Hacker News new | past | comments | ask | show | jobs | submit login
The most impressive linguistic feat I’ve ever seen (twitter.com/mattgemmell)
415 points by epaga 32 days ago | hide | past | favorite | 167 comments

I won't claim it's more or less impressive, but

Luis Antin van Rooten's Mots D'Heures is at first glance abstract (bordering on the nonsensical) French poetry ..

But is actually a homophonic translation of English nursery rhymes - that is, if you read the French aloud, it sounds as if you're reading the rhyme.


    Un petit d'un petit
    S'étonne aux Halles
    Un petit d'un petit
    Ah! degrés te fallent
    Indolent qui ne sort cesse
    Indolent qui ne se mène
    Qu'importe un petit d'un petit
    Tout Gai de Reguennes.
(If you read it aloud you will recognize a well-known scrambled egg...)

The effort is in making the French poetry make any sense, which van Rooten (a Hollywood character actor in his spare time) does admirably.


To hear it read aloud:


Using the desktop browser version of Google Translate, click on the listen/speaker icon below the French text in the left side panel

I had to listen to it three times before I was able to make heads or tails out of it. But it was worth it.

Amazing! Also works on mobile on the actual website.

I'm thinking this must be something that has been manually tweaked by someone on the Google translate team. Otherwise, how in the world does it pronounce it with the cadence of the Humpty Dumpty nursery rhyme?

The commas give it the right rhythm I believe. I can only hear French though, maybe because I never heard the English version.

Yes, I added commas at the ends of a few lines, which cues the speech engine to create more appropriate intonation and natural pauses. But parts of the second half of the poem are still not so intelligible in English

Thanks for that. It worked. I cut/paste the words before I saw your post. It gets so jammed together that way that it is hard to understand. The commas made it easy to understand.

Because poetry is the mastery of meter, and meter is a thing

This reminds me of a similar homophonic rendition, in which the French actually makes some sense (https://www.blueridgejournal.com/poems/mots06-etquirit.htm) - I actually first saw this in the Whole Earth Catalog, of all places.

   Et qui rit des curés d'Oc?
   De Meuse raines, houp! de cloques.
   De quelles loques ce turque coin.
   Et ne d'ânes ni rennes,
   Écuries des curés d'Oc.

That's pretty good. The translation notes are completely bonkers.

Apparently my knowledge of French pronunciation is nowhere near sufficient to figure this out. Based on the Wikipedia article, it seems the first line is supposed to sound like “Humpty Dumpty.”

If you read the English line first with a ridiculous French accent (like you might hear in Monty Python) and then read the French in an equally silly way, with a bit of imagination they sound broadly quite similar. Basically if you can make "Mots d'Heures" into "mother", you're well on your way!

You didn't fail, it's just that this supposed translation makes a lot of assumptions about how it will be read.

I speak French, and did not immediately recognize "Un petit d'un petit" as "Humpty Dumpty"; the "correct" reading relies on knowing that a native French-speaker might pronounce "hump" as "ump". It is not simply a mapping of syllables to phonemes.

In other words, it's an extremely targeted joke, intended at people who have specific expectations about how both French and English are read and spoken.

I know another French joke that is extremely targeted in exactly the same way. It's from Canada, and it pretty much only works when read aloud.

    Eh, did you notice there's no h in 'ere?
    It's 'cos they left it in h'Ottawa.
For people who don't know the accent, French people don't typically pronounce the h initial in words like 'here', (or 'humpty' for that matter) and they add an h to vowel initial words like 'Ottawa'.

Another joke we have refers to how the French pronounce the English word "happiness."

You need to imagine a French person speaking English with strong accent (e.g. Michelle in Gilmore Girls) to get “Humpty Dumpty” (which is obvious I guess, since the poem is in French, I suppose you couldn’t make it sound like an English or American accent)

At best it sounds like hunpety hunpety. It's said to be homophonic, but at best there's a resemblance. I suppose it matters how you're pronouncing the French -- in a modern pronunciation there's really no resemblance.

There was an old flash animation that did a similar thing with a Japanese rap using English words


Ah yes, animutation. This was peak meme culture back in the early aughts. Mmm, tasty nostalgia.

What's interesting about internet meme culture (back then and to this day) is that it's quite Dadaist. https://en.m.wikipedia.org/wiki/Animutation

Oh gosh, there's a all wiki full of that https://wiki.animutationportal.com/Super_Smash_Bros._Animuta...

Wow!! I had to go to the Wikipedia page to fully appreciate it. Thank you for sharing that.

I think this is the most made-up, arrogant comments I've seen on a post that haven't even taken a glance at the actual content. People all over asserting that it is has been done in special software, that there's typos everywhere, that there's double spaces all over the places, when none of it is true. I have to say I'm quite surprised at the number of totally certain and easily disprovable assumptions people on here and on the twitter thread are making.

