Re: Lojban word processor for Windows?
Avital Oliver
> Lojban word-processor? I am thinking of something for Windows, but of
> course there will be a bunch who would rather write it for Unix.
I would love to do it.
(Of course, my first version would be either Unix or Java).
If it is to be graphical, I'll do it in Java, if not, then a console application
in Unix.
ma...@kli.org
>Date: Fri, 17 Sep 1999 07:26:12 -0400 (EDT)
>Cc: loj...@onelist.com
>
>From: David Brookshire Conner <nell...@concentric.net>
>
>Oddly enough I was thinking about just this problem this morning as I
>was walking to work. In particular, Loglan's regular structure gives
>the text editor lots of help in e.g., automatically typesetting (a la
>LaTeX).
I was thinking along similar lines wrt typography and typesetting for
Lojban (I've been designing a Latin-character Klingon font, with
appropriate ligatures, etc). But you can't improve too much on some nice
cold monospaced Courier-like fonts for Lojban. It fits the language.
> > What I have in mind is something that supports Wordpad style editing (there
> > may be some open code already for such a simple editor, since Lojban uses
> > the standard alphabet, in which case the programming is mostly in the
> > utilities that follow)),
>
>Another approach, of course, is to write extensions for editors that
>support that kind of thing. Emacs comes to mind :-) So does
>Framemaker (which runs on everything, though is pricey).
>
>An Emacs major mode should be straight-forward. I'm a little surprised
>one doesn't already exist (or am I wrong here?). Easy stuff like
>recognizing sentence structure and piping text to the parser.
I *like* that idea. I guess I'm surprised I didn't already write a major
mode in emacs. Then again, I also have to keep reminding myself what
people are looking for. I still expect word-processors to be just really
really good typewriter/typesetter machines, not popping up templates,
word-completions (unless you ask), stuff like that. But I'll wager that
just an as-you-type cmene-checker would get a workout!
>An initial Framemaker extension would consist simply of a document
>template with lots of smart "paragraph" styles (with each "paragraph"
>in Framemaker being a lojban bridi). For example:
>
>StartUtterance - first sentence, following "paragraph" is
>ContinueUtterance - "autonumbered" to start with ".i "
>StartParagraph - autostarts with .ni'o, next para is
> continueUtterance.
>etc. formatting makes all this look pretty.
Maybe indents for .i?
>On another tack, a lojban font would be an interesting
>problem. Specifically, the ligatures would probably be different from
>an English font (as letter frequency is different), and would probably
>emphasize the cmavo.
See above. I gave that some thought, but didn't come up with anything.
Lojban doesn't feel right with much in the way of "real" ligatures (like
fi, fl, etc). Typography to make the cmavo more distinctive might be nice,
but I'm not sure how. Except maybe to emphasize ".i", but then if you
break lines at every jufra (which I personally don't do in writing most of
the time) you don't need it. Though you do want to make sure your font
does nice things with the apostrophe. Not to much and certainly not too
little. Again, monospace seems oddly workable. The overly wide space for
the apostrophe makes sure you can't mistake "ta'i" for "tai".
> > one which
> > would call up the place structure on a mouse click,
>
>A little trickier but again, primarily just needs a database of the
>place structures.
Would come in handy, no mistake.
> > or give the breakdown
> > of a legal but undefined lujvo with a different mouse click.
>
>and a database of the rafsi....
YES. For building and reading.
And one of cmavo?
> > And then
> > finally you could invoke the parser in a pop-up window for any selected
> > chunk of text.
>
>Emacs is especially good at this.....
M-x compile.
~mark
David Brookshire Conner
> Any of you programmers out there willing to try to come up with a simple
> Lojban word-processor? I am thinking of something for Windows, but of
> course there will be a bunch who would rather write it for Unix.
Of course :-)
Oddly enough I was thinking about just this problem this morning as I
was walking to work. In particular, Loglan's regular structure gives
the text editor lots of help in e.g., automatically typesetting (a la
LaTeX).
