduction to Aristotle’s take on the matter is, in my view, absurd.
It is also quite readable, as far as Ancient Greek philosophers goes.
I think a lot of software issues stem from 80s/90s standards that are nothing more than a burden in modern use. Most importantly:
Office software: Microsoft is horrible at UX and the amount of frustration I feel from using Word and Excel is only dwarfed by the realization that these issues are multiplied by the billions since this is the industry standard for office work. I'm seeing people around me bang their head on their keyboards trying to get some simple formatting to work in Word or getting Excel not to replace "Mo" with "Monday"... and they all blame themselves instead of that utterly broken piece of software! There's plenty of good a single, dedicated UI expert could do by just revamping the user interface standards of office programs but nobody seems to give power to people who understand user interaction. Also, how do you change a system that literally billions of people work with every day? We're stuck with digitally replicating a type writer on billions of computer screens because that seemed like a great idea in the 80s.
The HTML/CSS/JS web-stack: The more I think of it, there more I'm convinced the internet would be better if it dropped HTML, CSS and JS. The idea is interoperability and backwards compatibility but that rings cynical when you look at major websites barely even supporting browsers older than 8 years or so. The web is constantly changing. Why don't the tools? I understand that, in reality, this is a politics problem but it's still frustrating. In an ideal world, major computer companies should be able to come to an agreement of how to arrange a bunch of boxes with text in them without taming a few thousand competing and redundant CSS specifications. In reality, it would probably be Google dictating a standard only they can follow, with ads built into it as a dedicated tag. But this is about dreaming big, right?
The commitment to legacy support on the web isn't about old browsers supporting new content... it's about new browsers supporting old content. It's not political. The point is to avoid sites needing to be periodically updated to remain compatible with current browsers. If your website is functional now, you can rest assured it will be functional tomorrow, a year from now, even ten years from now.
Think about the mess if that weren't true... users would have to keep separate versions of browsers to use sites that are or aren't up-to-date on the specs. Web developers would be forced to constantly maintain and regularly migrate codebases, lest they get bitten by compatibility issues they have no control over. Or they'd have to (shudder) maintain multiple versions, in which case, you'd probably still end up with tools like Babel to deal with those problems.
But isn't that already the case? Mostly it's about old browsers supporting new content, but even the reverse can be a pain. I'm regularly baffled by the amount of old websites simply no longer working on modern browsers. There's some Flash widget here, some obscure styling hack there, broken links all around. If it weren't for the internet archive, few if any content from 20+ years ago would still be available. Most websites 404 after about 5 years of drifting into obscurity. Websites that have been around for long enough tend to have archived their articles or changed the routing and layout. It's insanely messy. Also I don't remember the last time I genuinely needed web content from more than 20 years ago. I mean, that Space Jam website is cute but... that's about it.
Now I'm aware of the value of a common standard, as brittle as it is, for internet communication. I don't expect HTML, CSS or JS to go anywhere within my lifetime and probably that's ultimately a good thing. But it's a horrible mess of a standard to build on and it's getting messier every year. In an ideal world, I'd happily give some backwards compatibility for a saner alternative. I'd jump on that in a heartbeat.
Now that IE is finally starting to actually fade away and modern browsers have adopted an automatic update strategy, legacy browser support is becoming much less of an issue than in the past. Most browser compatibility issues now come down to cutting-edge features whose standards haven't been officially finalized yet. There are an endless supply of developer tools available to make dealing with those problems effortless, but in any case, you'll never be forced into a position where you have to use those features.
Flash, and native web plugins in general, were proprietary, and neither W3C or ECMA were the ones who decided to kill them. That said, avoiding the sort of problems with Flash support disappearing is exactly the point of web's legacy policy. The Internet Archive literally wouldn't work if old parts of the standards were removed.
Also, obviously, the web standards can't guarantee that all content will be hosted indefinitely, only that it will work if it is. It's not exactly surprising you don't often run into content from 20 years ago because the web was significantly smaller than it is now. Wait another 20 years and see if you feel the same. It's also not just a matter of public websites... there are a lot more legacy web applications that businesses and government still rely on than you'd think. As we see with the IE situation, you could practically point a gun at the heads of people in charge of the decision to upgrade those, and they'd point to browser usage stats and refuse.