Welcome to the internet ¯\_(ツ)_/¯

It's not hard to find highly opinionated people making assumptions while confidently believing they're right. If you look closely, you'll notice this all the time on HN and Reddit and basically anywhere that allows people to comment.

The scary thing is that many readers will also often take anything said confidently enough to be fact. This is also basically how a lot of disinformation works. Say it authoritatively, and you'll become an authority on the matter to many people.

I've definitely been guilty of doing both of these things past, but I've learned now to actually read and think and do research before commenting or taking something I read at face-value.

>It's not hard to find highly opinionated people making assumptions while confidently believing they're right. If you look closely, you'll notice this all the time on HN and Reddit and basically anywhere that allows people to comment.

C'est la vie. The internet allows millions of people to opine on barstools.

It might simply be that our standards for the written word were too high before as it required a large investment to write/print/distribute a magazine or book.

> C'est la vie. The internet allows millions of people to opine on barstools.

It may be interesting if there was a social media platform that exerted some sophisticated control over the manner in which this phenomenon plays out. If adequately successful, the way "la vie" is, would then be different.

  Do you take a possibility
  you can be both right and
  wrong? The feat might not
  be as difficult as author
  seems it is. I personally
  can think of many ways to
  achieve it. Come to think
  of it, it may actually be
  even easier because lines
  in that guide afford more
  ways to be broken because
  they are longer. And I am
  not even native speaker.

  To match lengths you only
  need to make one or maybe
  two choices per line only
  to adjust a word by maybe
  one or two chars. This is
  nothing very difficult as
  any crossword solver will

"take a possibility", "as author seems it is", "not even native speaker".

It's simple to fulfil the requirements, making it flow naturally is not. Also, any mistake early on in any paragraph will have cascading effects.

The parent makes a good point. Maybe I am jaded, but this isn't a scarily impressive feat of linguistics. It is mildly interesting that someone put the time into this. I would be impressed if it were done on a typewriter. Maybe this rewrite helps clarify the point.

  Do you think its possible
  you can be both right and
  wrong? The feat might not
  be as crazy as the author
  claims to think. I myself
  can think of many ways to
  achieve it. Come to think
  of it, it may actually be
  even easier because lines
  in that guide afford more
  ways to be broken because
  they are longer. I am not
  even a native speaker.

Brick-texting isn't actually that hard, in short isolated bursts. Other geometric texts are not necessarily much harder, But, keeping it up over any length is somewhat impressive.

I think the most "simultaneously observed forms" were brick-text with an acrosticon down the left-most column, or brick-text in iambic pentameter. I have attempted observing all tree, if ended badly.

  As somebody said, I am just a
  random guy on the Internet. I
  learned this language already
  as an adult never attending a
  single lesson. There is a lot
  of other people that know the
  language way better than me.

  Thinking that it must be hard
  just because it is for you is
  a well known bias.

Gigablah, I can't reply to your comment because apparently I am rate limited.

I am not saying it is easy. I am just saying it is not "most scarily/stunningly impressive linguistic feat".

I hope you can appreciate there is a lot of space between "easy" and "stunningly impressive".

  It is easy to do this. All you
  do is write some text upfront,
  and go line by line and swap a
  few words to make the line fit
  in the margin. Use small words
  to make it easier.

Conversely, thinking that it must be easy just because it is for you is an even more well-known bias.

I can even think of an entire industry dedicated to countering this bias.

I agree, it definitely takes a bit of time but to compare it to Mozart is ludicrous

I suppose one only has to adjust the line length by at most half a word per line. It's impressive, but it's not exactly iambic pentameter.

Also the author gives himself the liberty to use one or two spaces following a period, and on occasion even following a comma.

I disagree that this is important. Search for a double space, and you will see only four that are inside a paragraph, and only one of those is after a comma. There are more stray spaces after the end of a line than there are double spaces inside of a paragraph.

Well, he did compare to Mozart. To pick a rando Mozart piece[1]:

1. Sounds like cutesy little throw-away dance with nowhere near the gravitas of a Beethoven or Mahler

2. Listen closely and hear that the phrases are written purposefully so that the music sounds bad when the musicians choose to obey the repeat signs. In other words-- the music flows beautifully if you ignore the repeat signs, but that flow is broken when you obey them and go back to the beginning of the phrase.

3. Musical jokes later in the piece depend on the players actually taking the "wrong warp" repeats

4. Listen closely and realize that Mozart actually left out the main melody of the piece. (In fact, that melody finally appears only once at the end of the piece)

5. Listen more closely and realize that one of the subsequent main melodies are based off the accompaniment figure for the missing main melody

6. Listen even more closely and hear the striking dissonances and imitative polyphony that is exactly the kind of thing admirers talk about wrt Beethoven, Mahler, etc.

