About this Site
Part of the downside of doing stuff is that in many cases, if you do it well, people won't notice a thing (they only notice when it's done badly). It's something of a stretch to say that my site is such a masterwork, but even so, I did some cool stuff (at least I thought it was cool) in making it, and maybe I should start to share some of that. A lot of it is ad-hoc kludges, but when it comes down to it, an awful lot of programming is just that. Someone needs to get something done, and pounds things together fast as you can just until it works. How many teeny little shell-scripts like that do you have kicking around in your bin/ directory?
I decided to try to map out the site before starting, for a change, if only to get some sort of handle on all the things I had to put on it without them slipping through my fingers (never can keep track of them all at once). I thought I'd try out FreeMind and see if I could use that to organize things. A tree structure is a good one to work with anyway. It's turned out to be pretty helpful.
I make a tree-like data structure mirroring the site's structure and use that to build the menus. I use the same trick on the KLI's site, but it's not all that easy to use, especially as things are developing and being rearranged and added. I eventually wrote a little text-file with the relevant information in indented format to show the tree-structure, and then a Perl script to parse that and make PHP statements that would build the tree appropriately. (Programs writing other programs is hardly new—it's basically what every compiler does—but it's still neat). Then when I realized that copy-paste from the FreeMind map would yield exactly the same sort of indented file, I moved the text file into a section of the mindmap of the site, and copy and paste it out when I need to.
site will be growing
(I hope!), a lot of its planned sections
aren't in place yet, so links will be broken. I generate internal links
with a little function that checks to see if the file is there, and if it
isn't, puts in text in
broken style (which the CSS defines), not a
link, and with a title saying that it's broken. That way users don't have
to find out the missing parts the hard way (404 errors), and when the
missing pieces are added in, all the links to it automagically start
working and I don't have to go and change anything.
I'm adding an Atom feed for the site so you can keep up with major changes I make, as they occur. At this point I'm maintaining by hand; we'll see if that becomes unwieldy.
Most sites use the
www prefix for web servers. Why does my site call
itself web.meson.org? I can understand the desire to have a
conventional web-server name, and to keep it distinct from the various
other names one's server might go under, in order to be able to factor out
the uses of the site distinctly from the actual hardware. But www
is a dumb name. It's a pain to pronounce. Pronounced correctly, it takes
a whopping nine syllables, not counting the
dot at the
end. Holy verbosity, Batman! And that's why people have taken to such
dub-dub-dub. Me, I
believe that the best and only appropriate pronunciation for the prefix is
wubba wubba wubba. But why even that? It's a web-site, right?
web is every bit as clear as www, if not more so, takes just
as many letters, and takes only one solitary syllable! (Don't you
try to whine that www is easier because it's one letter typed three
times. Just don't go there.)
OK, ok, yes, I do have an alias of www.meson.org for the site, so that will also still work. But it isn't the real name.
- FreeMind, pretty neat "mind-mapping" software, in Java so it runs on many platforms.