I get people's frustrations with certain aspects of the web, but I just totally disagree that the standards are a mess. Are they perfect? Absolutely not, but the notion that the standards committees are "insane" or don't know what they're doing, or that there is a simple and/or obvious solution to its problems is just absolutely ridiculous. Name a platform as big, as popular, as flexible, with as wide of breadth, as the web that is simple and clean and easy. There is none, and and there never will be. There will always be drawbacks no matter what approach you take -- you're always gonna piss some people off and make others happy.
The US government is a decently easy (if a bit passive-aggressive) example. See here for some improvements, particularly items 1 and 8. A few other important additions would be (renewable until they first get a deadly, chronic or untreated disease, for example) term limits for judges, proper legislation to make sure there are enough ballot counting stations for physical lines to not be a pain to wait in such as requiring a maximum distance to a polling station and a maximum number of voting age people per polling station (i.e 500 people per station, meaning the US would need to have ~-500k polling stations) and enforcing that.
Edit: and really reworking lame duck laws or nuking them entirely. Don't let people change laws after or weeks they're voted out of office?
Another, far less dragged out one one would be the human brain/behavior. I've talked about this here and here (second paragraph.)
For improvements, a few are already mentioned in the second comment linked, and those are:
A deeper, more fundamental desire for learning how politics works, same for science (-making), business and general economics. Imagine if one of the most important things for someone's that everyone wanted to know what their system of government or economics and following politics, legislation and business are and how these things are changing closely was something everyone did and wanted to do. It would be imbued in everyone's brain from birth that they're in a wider system and that they need to pay attention to it, take care of it or change it and that few things are actually mostly individual matters or at least that politics overwhelmingly aren't.
I'd also try to eliminate biases based on looks or immutable factors like race (skin color) or ethnicity, with probably an exception to language since learning languages is too hard to be done leisurely.
(Language also sounds like something with lots of technical debt. Wipe it all clean and replace it with lojban/toki pona?)
I'd also want to 'fix' a bunch of cognitive biases that have worn our their welcome like the dunning-kruger effect, creeping normality, confirmation bias for example and replace them with probably some sort of empiricism (although obviously scientists aren't gods either, they need to be checked and given a good work environment.)
To riff off of this one, I'd love to see a societal change where as a part of every child's education, they travel to several different countries (or at least cultures) for exchange-student style cultural immersion for several months. I like to imagine that instead of summer break every year, everyone just trades kids around so that they can experience they might not ever experience otherwise.
That's going to be difficult when the kids you probably most need to target have parents that aren't going to let their kids go to a school function on the other side of town.
True, but in this magical wish-land, this is already a cultural norm and a tradition that has been in place for a long long time.
Kind of a small, petty thing, but the Markdown spec's always had small idiosyncracies that bothered me. The semantic-style italics, bolds, and strikethroughs, auto-numbered lists, indentation creating code blocks, lack of spec around anchors, and lack of formal tables support (although, informally, the Tildes style is widely used) are all minor annoyances that I run into from time to time.
Although the first two can be justified by Markdown being designed for the semantic web, I still don't like it.
*asterisks*surrounding a word make it look bold, and_underscores_surrounding one makes it look slanted to me. Maybe it's early-exposure bias - I have a strong childhood memory of using underscores for italics and asterisks for emphasized words. Indentation creating code blocks has always boggled my mind - I can't think of a reason why it's necessary, and it's only ever served to mess up other formatting. Code fences, on the other hand, are fantastic, especially when used with info strings.And if I'm extending the base Markdown spec, I'd add in official support for LaTeX between
$and$$delimiters. The dollar sign is used too much for that to happen now, but hey, I do sometimes wonder if a website like Reddit had adapted that syntax early on, if math blocks would become more widely-used.I agree. It is not easy to create a format that is both understandable, but also doesn't unexpectedly fuck up user's text.
Markdown does only a mediocre job of it. That is to say, it is not terrible, but I bet I've seen it fuck up a user's text unexpectedly just as often as I've seen it help a user format their text.
I mean, Markdown doesn’t work because it’s logical, it works because it’s popular. A lack of logical consistency is not really an issue.
Don't get me wrong - I love Markdown. It's one of my favorite things to come out of the Internet, possibly ever. These nitpicks barely register as complaints in face of the bastion of usability, readability, and cross-site compatibility that is the CommonMark spec.
(Markdown's pretty awesome)
I would probably go with how we travel. I’d look into ways of improving and incentivising public transport over private travel. Private car travel is pretty wasteful, but for long journeys - especially in the UK - there’s nothing quite like it. The trains are terrible, taxis are expensive and pretty much have the same issue, and ride sharing firms exploit their “employees” greatly. But I wouldn’t know where to start, and especially while preserving/ensuring nature thrives.