7. Remember that it's difficult enough to write a cutesy little throw-away dance in the style of Mozart, much less pull off all these formal experiments with such subtlety that most modern performances continue to play this music as if it's just a cutesy little throw-away dance

And that's just a random example. For a specific example: if Beethoven had written the Introitus of the Requiem Mass instead of Mozart, it would count as the most sophisticated counterpoint that Beethoven had ever written. (And there are probably more impressive examples of Mozart's counterpoint than that.)

1: last movement of String Quartet in E-Flat Major, K. 428

Whether or not it's actually impressive is irrelevant, the point was that more people than usual were just making random things up about it.

As someone who loves Beethoven but could never get into Mozart, I would appreciate a link or dozen to actual, good performances of the pieces you mention as needing special attention. Thanks.

In its particular field (the brute force justification of monospaced text by creative word choices), it is just as superior as Mozart's work in his field.

But AFAICT this author's prowess in the field you defined is no more or less than a mastery of that singular skill.

For the author to be a "Mozart" (as in the historical figure, not the myth[1]) you'd need to show a richness of the author's approaches that makes the reader aware of the challenges of and breadth of the field. Something like this:

* the brute force justification of monospaced text by creative word choices

Oops, wait...

* the brute force justification of monospaced text by creative word choices which follow a rhyming scheme

Oops, wait...

* the brute force justification of monospaced text by creative word choices which follow a (line-based and paragraph-based) rhyming scheme

Oops, wait...

* the brute force justification of monospaced text by creative word choices which follow an intricate rhyming scheme where certain paragraphs are mashups of archaic German verse and 1970s gospel lyrics

Oops, wait... there are also some patterns of words where the pattern is written twice, but punctuation was inserted in a seemingly arbitrary place so that I didn't even notice the repetition on the first reading. Something like, "One two three four one. Two three four." Except each sentence was grammatically and contextually correct. Mozart did stuff like that-- there's an example somewhere in the clarinet concerto but I can't remember where.

1: I mean, there's no law against using the mythical Mozart here. But I imagine we'd be trolled by HN'ers if we did that regularly with, say, von Neumann or Einstein.

I agree that there's not a single instance of double-spaces in the text, and I thought that was a really bizarre claim, but there are some instances of misspellings that help it along.

For example:

1. missles instead of missiles

2. carefull instead of careful

3. futher instead of further

But they're at least all consistent so there's plausible deniability that the author just didn't know English that well. And ofc the tweets are still on-point saying that it's an impressive and unappreciated feat.

That makes it sound like the typo is made in just a few places to make the line fit.

Search missile, there are 0 results. Search missle and there are 217.

The consistent and plentiful usage makes me strongly believe it’s just how the author thought the word is spelt, not a cheap way to get their preferred formatting.

> they're at least all consistent so there's plausible deniability that the author just didn't know English that well

Sure but my disagreement was with

> help it along

Does it actually help them along if it’s that many items? It’s hard for me to imagine that 217 instances of a 6 letter word instead of a 7 letter word is much easier to format. My view is that trial and error, obsessively rewriting sentences to fit, is much more likely.

EDIT to clarify: what I mean is that I don’t think there was a master plan where words were misspelt to make this work. I think it’s rather brute forces with a lot of obsessive time and effort. Not in a bad way of course, I really enjoy what they did.

EDIT2: especially when rocket is already a 6 letter word, and likely could be used interchangeably in the guide. I feel that turning missile into one as well wouldn’t help the goal in the slightest, would only make it more difficult.

They're not just vowel removal abbreviations or similar though, they seem like things that'd be common errors - 'missles' is how a lot of Americans pronounce 'missiles', if her film and television is any indication, for example. 'Careful' is one of those 'eh what can I say, English spelling is weird' words anyway - 'care', 'full', why shouldn't it be 'carefull'? I'm not a teacher, but I bet that's a common error.

People were saying it was littered with double spaces and typos, rather than just a few instances, which is just incorrect

One of the later tweets says it's all "bricktext", which apparently is a thing. An old internet thing even. I may have to try it some time.

Does nobody remember the Jargon File?


Indeed, this is why we cant have nice things.

Welcome to the internet.

OMG I didn't realize other people had this affliction (I consider it a horrible OCD-like issue as I find myself often unable to send messages in editors that start to give me control over "squareness"): all of my tweets are the maximum length, all of my git commit messages are 50 characters, I gave a text-heavy talk a few years ago in a monospace font so every slide had a ton of bullet points that were all the same number of characters, and I often write messages where every line is the same length (I have linked a couple examples that are easy for me to link to off the top of my head).



I have never done this to anything as long as that manual (partly as I simply both don't let myself type things that are that long in a monospace presentation font and also try very hard to resist the temptation to ever, under any circumstance, manually word wrap text).

I see you haven't yet caught the coding version of that.

E.g. http://people.idsia.ch/~juergen/oopscode.c

(From http://people.idsia.ch/~juergen/oops.html)

Impressive, but very difficult to read in some places. I am definitely a fan of aligning similar lines, from equal signs to columns of numbers and variables; in the right dosage I find it helpful for readability, but your linked example also shows the problems when you overdo it.

