R
Roedy Green
Incidentally, but not too far off-topic (for the newsgroup, if not for this
thread ;-) mangled identifiers like abbreviations and that ghastly "Hungarian"
convention /really/ throw off my ability to read. Perhaps for people who "read
aloud inside" the mangling has little effect, but I can't read the things at
all. I worked for half a year on a codebase that made heavy use of Hungarian,
and by the end of it I was no nearer being able to read that code than at the
start. The same point applies to people who post in "text-ese" -- if you read
8 as the sound "ate" then you'll be able to decode h8 easily (or rather, with
no more difficulty than you would when reading real English), but for folk like
me, h8 is no more meaningful than, say, h7. I hardly ever event try to read
such posts, a simple 'u' for you is enough to put me off.
That is quite a profound insight. Your language processing can handle
either sounds or concepts it already has in its bank. If an
identifier is made up of random letters, you need a totally different
part of your brain to process it.
I attempt to bypass that by making up a pronunciation for a strange
identifier someone has used, that I can on demand turn back into a
string of letters. I might even try to associate some images,
emotions, or colours or personality with it to try to objectify it.
But code with ordinary English words is so much easier, even if the
result is longer. I guess then my speed reader kicks in.
When the code is under my control, the global rename function gets a
workout to remove these names as my first priority.