This seems like a great place to plug the Tau Manifesto.
This may sound weird, but I'd redo how (US) society deals with love. The way we express it; the way we teach our children about it; the way we think about it; everything. Here in the US, our popular culture is all pink hearts and boxes of chocolates, and of course buying things. It's easy to say, "Well that's just a caricature the media uses as a shortcut." Unfortunately, though, it gets drilled into you to the point that you find other people actually believe it despite what they say.
I'm not just talking about romantic love, either. I also mean the BS we teach our kids about familial love, too. "Your family will always love you." No. No they won't. They are humans with all the shitty problems that humans have and they very well may not love you. Or they may love you, but love themselves far far more. You shouldn't do stuff for them just "because they're family."
LaTeX is unmatched for its precision and extensibility, but requires hacks upon hacks to get to a comfortable position in it. Competing compilers, built on top of a language that hasn't been written in years, lack of unicode support, obtuse syntax, a nearly 2GB "minimal" install size on macOS... its a mess.
Many things but likely there'd be little userbase since people are still hooked on legacy code with little incentive to move on. Even in that case many of those same idiosyncrasies would still exist due to human nature. Namely ideologies that contain hate for others, or conversely praise for others, based on little to no substance or incorrect antiquated ideas.
Two things come to mind - scientific publishing, and modern economics.
Scientific publishing has far too much gatekeeping, lack of quality control, and there's even blatantly unscientific thinking present now. We've had a few meta-studies posted on Tildes that go into just how bad it's become. Today this publishing is more about signaling support for the existing paradigms than it is about pushing knowledge forward. Only the 'anointed' scientists are allowed to be taken seriously, heretics are figuratively burned at the stake if they challenge that core group. This is a consequence of the way publication works within academia. I think this will sort itself over time as more and more scientists publish their data outside of the journals that are stewarding this mess. We're still spinning through Khun's model and we'll get there eventually.
As for economics, what's needed there is replacing the 'greed is good' mantra with a better set of goals and measurements. Profit is a fine goal, but profit has to take a back seat to sustainability and I'd argue to human dignity. We can't continue to allow companies to externalize the costs of their bad behaviors - those costs need to hit them instantly, not be deferred by lawsuits or until they become someone else's (usually government) problem. I'm partial to making the board and even the shareholders personally, criminally liable for a company's misdeeds. If they can't keep ducking responsibility behind the fictions of corporate personhood and the structures they've built to shield themselves from these consequences they have no choice but to make sure the company stays on the level.
It seems like you can pick damn near any institution today and it's in a catastrophic failure mode of one stripe or another. We need to get better at regenerating our institutions. All of them always age out and fail, that's just the nature of orgs over a long time period. Somewhere near the end of that failure cycle is a place for renewal, we need to stop fighting against those changes and somehow make them part of the org's natural life cycle. Burning them down without having replacements ready doesn't sit well with me.
So, it takes a bit of hubris to declare that the way things are done as we know them—the result of constant evolution and disruption (and interruption, from bureaucracy and ~capitalism~)—are wrong. Hmm...with regards to technical things, at least, I'm always surprised by how well thought-out things were, early on. In early-days, it seems academics really tried to think through standards & protocols and tried to get them right from the get-go. I doubt I could think through such things with such depth of knowledge, so I highly doubt I could make anything better. It's just that most things get bastardized by greedy hands, is all. A rethinking of the system of capital would be the biggest change, I suppose. Then again... would we have such advancements if it weren't for the selfish goal of wealth accumulation? In that case, some form of indoctrination in the tenets of living together and contributing for the sake of everyone's well-being might have to be put in place...
ANYHOW, I think the current educational system (at least in the U.S.of.A.) could use a shake-up. Critical-thinking ought to be the main focus, with certain tools introduced (and certain tools to be learned via induction) to help in the process. First principles, first. Learn how to derive everything. This seems particularly touted towards math and science, but it doesn't have to be. The most history I learned came from an art history class I took in high-school, which started with some of the earliest art (e.g. the Venus of Willendorf) and extrapolated culture from that point. Writing ought to be emphasized more. Also, kids like to play. Let them play. How to we get kids to revere school, rather than abhor it?
America. Looking at the debates, it’s a dire warning that the wires have become a twisted nest and we need to do some serious management on those suckers. I don’t know exactly what, but we at least need a big bag of zip ties.