I think I should be glad that I rarely write long texts in monospaced fonts. I can't unsee that perfectly aligned game guide.

Why would you do this to him!

You're definitely not alone. In the following prose, I tried to minimize the number of non-natural hyphens, reduce excessive spaces between words, and ensure that each section fills all three columns entirely:


This was a fad in my local BBS around 1995. Not to these extremes, of course, but for a couple of months that year we'd justify our forum messages to the 72nd column without any extra space.

It's not too difficult, lines have enough words that there's always something you can change. After a while, as you approach the right margin, you can feel whether the rest of the sentence will fit, and you rephrase it as you write.

I see OP more as a feat of tenacity than of brilliance.


I think the interesting lesson here is that if you find yourself in the position of the tweet's author you should be thinking "Hm, maybe writing blocktext is easier than I'd expect" rather than "I have discovered an unsung linguistic genius".

The skills required are similar to those for writing in metre: have a decent vocabulary and a good feel for the different ways to structure an idea, plus a small bag of tricks for adding or removing a syllable / character.

I love finding things like this on my own, without being told about them. I think that is what the tweet thread is about—just the author sharing their joy about that.

It also might be that the tweet author is exaggerating for comedic value. That’s a thing people do.

  > have a decent vocabulary and a good feel
  > for the different ways to structure an idea
This sign of intelligence is also good for talking to people on their level: Explaining things to children and adults alike while meeting the expectations and comfort zone of both. I've always admired people who can articulate their ideas in a broad range of spoken formats for the intended audiences.

Interestingly, some of the best-known examples of people like this are all physicists: Einstein, and Hawking, and Feynman especially.

I did the same thing on Usenet around the same time period. I was surprised at how easy it was, actually.

Same here - I did this for a couple of months on Usenet and it was more a fun challenge than genuinely difficult. Doing it over that amount of text is quite impressive, admittedly.

This doesn't seem all that far from the art of writing tweets which are exactly 280 characters. Start by writing what you mean straightforwardly, then go back and edit it down. First restructure sentences, then replace words, then elide grammatical words, then juggle punctuation.

Now try doing thousands of them consecutively.

Extremely tedious (IMO), but totally possible if I for some reason wanted to. I've often had a long tweet (it's often an effective customer support avenue, but the character limit isn't conducive to conveying detail..) but as GP says, it's just a case of writing what you want and then whittling it down with a bit of rewording/rephrasing/abbreviation.

Thousands consecutively would 'just' take thousands of times longer, it'd be O(n), I just don't have any desire to do that, it's not fun for me.

Namely, that the following text employs word choices such that it forms justified blocks without having either added extra spaces or hyphenated words across lines:


Thanks. The title and the whole tweet-train is a pure click-bait.

It does at least give you a chance to see whether you can spot it for yourself.

So two tweets would be sufficient I guess?

Hahah, yes, I guess so.

Seriously. This is almost a genre at this point. Hysterically demanding rapt attention just to take an entire paragraph to say something that takes one sentence. It drives me crazy. I put it in the same bin as the scourge of 8-minute how-to videos that could be three bullet points and a diagram.

"Whoa, this is nuts: This Super Metroid walkthrough has monospace text that is manually justified through WORD CHOICE ALONE. [link]"

There. Was that so hard? And they still get to yell.

This linguistic feat also applies to code.

For my 2012 IOCCC entry [1], I spent many many hours tweaking the syntax and ordering of code to get the uniform looking layout. Which I ended up doing many times over every time I found a way to shorten the code. One can play with the total height, width of the arms, and height of the arms, to try to get back to uniformity.

I expect many other IOCCC authors have had similar experiences.

[1] https://www.ioccc.org/2012/tromp/tromp.c

I looked closer into this IOCCC entry, and must say I am very deeply impressed. For those wondering, I would recommend the accompanying article [1] as a starting point.

[1] https://www.ioccc.org/2012/tromp/hint.html

This is truly impressive.

The most impressive linguistic feat I've ever seen was on a Mastodon, which was named for the French writing group "Oulipo", whose members did work with constrained language. The work from this group you've most likely heard of (if you've heard of them at all) is George Perec's book "La Disparation", translated to English as "A Void", which does not (in either language) contain any word with the letter "E" in either case in it (so those of us who know at least a little French realize that means no "le" or "les"... wow...).

Anyway, this Mastodon was set up to implement the "no E" rule, and it was kind of a fun challenge to try to express yourself.

My posts (and most folks') was pretty tortured text.

The impressive bit was that there was a person there pursuing their Doctorate in Linguistics, and their posts... just read like ordinary English, with nothing tortured or strange about it. If you didn't know about the rule, you'd never guess it was in force. And they were fairly long, and covered a lot of ground... really amazing.