> What I have in mind is something that supports Wordpad style editing (there
> may be some open code already for such a simple editor, since Lojban uses
> the standard alphabet, in which case the programming is mostly in the
> utilities that follow)),
Another approach, of course, is to write extensions for editors that
support that kind of thing. Emacs comes to mind :-) So does
Framemaker (which runs on everything, though is pricey).
An Emacs major mode should be straight-forward. I'm a little surprised
one doesn't already exist (or am I wrong here?). Easy stuff like
recognizing sentence structure and piping text to the parser.
An initial Framemaker extension would consist simply of a document
template with lots of smart "paragraph" styles (with each "paragraph"
in Framemaker being a lojban bridi). For example:
StartUtterance - first sentence, following "paragraph" is
ContinueUtterance - "autonumbered" to start with ".i "
StartParagraph - autostarts with .ni'o, next para is
continueUtterance.
etc. formatting makes all this look pretty.
On another tack, a lojban font would be an interesting
problem. Specifically, the ligatures would probably be different from
an English font (as letter frequency is different), and would probably
emphasize the cmavo.
> with a Lojban word-list spell checker,
Should be a matter of assembling a dictionary and giving it ispell
(for emacs) or framemaker, making sure to tell it that this is a
different language. For that matter, internationalizing any of these
editors would be interesting. Both Frame and Emacs (Mule) have support
for this.
> one which
> would call up the place structure on a mouse click,
A little trickier but again, primarily just needs a database of the
place structures.
> or give the breakdown
> of a legal but undefined lujvo with a different mouse click.
and a database of the rafsi....
> And then
> finally you could invoke the parser in a pop-up window for any selected
> chunk of text.
Emacs is especially good at this.....
Hmmm....
Might have to do it.....
A Frame template will be easiest - anyone else out there have access
to Framemaker to be a beta tester?
Brook
---------
ASCII stupid question, get a stupid ANSI!
---------
Fancy. Myth. Magic.
http://www.concentric.net/~nellardo/
Bob LeChevalier-Logical Language Group
Any of you programmers out there willing to try to come up with a simple
Lojban word-processor? I am thinking of something for Windows, but of
course there will be a bunch who would rather write it for Unix.
What I have in mind is something that supports Wordpad style editing (there
may be some open code already for such a simple editor, since Lojban uses
the standard alphabet, in which case the programming is mostly in the
utilities that follow)), with a Lojban word-list spell checker, one which
would call up the place structure on a mouse click, or give the breakdown
of a legal but undefined lujvo with a different mouse click. And then
finally you could invoke the parser in a pop-up window for any selected
chunk of text.
lojbab
----
lojbab ***NOTE NEW ADDRESS*** loj...@lojban.org
Bob LeChevalier, President, The Logical Language Group, Inc.
2904 Beau Lane, Fairfax VA 22031-1303 USA 703-385-0273
Artificial language Loglan/Lojban:
see Lojban WWW Server: href=" http://xiron.pc.helsinki.fi/lojban/ "
Order _The Complete Lojban Language_ - see our Web pages or ask me.
David Brookshire Conner
ma...@kli.org writes:
> I *like* that idea. I guess I'm surprised I didn't already write a major
> mode in emacs. Then again, I also have to keep reminding myself what
> people are looking for. I still expect word-processors to be just really
> really good typewriter/typesetter machines, not popping up templates,
> word-completions (unless you ask), stuff like that. But I'll wager that
> just an as-you-type cmene-checker would get a workout!
Oh yeah - there will be some people that want an emacs mode just for
the chuckle value - "Look! I can get Zippy to talk Lojban! Now let's
psychoanalyze it!" (Emacs arcana - if you don't get it, don't worry.)
> >An initial Framemaker extension would consist simply of a document
> >template with lots of smart "paragraph" styles (with each "paragraph"
> >in Framemaker being a lojban bridi). For example:
> >
> >StartUtterance - first sentence, following "paragraph" is
> >ContinueUtterance - "autonumbered" to start with ".i "
> >StartParagraph - autostarts with .ni'o, next para is
> > continueUtterance.
