Hacker News Reader

Comment by salutonmundo | Hacker News Reader

Most sans-serif fonts poorly distinguish between "I" and "l", I've found. (This led to my dad thinking that artificial intelligence was called "Al", as in "Albert".)

View on HN

This is why I prefer a fixed width font when it can be ambiguous. They don't typically have the ambiguities between "l" (lower case "L") and "I" (upper case "i") or even the number "1", similarly between the letter "O" and number "0".

Personally, when handwriting, I got in the habit of striking a hyphen through the number "7" to disambuate against a possible number "1". Likewise, I'd strike through the letter "Z" to disambiguate from the number "2", and a diagonal strike through the number "0" to differentiate from the letter "O".

My handwriting is terrible, and it has saved me a lot of grief over the years, but especially in my engineering courses in college.

I don't know if that's actually traditionally true of fixed-width fonts. Courier, for example, is pretty bad, and widely used by people who aren't using a fixed-width font because they're programming.

Traditionally you'd use a lower-case ell for a numeral one, since there was no numeral one on your keyboard. Nor was there a separate zero. All of this "different characters for different uses" nonsense is for that newfangled electronic stuff.

"I had to code a database using just ones and zeroes."

"Oh, you had ones? We had to use the letter ell."

I’ve adopted all of these as well. Doing statistics in my undergrad, not understanding what you or someone else wrote was just the worst. The change of style just became an obvious necessity.

7 and Z have far less disruptive fixes.

For 7, just make the top long and the slash slanty. Don't do that for 1.

For Z, give the top a very slight curve, and optionally do the same for the bottom. That is, bend the middle of the line in toward the center of the letter. It might be enough to just imagine doing this, so that you don't end up with a curve going the other way. Give the 2 a proper curve instead of a sharp corner, and it will look different.

That's fine but the reason I use the extra bits is to try and make a character unambiguous for a reader who does not know my handwriting habits. In the case of Z I do both - your suggestion and put a bar across it. 2 gets a small loop at the bottom left.

Or (for 7) as in [0]

>in Japan, Korea and Taiwan 7 is written as ① in the illustration to the right. [1]

[0] https://en.wikipedia.org/wiki/7#Evolution_of_the_glyph

[1] https://en.wikipedia.org/wiki/7#/media/File:Sevens.svg

This is what I do I still use German/engineering Sevens and Zed's - it was that I was taught at 19 when I started work.

Same, I use crossed 7 and z, but I think of them as French (which was my first foreign language at school), and started using them in A-level maths (UK).

Similarly, but I dot my zeros to distinguish them from the null / empty-set symbol ∅ (and from the Scandinavian Ø).

> Personally, when handwriting, I got in the habit of striking a hyphen through the number "7" to disambuate against a possible number "1". Likewise, I'd strike through the letter "Z" to disambiguate from the number "2", and a diagonal strike through the number "0" to differentiate from the letter "O".

My handwriting also makes it hard to distinguish q from 9 in a situation where the baseline isn't clear.

I like and use your solutions, but the problem is that, if the reader doesn't know about them and doesn't have the other candidates to compare, a '1', a '2', or an 'O' has no visible mark to indicate that it's not a '7', a 'Z', or an '0'—that is, they are identified by missing, not additional, information. (I don't know any good solution, though.)

Surprisingly, in my experience, most people pick up on it with no explanation. Ive only had to explain it to a handful of people and to each, once. My handwriting is awful and has only gotten worse due to nuerigical issues. I look at my writing from my college notebooks versus current, and comparatively, my college writing is legible and I'm now producing chicken scratch that only I can decipher.

Sure! I agree that, if I saw your O side-by-side-with your 0, or even your 0 by itself, I would instantly understand. My point is just that, if I saw your O by itself and hadn't seen any of your handwriting before, I wouldn't know "aha, this is an O because it doesn't have a distinguishing mark", since I wouldn't know that your 0 does have a distinguishing mark.

My fantasy company would promote their wonderful new Al method of doing X, where X is just done by some guy named Al.

Investor: OH wow is that a Neural Network you guys are building?

You: Yeah, Al here is our resident Network engineer. Go ahead and ask him a question.

Investor: So do you use any kind of drop-out or just as-is?

Al: Yeah, we don't have too much packet-los-

You: OK Moving along...

Just say you trained a bunch of neural nets. https://xkcd.com/2173/

Are you the son of Scott Bakula?

Reply on news.ycombinator.com