FWIW, the Laravel source code has a prominent pattern to it, (comments, anyways) which is nice, for example

    | Application Timezone
    | Here you may specify the default timezone for your application, which
    | will be used by the PHP date and date-time functions. We have gone
    | ahead and set this to a sensible default for you out of the box.
[1] https://github.com/laravel/laravel/blob/8.x/config/app.php

Wait until he discovers Shakespeare, where every line is written in Iambic pentameter.

Or Dante's "Commedia" where each line contains 11 syllables

in triplets, that rhyme in chain

for 100 chapters,

for 4711 triplets

I believe that was actually the style of his Belgian cousin, Schuddespeer.

Or Georges Perec.

Reminds me of Gadsby, where the author wrote an entire novel without the letter e. Even more impressive is others have translated it into other languages, still without the e.


Likewas, thirty years later George Perec wrote La disparition[1] which is also devoid of the letter "e", and was also translated to multiple languages with similar constraints (with some variants, the Spanish version for instance is written without "a").

[1]: https://en.m.wikipedia.org/wiki/A_Void

In the Perec instance, it's what was called by the Oulipo group (of which he was a part), a Lipogram. Perec wrote a follow-up text in which the only vowel used was e. A Canadian author wrote a suite of five novellas each restricting themselves to a single vowel. In his case, he went through a dictionary by hand and created vocabulary lists for each novella.

I've read the English translation of La disparition and I'd note that (a) the act of translating a Lipogram is kind of its own special challenge (which might be also why the Spanish version chose to omit a instead of e—without e, a wide class of plurals and all third-person forms of to be are lost, although a rules out most feminine nouns) and (2) the writer/translator ends up building a vocabulary of standard circumlocution to avoid the forbidden letter, e.g., “this/that man” in place of “he.”

There are also books written in an E-Prime english. In E-Prime you can't use any form of "to be"(is, are etc.).


Though that's less restricted writing for the joy of it and more a philosophical claim that "nothing is, everything does." In other words, the use of "be" to define two things to be equivalent acts as a short-circuit to convince people of a relation without doing the work to demonstrate it.

"Ella Minnow Pea" by Mark Dunn is in a similar vein.


That's called ‘constrained writing’ in general, and there are lots of different approaches and examples of it: https://en.wikipedia.org/wiki/Constrained_writing

The title seems a bit of a clickbait. For those that don’t want to click:

It’s a plain text guide to Super Metroid, manually justified via word choice.

Yeah, there's really no need for a long Twitter thread about some minor detail that can be summed up in one tweet.

It takes the author six tweets to get to the explanation. Like, get on with it, already.

Composing such texts is an art form indeed, and really hard. Amazing achievement.

Once in a while when composing an email, the first three lines or so just happen to have exactly the same length. I always take this as a challenge to continue this style for the remaining lines (no cheating with white spaces) and I have wasted (?) hours on this in the past.

Did this for my wedding invitation.

Kept rewording it and changing punctuation until everything lined up near perfectly. I suspect that there are tons of people who worked in marketing or communications who have done this throughout the years. Especially back in the 80s and earlier, when text-heavy marketing was in fashion.

A similar example: in Laravel's codebase, every line in a doc block is 3 characters shorter than the line above it.

Example: https://github.com/illuminate/database/blob/master/Eloquent/...

I recently discovered Super Metroid on the SNES emulator in Nintendo Switch.

The game is so good! Compared to the other games in the emulator library, it’s like night and day.

I have no idea about other games in 1994, and how does it compare. Doom is one year earlier and it’s much more advanced, so, who knows.

Super Metroid is, improbably, still one of the best "metroidvanias" ever made. The best, in my opinion. If you use the randomizers that people have developed in recent years, it's infinitely replayable.

It helps that it was among the later SNES games, so developers had more time to learn the hardware. Chrono Trigger (1995) is also one of the platform's best, most polished titles.

Not just randomizers, but mashed-up randomizers:


Once upon a time I suffered from severe OCD. I fully justified all of my code comments in C to 80 columns by word choice. No extra spaces. No random hyphenation.

You feel pain while you're doing it, but you can't stop.

I'm glad therapy helped me.

I don't think I have ocd but this is something that I do without thought. Usually, I'll catch myself after a minute or two. But if a paragraph is almost there, it becomes difficult to work with that paragraph on-screen until I scratch the itch.

I genuinely don't think it's that impressive, you just write a line to say something and add, subtract or change some words to fit. Painstaking yes, most impressive linguistic feat? no

It's easier than the author appears to suspect.

I wrote all my status reports for work, commit comments, and quite a few emails in 'monospace full-just' for many years. I'm not sure why I stopped doing it habitually, though I still do it sometimes when I notice my messages naturally came out close to it. It may be just due to using the web more often, which seldom displays text with a monospace font.

I don't recall anyone ever noticing, nor did I expect them to.

Seeing another commenter mention BBSes in the mid-90s, perhaps that is where I picked up the practice.

The metroid text seems to change width periodically in varrious blocks, which is also a property of the monospace-full-just text that I wrote.

