> It is not so surprising. When you want it, you cannot make it. And when you > can make it, you no longer want it.
Yes, but... Purely at the level of personal skills-acquisition, I agree with you. Writing such a tool can be as instructive as using one. Even before you've got it working, the tool has often done all it can for you. So you put it aside, unfinished. Smoothing the way for novices is a different matter. The motivation is different, and must be enough to sustain the developer through to the end. It can be a thankless task... Several times in my career I've been on projects to develop high-grade teaching materials for degree courses -- and trying to sell the results, or bid for funds to make it happen. Demos, tools and case-studies. You soon notice that the better-designed the tool is, the more efficiently it delivers its message and the less time the student spends on it. This doesn't impress headmasters and course-leaders. They say: "I'm not spending all that on a tool the student will only use for a week." Yes, but what if it took a term for students to master the topic without the tool, and some never did -- to their detriment? Let me tell you about the IT degree course at my university which eliminated subsidiary mathematics (30 hours) in favour of 2 hours' "Excel formula development"... (But not on this list.) Ian On Mon, Nov 22, 2010 at 1:15 AM, Dan Bron <j...@bron.us> wrote: > Ian wrote: >> can't believe [a tacit-to-explicit translator] >> hasn't been done already. > > It is not so surprising. When you want it, you cannot make it. And when you > can make it, you no longer want it. > > For the subset that can make it and still want it, it is somewhat difficult > to remember exactly what you wanted. For example, where is it appropriate > to cut a tacit one-liner into smaller chunks to be distributed among several > lines in the resulting explicit definition? Certainly not at every @ . > Etc. > > On my J todo list for several years has been a J "coach" similar to the > "Regex Buddy" at [1]. I've just never got up the initiative to do it. On a > more promising front, Bob Therriault has recently been building web videos > that demonstrate certain J primitives and tacit expressions graphically [2]. > > Personally, I learned tacit by writing, rather than reading, a lot of it. > J's default interface encourages this habit; it is productive & fun to build > programs in the IJX window, which effectively limits you to one line. Using > IJSes & flipping back and forth was enough of an additional friction that I > never developed a taste for explicit code and its multiple lines. But > having written enough, I also find tacit easy to read. Whether it's easier > or harder than explicit (or than Java, or English, etc) depends on the > application. > > -Dan > > [1] Regex Buddy: http://www.regexbuddy.com/regex.html > [2] Bob's animated explanation of J adverbs: > http://bobtherriault.wordpress.com/2010/11/17/those-tricky-adverbs/ > > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
