Home> urticator.net Search About This Site Domains > Glue Stories Art Memes The Mind The Body Language Philosophy > Strategies Other Other (2) Anti-Consumerism Environment Free of Distraction Time Easy Deflection Flexibility No Eking Voluntary Simplicity Happiness Mind Maps > Journaling Too Much Is Eventually Enough Deliverables
|
Days of the WeekI enjoy working with data in fixed-width formats, so naturally I also like having codes for things, especially fixed-width codes. Here I'd like to talk about codes for the days of the week.
3. If we take the first three letters of the names of the days, we get an excellent set of codes, familiar and unambiguous. The only problem is that the codes are too long for an enumeration with only seven values. 2a. If we take the first two letters, we get a set of codes that's unambiguous but flawed. If you want to read the codes aloud, well, for two of them, there's no vowel, and for most of the rest, the vowel has the wrong sound because the following consonant is missing. For example, the “u” in “Sunday” sounds like “uh”, but the “u” in “Su” is more naturally pronounced “oo”. Actually, the problem isn't that I want to read the codes aloud. When I see a code, my brain automatically pulls up the name of the day (because that's the point) and the sound of the name (because for me everything is verbal), but it also automatically tries to turn the code into sound. If the two sounds aren't compatible, that creates a bit of unpleasant tension. The alarm clock on my phone uses two-letter codes. 2b. We can solve the problem of codes with no vowels by adding vowels. Amazing! The resulting codes aren't good for anything, I just like to use them to construct exotic-sounding names like Wethufri Sasumotu. 2c. The codes in set 2b happen to be generated by the rule “cut after the first vowel”. But, isn't the rule “cut after the first group of vowels” obviously nicer? The only difference is that Tu becomes Tue. 1a. If we take just the first letter of the names of the days, then by unhappy historical chance we get a set of codes that's ambiguous. To be specific, the codes T and S appear twice each. The set can still be used, but only in places like at the top of a calendar page where it's clear from the context which day is which. 1b. We can solve the problem of ambiguous codes by adding letters, and what we get, in my current opinion, is the cleanest and most consistent set of codes. I don't think I've ever used it for anything, though. I guess I prefer short codes to clean and consistent ones. 1c. If we replace Tu with T, we get another strong contender. Is there any way to justify the change? Not that I can see. The rule “cut before the first vowel, then disambiguate” explains what the appeal is, but also requires us to replace F with Fr. 1d. I have a paperweight that's also a perpetual calendar. I checked to see what set of codes it used, and now I feel obliged to include that set in the table, even though it's strange. I did some work with negative digits recently, so the red S makes me think that Sunday is negative Saturday. 1e. Now it's time for a story. When I was in junior high and high school, computers were just starting to become common. As one example, the junior high school was on the cutting edge in that it had a computer facility … and if I remember correctly, that facility consisted of a couple of teletypes tucked into a nook under the stairs. To be clear, I'm talking about the kind of teletype with a keyboard, a seven-pin print head, and a roll of yellow paper. What did we (the students) do with the teletypes? I think we used them as calculators occasionally in algebra class, and also used them to write and run simple programs in BASIC. The only reason I wasn't excited about the programs is that by that time we had an Apple II at home, with BASIC and many other fun things. As another example, the schedules at the start of each semester and the report cards at the end were all printed by a mainframe somewhere and then mailed out. I had saved most of my schedules and report cards in a folder, but at the time of the fire, that folder was in a box in the one corner of my apartment that actually did catch fire, so you'll have to trust my memory. So, imagine thin slips of paper, mostly white but preprinted with some blue design elements, covered not too densely with rows and columns of characters smaller than the ones produced by the teletypes. The letters were all upper case; the print color was different for each slip, somewhere between black and medium gray depending on how much the print ribbon had been used; and there were columns for things like the name of the class, the name of the teacher, the room number, and the time. That brings us to the point. Because the schools had flexible schedules, there was also a column for the days of the week that the class was held, and because the fixed-width format was a constraint, one of the schools had constructed an unambiguous set of single-letter codes by making the interesting decision to let the code for Thursday be R. For some reason, that made a big impression on me. As I remember it, the school was actually the local university (see My Mathematical Education), and it had also decided to let the code for Sunday be N, but I'm less sure about those details. 1f. Sometimes it takes me a while to figure things out. After I started keeping a journal, it took me about a year to realize that it was useful to record not just the date but also the day of the week. At the time I was using my best penmanship (in pencil) so that the journal would look good, so it was natural to write out the full names of the days. I gave up on the penmanship after two years, but it took me another three years to get tired of the full names and start experimenting with codes. Then it took me another two years to finish experimenting! The set of codes that I happened to settle on … well, I'm not sure how I want to describe it. Maybe it's a hybrid of sets 1c and 1e, or maybe it's just set 1e with R replaced with Th. I still use the same set of codes today. I really wanted to like R. I kept trying it out, thinking that eventually I'd get used to it, but I never did. See what I said earlier about unpleasant tension, I guess. I don't know why I didn't have the same problem with N. Although Th has twice as many letters as R, it doesn't take twice as long to write. I write R as a vertical stroke followed by a stroke with two parts, while I write Th as a vertical stroke followed by a stroke with three parts. 1g. One of the ideas that I experimented with was to let the code for Thursday be Θ (theta) because it represents a “th” sound. I was in graduate school at the time, surrounded by equations and Greek letters, so it was a natural idea to come up with. That's an upper-case theta, by the way. In print, it has that fancy detached crossbar, but when I write it by hand, I just do it all in one stroke, same as lower-case theta (θ) … start on the left side, go down and around, then across. I just make it about twice as wide when it's upper case. A better idea that as far as I know never occurred to me until I was writing this essay is to let the code for Thursday be Þ (thorn) because it represents a “th” sound and was used in English in the past. One trace remains even now: near the end, “y” was used as a substitute for “þ”, and that's why the “ye” in “ye olde” is supposed to be pronounced “the”. The Wikipedia article on thorn has many other interesting details. That suggests another idea, that we could let the code for Thursday be Y, meaning thorn, but I guess that's too obscure, even for me. 1h. Later I also tried out set 1e with R replaced with Θ. I don't have much to say about it, but it's a nice set of single-letter codes. That's all the sets of codes I want to talk about. I know I've been jumping all around, so here's a table that summarizes the ten (2×5) options for short codes that seem plausible to me. Six of the ten appeared in the original table.
|
See Also@ December (2021) |