  Ok so I get that it could be quiet
  difficult to write sentences which
  must be of a fixed length. However
  hyperbolically calling it the most
  scarily/stunningly linguistic feat
  is ridiculous in the extreme. I am
  certain that a person sufficiently
  motivated, could produce a similar
  document. Comparing something like
  this to the work of Mozart is just
  nonsense. I mean, I couldn't write 
  a piece of music like Mozart if my
  life depended on it.

Yea, we uaed to have to do this as an exercise in my public school English class, is that not a thing anymore?

It's cool, to be sure, and Ive actually used the guide over the years, long before they were sold to Cnet/gamespot.

I really miss the old gamefaqs, some truly amazing finds and collaborations happened there, before it all became a C/P fest with re-hashed and re-mixed same guides over and over again. Or just info/pics pasted from other sites.

The ASCII art that was uaed and consigned back in the day was killer too.

I guess I'm just an Old Randy Marsh now, 'all this new stuff is crap warble warble!'

I don't know. Translating haikus across languages so that they remain haikus of equivalent meaning and beauty seems vastly more impressive.

I remember doing this by accident in an email back in the days in which all interactions with a computer were in monospace (on a 80x24 terminal screen). Fiddling with margins to make it work is kind of cheating (and curiously, it seems to be easier with narrower line lengths than wider ones). Doing it for an extended period of time with consistent margins is a real challenge.

I read the first few pages and no doubt it is an enjoyable read. If I had any interest in the game, I would have read the full thing. But there is no reason why if the same text would not have been any better without graphics and more styling?

Is this confirmation bias? the moment someone makes a claim of excellence, or genius, instantly number of people realize "how it hit them"!!

The guide was last updated in 2001. At that time, GameFAQs only allowed plain text guides to be uploaded. The guide's author did not choose to write manually fully justified plain text instead of rich text with graphics – they chose to write manually fully justified plain text instead of left justified plain text.

It's impressive but in practice, as a guide, it's very difficult to use. There is one path to follow with no overall map or routing to skip things. So to use it to figure out one tricky section often means reading 4 extra paragraphs just to find a landmark and trace from there to where you're stuck before you can follow along.

Ah perfectionism, I do this all the time picking such words to align a word with its opposite word on the previous line (for example, an old name vs a new name to be able to compare them visually when they are aligned). This improves readability. Actually this is not hard to do.

I remember doing this with some homework assignments back in college, the very many filler words available in English make it pretty easy. The linked text uses copious fillers and doesn’t even maintain the same width across paragraphs.

Text blocks like this was a thing in a BBS-style forum I used in the nineties. There were topics where all posts were expected to be in block form. They would also turn up in other topics here and there.

I regularly wordsmith comments in my column-constrained code (e.g. 100 columns) for prettiness—I don’t like to have just one or two words on the last line of a multi-line comment, so I will normally fiddle things to get that last line either empty or a bit fuller. I will just about always do this if it’s a single line plus a word or two. I also definitely enjoy occasionally indulging in the style of perfect line-filling this submission is talking about; but even when I don’t go to the trouble of filling lines, I regularly tweak line break positions to balance line lengths and to start sentences or phrases at new lines.

This style of craftsmanship, embracing constraints and optimising for beauty or even whimsy, is something that is sadly often not possible now with the diverse presentation media we employ: in the past we would adapt the medium and presentation to the content, but we’ve shifted to mostly adapting the content and presentation to the medium.

A paper book had static presentation, and so you could tweak things to your heart’s desire; and if someone had bad eyesight, you could produce a larger version of it while keeping everything else intact, if you desired. One particularly good example of that is static paginations of certain Bible translations; for example, the KJV has a couple of extremely popular paginations, so that one person may have a small Bible with tiny text and another a large Bible with wide margins and large text, but they’ll still turn the page at the same time. (This can lead to church services where most of the audience turns the page during a reading of scripture at the same time!)

But in a digital world, your content needs to adapt to display on devices both tiny and huge, and you will have limited control of its presentation. Static layout like this Super Metroid guide no longer works: at 73 columns wide, it’s too small for display on portrait phone screens without introducing scrolling (awful), additional wrapping (ugly) or reflowing (losing the craftsmanship, and impossible to do fully automatically anyway).

Responsive design is all about embracing the fact that you don’t know the medium, and working with that. And yes, it lets you do some nice previously-impossible stuff and work better on a diverse range of devices, but it also loses something.

I once noticed, part way through writing an essay in school, that I had started the last three lines with the letter w; I ran with it and made every line for the rest of the essay start with the letter w.

I could go on to talk of a great many more things where we’ve lost something over the techniques of old due to low-quality computerisation of something formerly done by experts; but also that better computerised alternatives are often available which claw back some or most of the deficit, typically by more deliberately attempting to imitate what the experts did rather than doing the easiest thing; greedy line-breaking is nigh ubiquitous, but there is also Knuth-Plass; most computerised music engraving is fairly atrocious, but LilyPond exists.