> >etc. formatting makes all this look pretty.
> Maybe indents for .i?
My first try on this (since writing the email earlier today) assumes
each bridi has it's own "paragraph". The first bridi in a .ni'o
paragraph has no automatic text, and is a "run-in" heading, which
means the next para starts on the same line this one finishes on. You
use a new .ni'o style for a .ni'o paragraph, which starts on its own
line. The ".ni'o" is inserted automatically, and the ".ni'o" hangs
into the left margin, which makes paragraphs easy to spot. It's also
consistent with ".ni'oni'o" and such.
So text looks like this:
mi vecnu ta la djan. .i la djan. mi i vecnu .i mi tavla la
meris.
.ni'o la lojban. melbi mi
.ni'o la zo'is. melbi mi .i la zo'is. mi melbi vecnu vau
It might be nice to automatically decorate the .ni'o and the .i (e.g.,
in bold, or a slightly larger font, or something) to make them a
little more obvious. Of course, a lojban font would handle this (see
the other thread).
> > > or give the breakdown
> > > of a legal but undefined lujvo with a different mouse click.
> >
> >and a database of the rafsi....
>
> YES. For building and reading.
Sure - tab completion! This *is* emacs we're talking about!
> And one of cmavo?
Yep. If it is syntax-aware, of course, it should only suggest legal
cmavo for wherever point is. And with tab-completion again.
Brook
---------
"If you can't make it good, at least make it look good."
-- Bill Gates on the solid code base of Win9X
ma...@kli.org
>From: sd...@csis.hku.hk (Lee Sau Dan ~{@nJX6X~})
>Date: 18 Sep 1999 04:00:19 +0800
>
>>>>>> "David" == David Brookshire Conner <nell...@concentric.net> writes:
>
> David> My first try on this (since writing the email earlier
> David> today) assumes each bridi has it's own "paragraph". The
> David> first bridi in a .ni'o paragraph has no automatic text, and
> David> is a "run-in" heading, which means the next para starts on
> David> the same line this one finishes on. You use a new .ni'o
> David> style for a .ni'o paragraph, which starts on its own
> David> line. The ".ni'o" is inserted automatically, and the
> David> ".ni'o" hangs into the left margin, which makes paragraphs
> David> easy to spot. It's also consistent with ".ni'oni'o" and
> David> such.
>
>Rather than having the compute suddenly spit out some text, I would
>prefer that it automatically recognizes that I'm starting a new
>paragraph when I type ".ni'o" and hence does the right thing for me
>(e.g. hangs ".ni'o" to the left, ".ni'oni'o" further left, etc.)
>
>Well... this is just my personal taste.
I'm with you. When I programmed in Pascal (it was for school, sorry) I
always used to disable Pascal-abbrev-mode because I didn't like the machine
trying to complete my code for me. Maybe I don't WANT begin/end pairs. Or
whatever. Or maybe I just like typing everything in.
>Has anyone of you brought up syntax-highlighting using font-lock-mode?
>It'd be nice to have the gismu, rafsi and cmavo displayed in different
>colors. Better still, different categories of cmavo can be displayed
>in their own colors, with open cmavo have the same color as the
>closing ones (such as "le" and "ku").
In fact, paren-matching would be a HANDY thing (and more
device-independent) for terminators.
~mark
Piermaria Maraziti
> Any of you programmers out there willing to try to come up with a simple
> Lojban word-processor? I am thinking of something for Windows, but of
> course there will be a bunch who would rather write it for Unix.
I wish to do something similar.
With a dictionary online and a parser that could suggest the continuation
of a sentence and another that can help you translating from lojban
interlinearizing with translations and making the syntactical tree.
But I have no time, as for now, to do it... :-(
Perhaps I can support someone who wishes to try.
