"Walter Bright" <[email protected]> wrote in message news:[email protected]... > Andrei Alexandrescu wrote: >> That definition is what was discussed in this thread and alleged to be >> anything but beautiful. > > I've been in museums in europe where they proudly display ornate swords > and armor as "beautiful". I always kinda thought otherwise, because all > that decoration and encrustation was not what the weapon was for. More > interesting were the weapons with a single minded deadliness to them. I > suppose it's the engineer in me <g>.
My take on that: "Beauty" usually refers to something being appealing to the senses, particularly sight. With something like code, though (as opposed to a physical object) it makes a little bit more sense than it normally would to use the term "beauty" metaphorically (ex: "functional qsort is/isn't beautiful"), because applying the traditional concept of "beauty" to code would amount to little more than an assessment of the formatting (font, spacing, etc.) and would not carry much, if any, relation to the actual content of the code. That content, of course, being the very thing that makes code code in the first place. I suppose a similar thing could be argued for swords (because one could define "sword" in terms of its intended function, just like "code"), but I think the argument there would be weaker because it's much easier to define something like "sword" in terms of its structure (shape, material, etc) than to define "code" in terms of its structure (one would probably have to start with a definition of certain glyphs and then find a structurally-oriented way to distinguish sequences of glyphs that do or don't qualify as "code"). So the criteria you're using to determine suitability of the label "beautiful sword/armor" is something I'd prefer to attach to a label more along the lines of "good sword/armor" or "well-designed/engineered sword/armor" (which, of course, IMO, is every bit as worthy of admiration as a pretty-looking one).