> something that is sadly often not possible now with the diverse presentation media we employ

I miss using <tables> and 1px invisible GIFs to create pixel-perfect web layouts from Photoshop. But there's also something to be said for creating a perfectly re-arranging fluid layout.

Resist the flow. PDF is the answer.

No it’s not. It allows you to get the fixed medium that was lost, yes, but loses the flexibility that is far more important given the reality of the devices people have—and that’s only talking about screen sizes, not even the other problems of accessing and viewing PDFs.

Notice how clear the document is even with plain ascii formatting.

This reminds me of old time military documents - they were peculiarly orthodox and notably very clear.

One of my math professors told us about his friend in graduate school who could justify his papers on the fly. Maybe it's easy for some people.


When someone can write, they just.. can. We mere mortals, only read, in awe..

Can someone please explain to me what is happening here. I'm an old fart. Is this a "blog" where every sentence is a twitter post? Is that to generate more views or something?

Yes, basically, it's written on Twitter because more people will see it there. On Twitter it needs to be broken into multiple posts anyway, but the post is broken into a few more than necessary so it breaks at more natural points and fits the common style on Twitter for post series like these.

I don't care for it myself.

I see, thank you.

It's pretty common on Twitter if you've got more to say than fits in a single tweet. There are even tools to let you manage it (i.e. write a thing and break it up and post individual tweets in order).

Sometimes you'll see them numbered like /1, or if the author knows how many there will be in advance, 1/13, 2/13, etc.

It's not uncommon to put a separate link in many of the individual tweets in the series.


> One who thinks that the word ‘missiles’ is spelled “missles”, but a Mozart nonetheless. Gives me chills every time.

I guess that's one of the trick to shorten a line of text.

This is really impressive, but not really unusual. Poets have written epics under more severe constraints.



Do we have to have this comment on every single twitter post? Jesus, we get it, you don’t like threads, get over it. Either use a workaround so it outputs blog-like post or don’t read it.

But for the love of god, stop complaining about. Every. Single. Post.

Since it's against HN guidelines, I think it's fair to flag such posts at this point.

> Please don't complain about website formatting, back-button breakage, and similar annoyances. They're too common to be interesting.

We should have this comment on every single Twitter post, but it should be mandatory that it also includes the ThreadReader link.


Yes we do have to have this comment until people stop writing things in twitter threads.

> But for the love of god, stop complaining about. Every. Single. Post.

Then stop using Twitter as a way to write longform. :)

Yes, but the OP should have quoted the meat of the article so we don't have to wait for the slowly-loading appearing text to appear on scroll:

>It’s a plain text complete guide to Super Metroid, which is _manually_ fully justified… via word choice.

Yes, they could put that on a blog. It would be cool if visitors on the blog could comment there to interact with the author. Maybe not just at the end but on every paragraph or sentence of their articles, wherever it makes most sense for the comment. Not having to host the blog oneself would take friction out of the process and allow a lot more people to participate, so maybe that's also wanted. While we are at it, an API would be great, since some people might like to automatically collect the content and reformat it to their needs. Somebody make this, please.

It already exists, quite a few in fact, except they are not called "blogging software" and they are not necessarily marketed to "bloggers". There is Slab, Confluence and many others.

I see what you did there. That's Twitter.

I just don't click on them any more. As soon as I see Twitter I come into the thread to see if someone has posted it in another format. If they haven't, then I just won't see it this time. No big deal. There are countless things I won't see today anyway.

> One who thinks that the word ‘missiles’ is spelled “missles”, but a Mozart nonetheless.

Maybe he had to create a typo here or there to keep things fully justified.

> The best, most delicious part is that untold thousands (tens of? hundreds of?) of people will have read it during the past two decades, and NONE OF THEM CAN SEE IT!

Because they don't care? Why should they?

Hey downvoter, would you mind to explain? I genuinely don't understand why I or anyone else should care if some person wasted a lot of time on formatting a text to paragraph boundaries. Is it forbidden to ask such questions here?

I genuinely don't understand why I or anyone else should care if some person wasted a lot of time on formatting a text to paragraph boundaries.

It's nice to take an interest in what other people do, and to show that you're interested, even if it isn't something you'd do yourself. It's the kind thing to do. Kindness is an admirable quality in people. Asking "why should anyone care?" shows a lack of understanding of that.

Please note what the author did say: "thousands (tens of? hundreds of?) of people will have read it during the past two decades, and NONE OF THEM CAN SEE IT!". I'm not a native English speaker, but isn't it sufficiently qualifies as "anyone didn't see"?

That doesn't say that the author thinks everyone should've seen it, or should care. Just that it's a work that almost noone sees while it's just there in plain sight. And the author finds it intriguing knowing about a thing that many have seen, but don't know they have.

This is hacker news. People come here to read about other people doing cool/impressive stuff just because they can. Not because they have to.

