@ShutaoTang Wouldn't it be trivial to define it yourself? Like DifferenceQuotient[f_, h_] := ((f[# + h] - f[#])/h &) and use it like DifferenceQuotient[Sin, 1][E] or even Limit[DifferenceQuotient[Sin, h][x], h -> 0]
Heh, everything's remained mostly the same, except that these days you can't read two lines in the documentation without running into 5 "Wolfram"s and nothing works without phoning home to W|A to quantity-entitize the expression.
@Guesswhoitis. good to see you're back. I'm curious to the details of your absence, but there is no need to go all public about this. Is there any way we could help you? A little fundraiser or so?
@SjoerdC.deVries Thanks for the greets. The past few months have been quite complicated and messy, so it's good that I can go back to you guys again. :)
"Is there any way we could help you?" - if there is, I'll let you know. For now, I'm crossing my fingers that I can get my old computer back from the pawn shop.
(Even if I didn't pawn it, I'd need to mostly install stuff from scratch, anyway; I had a nasty hard disk failure before said pawning.)
@Guesswhoitis. We've Dataset which sounded really useful, but which is currently far from that as the current implementation with Association takes about 10 times the memory a normal table would take. It's query syntax is quite complicated too, and I often find it necessary to convert from and to lists using Normal for many tasks.
@OleksandrR. I'm touched by all the concern; thanks. :) It's... okay here, but it could be better. Nevertheless, I should be able to contribute again. Maybe more when I get my computer back. :)
@Guesswhoitis. well, I hope your reappearance here is a sign that things are improving for you. It's been a long time. I really missed your contributions. (You're far better at math than I am and much better read in that field, so I learn something from practically everything you write!)
@SquareOne CurrentValue is one that it's probably easiest to just read the doc page. Regarding language design, I definitely think it should be expanded to include all keyboard keys instead of just modifier keys. There are also potential suggestions if some other values (besides MousePosition, MouseOver, etc) are common enough to warrant additional shortcut symbols.
@IstvánZachar @Guesswhoitis. Ah, BarLegend has to delay evaluation of the colors until displayed I would guess because it doesn't know its size or how many samples of the blend function are needed until displayed.
@MichaelHale to be correct: it holds the first argument, if I'm right.You must be right @Michael, but I simply did not expect BarLegend to act as a scoping construct...
Ah. "BarLegend is displayed in StandardForm as a graphics object. In InputForm, it is displayed as an explicit list of color function, data range, and contour values."
@IstvánZachar Well it doesn't hold them. It's just the first argument is a function, which is held, that doesn't get applied to any values until the legend is displayed in standard form. By that time max has been reset.
I've noticed there had to be something different about legends because they aren't selected as part of the graphic when you select a plot that has a legend.