It's optimized for nothing, as far as I can tell.<p>Efficiency: letter frequency is not taken into account.
Internationalization: no capital letter/lowercase distinction. No final letter forms. No circumflexes, accents, gravures, hats, dots, umlauts... punctuation is not specified.
Physiology: no attempt to take into account how human eyes work.<p>If this were going to work, Braille would be a lot more popular among sighted folk.
Typographically, this has horrible legibility: the letters are difficult to distinguish and recognize on an individual basis. This contributes to poor readability when the letters are grouped to form passages of text. This seems much more about compacting information in a human-decipherable format than creating a more readable alphabet.<p>There are no ascenders or descenders in a traditional sense, no contours or apertures, and there is no stroke contrast. Also, the baseline is thrown off in many words. Put this all together, and you have text that's miserable to read.<p>This isn't even to mention the difficulty of getting readers to abandon tradition.<p>In all, I'd consider this an interesting experiment, but a practical failure.
Is this a piece of satire?<p>I might be able to use it, but I still feel that I would need space between characters otherwise there is a risk of error, without needing to slow down and make sure things are aligned properly. Also the examples make it apparent that the font size would have to be nearly doubled for accuracy purposes.
I can't tell whether this is supposed to be serious. Here's a simple test, show someone the dot => alphabet mapping for A to E. Next, write a dot in the middle of a plain white piece of paper and have your test subject tell you which letter it is.
That makes about no sense (except as really deadpan satire), the strength of the human visual system is in "chunking" shapes, and making letters into clouds of disconnected dots is about the last thing you want to do to it, if you don't have to. If you do have to, there are already well-established systems that do this (Morse code and Braille).<p>Teletypes switched to printing actual letters as soon as they could, even though there were people very highly trained in reading the dots and dashes.
Merely black and white? Sure, writing was developed when there was just black ink and parchment, but now even free (with service plan) handheld devices have rich color displays.<p>Convert each letter into the trinary representation of its 0-based ordinal position – 26 is conveniently close to 3^3, isn't it? Use each digit as a scaled R/G/B value.<p>So for example, 'a' is letter 0, trinary 000, HTML color #000. 'g' is letter 6, trinary 020, HTML color #0F0. And so forth.<p>Every one of Dotsie's dots can now be a whole letter! I can fit the full text of 'Huckleberry Finn', 'Pride and Prejudice', and 'The Adventures of Sherlock Holmes' on my 1920x1200 monitor – and still have room for the 'Kama Sutra'! (Free Gutenberg Project plain-text editions, of course.) Yes, it may require a jeweler's monocle to read, but think of the space optimization.<p>The color-blind can join the blind in relying on braille.
<p><pre><code> " The latin alphabet (abc...) was created thousands of years ago, and is optimized for writing, not reading. "
</code></pre>
... that's demonstrably false. Cuneiform is optimized for writing (with a stylus). Gregg shorthand is optimized for writing with a pencil. The Latin alphabet forms have lots of redundancy (though not quite enough) in 2D visual space in order to be clearly recognized and distinguished. Lower-case is better than upper, in most fonts.
Really interesting, I don't get it why everyone is so eager to shoot this down. Of course you can't know the benefits before you learn it. Also the pure dots present just the base system, typography and different typefaces can evolve from there with more aesthetics, redundancy, readability etc. I'm definitely up for it for science/entertainment!
I don't propose that Dotsies is without flaws or destined for any amount of adoption. Having said that, here is my parody of the reaction the poor sap (let's call him Marcus Librus) got who originally posted the latin alphabet on the ancient Roman version of Hacker News (let's call it Papyrus Tidings):<p>Brutus Quidus: I've looked at your alleged alphabet for 30 seconds and have discovered all of the flaws. It is clear it will never work. First off, your O and Q are nearly impossible to distinguish. What if there's a piece of dirt on my scroll where I put the O? It's obvious you didn't do any research.<p>Claudius Acerbus: Is this serious? This will never ever work. Besides the achilles heel that if you have an N next to an I people will think it's an M, that circle thing looks just like a D. And if you have 2 D's on top of each other people will think it's a B! Christ! Err, wait, not born yet. You'll see!
if this idea has any legs at all, the first improvement will be to distinguish vowels from consonants ... and then to make sure common letter sequences that have particular meaning (English), -ing, -ly, -tion, un-, in-, etc., stand out visually. I can live without upper/lower-case distinction for most reading.<p>and common 1- and 2- letter words need to look unambiguous when floating independent of a readily visible base- and top-line you'd see in most longer words.
I was surprised by the lack of research justifying the claims about the latin alphabet and/or supporting the concept of dots. Does anyone know of any research in this field?