Plus, time enjoyed is never time wasted, and clearly the author of the guide had a lot of fun with this.

Well, I have nothing against that. I just wondered why they are surprised that others don't find it so interesting? And bearing in mind that I may miss something I asked for clarifications. Should I be hated for that enough to be kicked out of the community (b/c that's what basically downvotes do)?

I didn’t downvote you, but your comment was probably downvoted because it comes across as rude and dismissive and somewhat like a teenager trying to impress a crowd with how cool he is for not caring.

I'm not a native English speaker, so I'm not qualified enough to counter that, you may know better. But didn't the thing that is was a question not a statement change something in the reasoning?

A downvote is not hate.

you've conveniently missed the main idea: it eventually expels a person out of the community

No, it really doesn’t. If you feel shunned or “expelled” then that is entirely within your own mind. Only you can change your mind.

Personally, I think it's interesting when I see someone has created something which accidentally conforms to a tangential restricted form.

Like, an acrostic, or Bach hiding his initials in the notes. Also, it's not formatting. :)

As for the downvote, I have no idea! It seems this is a divisive topic -- my first comment, above, must have raised many's ire. Also, I think asking why you are downvoted is also verboten. (see: https://news.ycombinator.com/newsguidelines.html) shrug

Michael Hart, of Project Gutenberg fame, was known for this style, only, without typos:


Quite the hyperbole. The author even points out how this was done without realizing it — purposefully misspelling certain words to get the right character counts.

Chapter 1: 32 justified lines, "futher" for "further".

Chapter 2: 18 justified lines. "missles" for "missiles" but is consistent throughout.

Chapter 3: 61 justified lines. "missles" consistent again. "limitid" for "limited" (same length). "powerfull"/"usefull", yes, but again looks to be consistent throughout. "liek" for "like" (same length). "essentail" for "essential" (same length).

Chapter 4: 88 justified lines. "missles" and double "l" consistent. "mroph" for "morph" (same length). "eleminate" for "eliminate" (same length). "insure" for "ensure" (same length). "wreckless" for "reckless" but in a non-justified line.

Every line is justified when you had good reason for writing it.

Wow, I stand corrected.

Exactly. It might be impressive if it's consistent, or if the number of misspellings is rare. But starting off with a correctly spelt document and then achieving justification by conditionally removing a letter is... not that hard.

Ought to be possible to test this hypothesis (is the justification done by conditional misspellings?) with some simple scripting.

edit: however, it appears that's not what the author did.

Just read the first paragraph of story and... I can't see any misspellings. (Pasting here, the spellcheck points out Samus, Aran, Metroid and "molted", which I thought was OK.) (E: ok, "futher" for "further".)

A couple of missed words or odd linguistic choices, but compared to the number of lines, not many.

The one specific misspelling that the tweeter points out is consistent throughout.

I was curious so I did the requisite short bit of scripting.

I think you're right, the misspellings are independent. For a few words (like "begining") the text contains the correct and a misspelt word. But for most misspelt words ("carefull" for example) the text contains only the misspelling.

Thus, the text being fully justified is independent of the misspelling.

I think it could have been written in something like MS-DOS edit, which doesn't have a spell check feature anywhere in sight. So it never saw a spell checker, they didn't have too much awareness of the errors. Maybe they also didn't care about spelling too much - thinking up words that fit in the spaces is fun but correcting spelling errors isn't much fun.

The document being an in depth write up of Super Metroid probably also says something about the person's character!

I find it remarkable and astounding that someone was able to spend multiple tweets explaining what the author “accomplished”.

Really, scarily, stunningly? When words are used this way they lose their meanings.

And the comparison to Mozart? My God. How insulting to the compsoser.

Clearly, this guy hasn’t seen much.

This could have been easily done in software and it probably was

From the guide's FAQ:

  03. What program did you use to justify the text?

      None. I just chose words carefully so that everything lined up on the
      right hand side. Everything was done with an ASCII editor.

Which software?

I don't think it exists. Perhaps we should create it.

I manually justify paragraphs of monospace text the exact same way all the time, for my code and git commits. It's an added cognitive burden that should ideally be automated by M-q in Emacs or something.

With WordNet we can get a good thesaurus for permuting words. Combine that with CMUDict and IsleDict and we can even reshape the text into something like iambic pentameter.

A thesaurus won't work. There are not many true synonyms around; often, synonyms change the meaning of the word, or interact with the meaning of the context. You also need to change inflections, word order, prepositions, and sometimes the sentence and discourse structure to make a good fit, while maintaining the consistency of vocabulary and readability throughout the current context. It's quite a complex task.

Well... it's not that impressive is it, specially if achieved by mispelling. It's laborious sure, but... For instance 16th century poet Luís Vaz de Camões wrote a wonderfully written Homeric poem of +1000 stanzas of 8 lines with the exact same rhyme scheme. That is a much more impressive and unatainable feat by 99.9999% of the population.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact