![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
Speaking of male and female pronouns, I just found this interesting little history of the words "he" and "she" on Michael Quinion's World Wide Words site, the response to an insightful question from an 8-year-old. It turns out that there was a period in the 12th century, while English as we know it was still coalescing, in which it had no separate masculine and feminine subject forms for its third-person pronoun; men and women were both "he" because the Old English female pronoun "heo" was no longer pronounced differently. But this wasn't a fully gender-neutral third-person pronoun like some languages have, since the object forms "him" and "her" had not converged; and eventually the word "she" appeared.
Dec. 6th, 2003
Graphing stuff
Dec. 6th, 2003 10:39 amThe program that originally sold me on Macs was Graphing Calculator. I think it was Mac-only at the time. A limited free version of it actually shipped with Power Macs. Just seeing those 3D graphs spinning around was pretty impressive back in the mid-nineties.
Graphing Calculator still works fine in the Classic environment under Mac OS X, but it's pretty clear that it's never going to be ported to a native version. The problem is that the Mac version doesn't use OpenGL, but is instead heavily dependent on the QuickDraw 3D API that was part of the legacy cruft dropped when Apple created the forward-compatible Carbon framework. There are emulations of QuickDraw 3D for OpenGL, but it's hard to say as an outsider how easy the port would be or whether there would be difficult licensing issues for a shareware program. People are working on OS X-native Graphing Calculator clones, of course; a promising shareware one I've tried is Graph-O-Matic. But it's got a way to go.
Also, Graphing Calculator is better as an instant visualization tool than as a means of preparing usable scientific graphs. Its 2D graphs are incredibly slick, but in 3D it lacks such niceties as labeled axes, and the rendering options aren't really what you'd want in a publishable figure. And, of course, everything is at screen resolution; you can't get vector PostScript output. So for serious graphing I used to fall back on the ancient copy of Mathematica that I bought back when I could legitimately get an academic discount. It's probably gotten better by now, but Mathematica's graphing, while powerful, was always kind of painful to deal with, since such things as view parameters were handled through the command line interface, with graphical tools as adjuncts for creating command-line parameters. It was pretty slow, too, since it did absolutely everything through its PostScript-based rendering model in software. And Mathematica just gets more and more astronomically expensive.
When it comes to scientific software of any sort, you're probably going to end up falling back on the open-source community sooner or later. There's the venerable gnuplot for making useful if unspectacular-looking graphs, of course, and all sorts of other programs hook into it. And there are all sorts of other packages like NCAR Graphics. I haven't played with any of these much recently.
For really nice-looking 3D plotting with a better user interface, an OpenGL-based Unix/X11 program that seems to have lots of potential is Dynagraph (not to be confused with many other things called Dynagraph), which is available as a Fink package for Macs.
Dynagraph is not a full-fledged Graphing Calculator replacement (it just graphs things and is only for 3D work), but in some ways it is much more powerful. It seems to be an attempt to reproduce the 3D plotting capabilities of symbolic math packages like Mathematica and Maple, plus the features that OpenGL provides, such as smoothed surfaces and the ability to spin the graph interactively. I like the division of labor between the command language for function entry, and direct graphical manipulation for view, lighting and shading parameters. It labels axes, produces PostScript output on demand, and can even use fancy Level 3 PostScript features that are good for making shaded plots.
There are things that could be better. Development so far seems to have concentrated on the fancy visual stuff at the expense of other things. There needs to be a real user interface for saving graph files, and "You goofed!" is not an adequate all-purpose command error message.
(You may have some trouble constructing it in other ways right now due to this, which is worrying. I hope that the attacks there and on Debian don't essentially shut down the free-software movement by eroding the trust in code; the security benefits of open source only apply if someone actually inspects the code and if you're sure that the code you've got is the same code that was inspected.)
Graphing Calculator still works fine in the Classic environment under Mac OS X, but it's pretty clear that it's never going to be ported to a native version. The problem is that the Mac version doesn't use OpenGL, but is instead heavily dependent on the QuickDraw 3D API that was part of the legacy cruft dropped when Apple created the forward-compatible Carbon framework. There are emulations of QuickDraw 3D for OpenGL, but it's hard to say as an outsider how easy the port would be or whether there would be difficult licensing issues for a shareware program. People are working on OS X-native Graphing Calculator clones, of course; a promising shareware one I've tried is Graph-O-Matic. But it's got a way to go.
Also, Graphing Calculator is better as an instant visualization tool than as a means of preparing usable scientific graphs. Its 2D graphs are incredibly slick, but in 3D it lacks such niceties as labeled axes, and the rendering options aren't really what you'd want in a publishable figure. And, of course, everything is at screen resolution; you can't get vector PostScript output. So for serious graphing I used to fall back on the ancient copy of Mathematica that I bought back when I could legitimately get an academic discount. It's probably gotten better by now, but Mathematica's graphing, while powerful, was always kind of painful to deal with, since such things as view parameters were handled through the command line interface, with graphical tools as adjuncts for creating command-line parameters. It was pretty slow, too, since it did absolutely everything through its PostScript-based rendering model in software. And Mathematica just gets more and more astronomically expensive.
When it comes to scientific software of any sort, you're probably going to end up falling back on the open-source community sooner or later. There's the venerable gnuplot for making useful if unspectacular-looking graphs, of course, and all sorts of other programs hook into it. And there are all sorts of other packages like NCAR Graphics. I haven't played with any of these much recently.
For really nice-looking 3D plotting with a better user interface, an OpenGL-based Unix/X11 program that seems to have lots of potential is Dynagraph (not to be confused with many other things called Dynagraph), which is available as a Fink package for Macs.
Dynagraph is not a full-fledged Graphing Calculator replacement (it just graphs things and is only for 3D work), but in some ways it is much more powerful. It seems to be an attempt to reproduce the 3D plotting capabilities of symbolic math packages like Mathematica and Maple, plus the features that OpenGL provides, such as smoothed surfaces and the ability to spin the graph interactively. I like the division of labor between the command language for function entry, and direct graphical manipulation for view, lighting and shading parameters. It labels axes, produces PostScript output on demand, and can even use fancy Level 3 PostScript features that are good for making shaded plots.
There are things that could be better. Development so far seems to have concentrated on the fancy visual stuff at the expense of other things. There needs to be a real user interface for saving graph files, and "You goofed!" is not an adequate all-purpose command error message.
(You may have some trouble constructing it in other ways right now due to this, which is worrying. I hope that the attacks there and on Debian don't essentially shut down the free-software movement by eroding the trust in code; the security benefits of open source only apply if someone actually inspects the code and if you're sure that the code you've got is the same code that was inspected.)