It would be really, *REALLY* nice for the BPFK if we had a large
corpus of Lojban text all in one place.
To that end, what I'd like is a system that does this:
1. Sends email to a fixed address (probably the BPFK list) on every
2. Accepts input of either uploaded files or URLs, with
descriptions and such.
3. Presents a list of same that allows for changing the URL or
description, or re-uploading if it's a file.
4. Once an hour or so, runs through everything, strips crap like
HTML out, and makes one giant text file out of it, with delimiters
saying which part came from which. Provides a constant URL to the
Things I can help with:
A. You're welcome to an account on lojban.org ; it's a Debian Linux
B. I have code that checks on a line-by-line basis for text that is
at least 80% Lojban; we could have the sources stored by this system
optionally (a checkbox or something) pass everything through that
C. Stripping HTML or TeXInfo or just about anything else is
absolutely trivial for me; I can help with that part.
This looks to me like about a 5 hour job; if it looks much longer to
you, either you need to find a better web framework or I've failed
to communicate what I have in mind.
They say: "The first AIs will be built by the military as weapons."
And I'm thinking: "Does it even occur to you to try for something
other than the default outcome?" See http://shrunklink.com/cdiz
http://www.digitalkingdom.org/~rlpowell/ *** http://www.lojban.org/