I might be convinced if I could see a list of words. The symbols are too difficult to distinguish individually, but common words could eventually become recognizable, perhaps.<p>My opinion is: if you're going to learn something enough to be fluent in it, why not make it a real language spoken (or in this case, written) by actual people?<p>And finally, I'll give the author twenty bucks if he can read arbitrary text that I supply. (This would have to be an in-person thing.)
When I read something (mentally), I can see that there are two parallel processes that take place:<p>1. The "physical" aspect of mental reading - mentally sounding out the words that are in front of me. This is quite fast already and is pretty mechanical.<p>2. Converting the words into meaning and registering it in my mind. This is the much harder task, and you actively need to focus to do this.<p>To see that these two are indeed parallel processes, notice that it's possible to "read" something in your mind without any of it actually registering while you think of something else, say a narrative about just how awesome your life would be if X or Y were true. I often have to actively break the internal narrative and pull myself back to focusing on what's read.<p>The focusing part is the real bottleneck, not the mechanical reading part. This is painfully obvious especially when you are trying to learn information-dense material such as Mathematics, where you have to pause to recall definitions, follow argumentation, and so on.
If the creator happens to read this, I'd suggest upping the font size significantly.<p>There's a reason children's books have such a large font. When we learn how to read, it's much easier to focus on large single characters. Only after we become accustomed to reading individual characters can we then start to recognize words.
Christ people! This is so stupid I have to think we have been trolled with this one.<p>"Optimized for writing, not reading?" You have to be kidding me, right?<p>Yes -- and the Romans had indoor plumbing, aquaducts, and crosswalks. All just as antiquated as the serif.<p>See there? I've fed the trolls.<p>-- I would love to see the brouhaha this would elicit from typophile.com.
I always thought we could create a better language based on more efficient grammar rules and language habits. I just don't think the rest of the world can follow. I'm intrigued to try this out though. I'll see how it goes over the next couple of days.
Making a modern condensed character set is a nice idea, but this is a really, really bad execution.<p>Many letters are the exact same shape, shifted vertically 1/5th of a character height? A-E are completely identical but for position. This is terrible enough to be a show-stopper on its own.<p>Braille has all the advantages of this, but is FAR more readable, since each letter has its own distinct shape. You say in a different thread that "Braille has the same proportional shortcomings as normal text". This is false. Braille characters are a grid of a fixed number of dots (6 or 8), and every character is the same size. It also has numbers, punctuation...
Wait, what problem does this solve? I have to agree with @nhebb in doubting whether this can be serious.<p>The idea is provocative, but contains the seeds of its own destruction. If the deficiency in Latin script is that it is (putatively) "optimized" for one function and not another, why make the same (or rather opposite) mistake <i>on purpose</i>?
trogdoro, I love this idea, but I'd suggest choosing dot combinations that visually approximate the shape of letters where possible, For instance:<p>- lower case L (l) would be a long vertical bar<p>- i would be the lower two dots filled and with a dot in the second highest position<p>You could then make some other common letters look like a vertical cross section:<p>- o could be the bottom dot and the middle dot<p>- s could be the top, middle, and bottom dots<p>- m could just be the bottom dot<p>- w could just be the top dot<p>More complex letters could represent crossing lines as gaps. For instance:<p>- t could be a vertical bar with the second highest dot empty<p>- k could be a vertical bar with the middle dot empty<p>You could add a sixth dot underneath the normal line for letters which commonly drop down below. For instance:<p>- j could be an i with an extra dot underneath<p>- p could be an o with an extra dot underneath<p>- g could be an un-dotted j<p>- y could be a v with an extra dot underneath<p>You could take advantage of certain letters which are the inverse of each other by inverting the dots (white dots become black, and black dots become white. For instance:<p>- z could be the inverse of s<p>- q could be the inverse of p (but with the lower dot still filled in)<p>For letters which are upside down versions of each other, just make the symbol upside down<p>- w would be an upside down m (as already shown above)<p>- n would be an upside down u<p>You could also add an extra dot above the normal line of reading to denote capitalization (kinda like a representation of the shift key). Doing this would make for 7 dots total.<p>I'm sure that not every letter would be representable using rules like this, but the less common letters could get the less obvious patterns.<p>The nice thing about doing it this way is that some words will kind of look like the shapes we have already memorized for those words. Kind of. If you squint.<p>This is a fun experiment, and I commend you for trying it. I look forward to seeing vs 2.0 :)
I added a link to an interface for trying your own mappings. It didn't get any notice, so adding it here as well:<p><pre><code> http://dotsies.org/design-your-own/</code></pre>