Ok, so now Reader has powerful search capabilities, with the option of using regular expressions. And whacky font sizes aren’t such a problem anymore. Something just occurred to me however, and that is that a "web style" search might be an option, since people are probably pretty familiar with the style by now. That is that words separated by spaces are searched for as separate words. So searching for "hello there" would match both "hello" and "there" rather than "hello there". Currently to do a search for one of many words, the more rancid RE style syntax must be used, eg "(hello|there)". Not quite as user friendly methinks. Ah well, I’ll get right onto it…

Hey, wait a minute… did I just lose a day again?

Hmmm… what else… oh yeah, eLibrary got another few tweaks, filtering out some duplicate names in there. I have to come up with a way to filter out all the different versions of a text in the standard view, while still allowing access to them.

Heres a nice thing to read if you ever have to help an inexperienced person use a computer. Other nice articles at the site too. What a nice person!

Decisions

JujuEdit supports most basic text formats (unicode formats, ASCII, Language Based code pages), but Reader does not. Current code that might otherwise be shared must be duplicated because of this difference. It looks like the thing to do is to use the same text handling code for Reader as for . Reader doesn’t really need it right now, but it looks like sooner rather than later might be a good idea.

It’s hard to gauge the need for a multi-language product, when you’re a monoglot like m’self. We are very ignorant of the rest of the world you know. As an english speaking programmer it’s very easy to get resentful about non-english language support, because other languages are not only different, they often involve much more complicated character sets. For example, any english sentence can be written using solely the 26 letters of the english alphabet. Start supporting a few other European languages however, and character variants are required (what with umlauts and circumflexes and other such exotic animals), and once beyond the latin style alphabet (to Arabic and Hebrew scripts for instance) you end up goggling at the question of how you are going to support right-to-left text on screen, as well as handling a complex ligature structure (where 2 characters are combined to form another). This is already way beyond my experience.

While I can happily render individual Arabic glyphs in JujuEdit, I have no idea whether they will look intelligible to a native reader. Asian character sets are also scarey, because they’re just so damn big! So basically, I’m apologizing in advance here to people who may (or may not) be disappointed that my "international" support extends no further than being able to render glyphs from within extended font sets.

It’s a shame that a new language wasn’t created at exactly the same time computers became popular. If we all spoke a modern kind of Esperanto with an alphabet of exactly 16 characters, it would simplify things for us programmers enormously…

This entry was posted on Friday, October 11th, 2002 at 2:36 am and is filed under Misc. You can follow any responses to this entry through the RSS 2.0 feed. You can skip to the end and leave a response. Pinging is currently not allowed.