From: Mitch Tishmack Date: 2005-10-23T11:27:41+09:00 Subject: Re: Rubyists into Lojban? I have dabbled into Lojban a bit. Not anything too serious at the moment. I am learning French and German so I have my mind occupied enough for now But I have been building a mini flash-card-like program for my own aid in remembering vocabulary, I treat similar words as sets. I have been trying to keep it semi-generic. But to expound on your annotator program. I would rather have something similar to this: http://www.rikai.com/perl/Home.pl Sure written in perl, and more for kanji to furigana/katakana/ hiragana etc... But the general idea is much more user friendly than forcing someone to click on a link to obtain information. Although I would personally prefer using invisible floating divs that are shown when you mouse-over the word. I don't want information underneath or over when I am reading. The main reason is it acts as a crutch when reading. Use it until you learn it then disregard. I loved the advice of my first language teacher, don't translate when you are reading. Memorise the words and read/think in the language as much as possible. Something like this is what I think would be the closest to what I would build: http://www.brunildo.org/test/CPop.html Or similar to Eric Meyer's original pure css pop-ups as documented here (without fixed positions however): http://www.meyerweb.com/eric/css/edge/popups/demo.html I just need to get ruby embedded into an Objective C program for OS X and then the fun begins for me. But documentation for this feat is less than stellar and my attempts core dumped... alot. (oops :) ) Anyway, I need to get some sleep. I have to go to work in about 6 hours and patch stuff. Mitch On Oct 22, 2005, at 6:31 PM, Hal Fulton wrote: > Ryan Leavengood wrote: > >> I've had some forays into learning Lojban in the past, as well as a >> bit of Esperanto. In fact I can recall a few years ago when I used >> IRC >> more, my two channels of choice were either #ruby-lang or #lojban. >> Unfortunately I never did get too far in learning the language, and >> have forgotten most of what I did learn ;) >> But part of my goal was always to try to create a decent >> human-computer speech interface based on Lojban, which due to its >> logical and regular nature would probably be easier than using >> English. This would mostly be for fun, learning and my own personal >> use (since I doubt the average person would learn a new language just >> to talk to their computer.) >> So, anyhow, let me know what you are planning Hal. >> > > Well, I've thought of a few things. > > 1. An ri-like utility that works at the command line -- give it a > Lojban word, get back all info about it > > 2. A program (using PDF::Writer) which would hopefully make TeX > unnecessary in spitting out a nicely formatted dictionary > > 3. A program that would make what the Brits used to call a "trot" -- > a text with the hard words translated in small print underneath > each line. And a page-oriented variant on this. > > 4. A nice parser. > > 5. A sentence diagrammer?? See my blog. > > 6. An annotator. Here's a first pass at it: > http://hypermetrics.com/pikmin.html > > As for the annotator, I have bugs to fix and features to add. One > big feature I'd like is the ability to create an entire downloadable > annotated document (which someone could then host on his own site > or view offline). It also needs to be updated in where it gets its > content. > > 7. I thought of a text-to-speech thing (not voice recognition), but > I would do it by cheating -- convert the Lojban to phonetic gibberish > (e.g. "gismu" -> "geese-moo") and run it thru an ordinary TTS program. > Trouble is, cheating would sometimes get difficult, and many TTS > programs don't have certain phonemes like Lojban {x} (unless you > select a German voice or something). > > > Those are pretty much all my thoughts. > > > Hal > > >