Hmm, apart of that, I strongly suggest to use Java to do it: this for
three reasons: multiplatform support (win 95 and unix version comes easy),
a strong and very good language, and the program could also be put in an
applet, on the internet, with little effort.
Ciao!
-------------------------------------------------------------------------
Piermaria Maraziti - pier...@maraziti.it - http://piermaria.maraziti.it
ait anuas [Ex Arcano] - ainulindale: - Discordia l'Eterno - +3934735GILDA
http://gilda.it http://www.pathos.it http://discussioni.org ICQ:744473
Gran Siniscalco del Leale Ordine della Cavalleria et Stregoneria Italica
ma...@kli.org
>ma...@kli.org writes:
> > I *like* that idea. I guess I'm surprised I didn't already write a major
> > mode in emacs. Then again, I also have to keep reminding myself what
> > people are looking for. I still expect word-processors to be just really
> > really good typewriter/typesetter machines, not popping up templates,
> > word-completions (unless you ask), stuff like that. But I'll wager that
> > just an as-you-type cmene-checker would get a workout!
>
>Oh yeah - there will be some people that want an emacs mode just for
>the chuckle value - "Look! I can get Zippy to talk Lojban! Now let's
>psychoanalyze it!" (Emacs arcana - if you don't get it, don't worry.)
Hee... Considering people STILL screw up cmene construction, I know it will
be useful. Ask Lojbab what a pain in the nether regions I was about
knocking down people's faulty cmene.
> > >An initial Framemaker extension would consist simply of a document
> > >template with lots of smart "paragraph" styles (with each "paragraph"
> > >in Framemaker being a lojban bridi). For example:
> > >
> > >StartUtterance - first sentence, following "paragraph" is
> > >ContinueUtterance - "autonumbered" to start with ".i "
> > >StartParagraph - autostarts with .ni'o, next para is
> > > continueUtterance.
> > >etc. formatting makes all this look pretty.
>
> > Maybe indents for .i?
>
>My first try on this (since writing the email earlier today) assumes
>each bridi has it's own "paragraph". The first bridi in a .ni'o
>paragraph has no automatic text, and is a "run-in" heading, which
>means the next para starts on the same line this one finishes on. You
>use a new .ni'o style for a .ni'o paragraph, which starts on its own
>line. The ".ni'o" is inserted automatically, and the ".ni'o" hangs
>into the left margin, which makes paragraphs easy to spot. It's also
>consistent with ".ni'oni'o" and such.
>
>So text looks like this:
>
> mi vecnu ta la djan. .i la djan. mi i vecnu .i mi tavla la
> meris.
>
>.ni'o la lojban. melbi mi
>
>.ni'o la zo'is. melbi mi .i la zo'is. mi melbi vecnu vau
I've started these days doing stuff more like:
ni'o
le broda cu brode .i mi brode le'i brodo gi'e brodi .ije do .iu se panci
ni'o
mi do citka
i.e. line-break after the ni'o, no line-breaks between jufra, but
whitespace between paragraphs (i.e. before ni'o, but not after). See
http://www.kli.org/kli/langs/KLIlojban.html for an example.
Indentation helps in reading, but I'm starting to think it may be unwieldy
in practice. Lots of major-league hanging indents (and worse, nesting)
mean a lot of wasted space.
>It might be nice to automatically decorate the .ni'o and the .i (e.g.,
>in bold, or a slightly larger font, or something) to make them a
>little more obvious. Of course, a lojban font would handle this (see
>the other thread).
Note that in Courier-ish fonts, {.i} is pretty marked already. That period
helps.
> > > > or give the breakdown
> > > > of a legal but undefined lujvo with a different mouse click.
> > >
> > >and a database of the rafsi....
> >
> > YES. For building and reading.
>
>Sure - tab completion! This *is* emacs we're talking about!
You bet.
> > And one of cmavo?
>
>Yep. If it is syntax-aware, of course, it should only suggest legal
>cmavo for wherever point is. And with tab-completion again.
Oooh, clever.
~mark