Tangent the second: ASCII
Christopher Doty
Michael Everson
> I'm wondering how closely tied people are to the idea that, if a hypothetical thing happened where Lojban got, say, tone, it should stick with ASCII-only characters?
It already doesn't. People use the curly apostrophe for the apostrophe; it is outside of ASCII. Specifications discuss specifically the use of guillemets; these are outside of ASCII. Specifications also discuss specifically the use of á é í ó ú for anomalous stress; these are outside of ASCII.
Michael.
Ivo Doko
> You received this message because you are subscribed to the Google Groups
> "lojban" group.
> To post to this group, send email to loj...@googlegroups.com.
> To unsubscribe from this group, send email to
> lojban+un...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/lojban?hl=en.
>
Well in my humble opinion there's absolutely no downside in staying
within ASCII *if it is feasible*. If, however, it isn't, then
obviously the only way to go is Unicode. But as long as it's possible
to stay within the albeit a bit tight confines of ASCII I don't see a
reason for intentionally inventing things for the sole purpose of
having to step outside it.
Alan Post
> On 7 Apr 2010, at 18:45, Christopher Doty wrote:
>
> > I'm wondering how closely tied people are to the idea that, if a hypothetical thing happened where Lojban got, say, tone, it should stick with ASCII-only characters?
>
>
I'm undecided on this question, as I really benefit from dealing
with ASCII, and in general prefer restricted character sets for
the benefit of maximum interoperability.
5 years ago, I would have really opposed non-ASCII characters,
whereas today I don't hold nearly so strong a view. I can surmise
from this that my future self will more likely tend toward deciding
to support non-ASCII characters.
I'm particularly drawn to the aesthetic of � � � � � to mark stress,
rather than capital letters. I like this because I'm more
confortable with diacritic marks than I am with the idea of a letter
having two distinct written forms. On that particular example alone
I get interested in non-ASCII characters.
So I would say I'm not particularly 'closely tied to the idea of
Lojban sticking with ASCII-only characters.'
-Alan
--
te djuno lo do sevzi
Michael Everson
> I'm particularly drawn to the aesthetic of á é í ó ú to mark stress,
> rather than capital letters. I like this because I'm more
> confortable with diacritic marks than I am with the idea of a letter
> having two distinct written forms.
For one thing, they're portable. You can search for á vs a more easily in many environments than you can for A vs a.
Michael
Craig Daniel
>
> I'm particularly drawn to the aesthetic of á é í ó ú to mark stress,
> rather than capital letters. I like this because I'm more
> confortable with diacritic marks than I am with the idea of a letter
> having two distinct written forms. On that particular example alone
> I get interested in non-ASCII characters.
I use stress marked by accents (when I must mark it at all, which is
seldom) in handwriting.
Am I correct in remembering that even in standard Lojban orthography
it is not invalid to mark stress on only the vowel, rather than the
consonants of the syllable as well?
If so, I suggest that the two approaches can be merged harmoniously.
All it takes is the creation of Lojbanic fonts wherein the accented
vowels (and accented versions of the consonants able to be syllabic)
are the "capital" forms. From the perspective of a computer, this
means staying in the ASCII realm; it also means that although the font
isn't the one you're used to using, the characters used within that
font are the standard ones and you're not actually deviating from the
ASCII orthography - just displaying it differently to anyone who
happens to have the right font. For everyone who likes things to look
the way they do with capitalization, they just don't install the font,
and they see your (still correct) Lojban rendered the way they're used
to.
Anyone know how to use their favorite font editor? If so, hack up your
favorite open-sourced font that's suitable for body text, distribute
it, and because writing with that assumption doesn't actually break
the standards (unless I'm wrong about the vowel thing, in which case
you want to make the consonants have identical capital and lowercase
forms for the same effect and then you'll have things break whenever
you want to accent a syllabic consonant) it requires no validation
from the LLG or the BPFK - just people installing a new font for
reading Lojban in, and that only if they feel like it.
- mi'e .kreig.daniyl.
komfo,amonan
But in Lojban there's no difference between a word with 'A' and a word with 'a'. They're always the same word. So case-insensitive search finds both versions.
mu'o mi'e komfo,amonan
Timo Paulssen
> For one thing, they're portable. You can search for � vs a more easily in many environments than you can for A vs a.
>
> Michael
>
That may be, but if you have a text that has "mi n�lci lo br�vla ki'u lo
du'u mi g�rku" and you search for "gerku" - will your search algo find
it? or treat it as a non-match?
Will you find "gerku" in "mi nElci lo brIvla ki'u lo du'u mi gErku"? I
would think so. Also, the stress is basically only ever interesting if
you have spaceless text, but then you have considerably more problems
(for example the huge lynch mob in front of your gate).
Now, i agree, that � looks a bit nicer, but this argument you brought
forth is not a good one.
- Timo
Michael Everson
> Now, i agree, that á looks a bit nicer, but this argument you brought
> forth is not a good one.
The other argument is that á can't be "accidentally" lost in an all-upper-case or all-lower-case operation.
:-)
Craig Daniel
<timo...@perpetuum-immobile.de> wrote:
> That may be, but if you have a text that has "mi nélci lo brívla ki'u lo
> du'u mi gérku" and you search for "gerku" - will your search algo find
> it? or treat it as a non-match?
> Will you find "gerku" in "mi nElci lo brIvla ki'u lo du'u mi gErku"? I
> would think so. Also, the stress is basically only ever interesting if
> you have spaceless text, but then you have considerably more problems
> (for example the huge lynch mob in front of your gate).
> forth is not a good one.
> - Timo
There's one place in non-spaceless text where you need stress: names.
That's the only place I ever bother marking it, or have seen others
doing so outside of examples of talking about Lojban morphology and
things like the tosmabru test, since nobody actually writes spaceless
text.
- mi'e .kreig.daniyl.
komfo,amonan
That may be, but if you have a text that has "mi nélci lo brívla ki'u lo
du'u mi gérku" and you search for "gerku" - will your search algo find
it? or treat it as a non-match?
to ta'a u'i mi mutce co nelci ta noi ke'a bridi toi .i mu'o mi'e komfn
Jonathan Jones
On Wed, Apr 7, 2010 at 2:43 PM, Timo PaulssenThere's one place in non-spaceless text where you need stress: names.
<timo...@perpetuum-immobile.de> wrote:
> That may be, but if you have a text that has "mi nélci lo brívla ki'u lo
> du'u mi gérku" and you search for "gerku" - will your search algo find
> it? or treat it as a non-match?
> Will you find "gerku" in "mi nElci lo brIvla ki'u lo du'u mi gErku"? I
> would think so. Also, the stress is basically only ever interesting if
> you have spaceless text, but then you have considerably more problems
> (for example the huge lynch mob in front of your gate).
> Now, i agree, that á looks a bit nicer, but this argument you brought
> forth is not a good one.
> - Timo
No, that's the only place you need it in standard text, and even then, only on non-standard stressed names.
In spaceless text, you have to mark stress on all brivla:
locribecutAvlami
That's the only place I ever bother marking it, or have seen others
doing so outside of examples of talking about Lojban morphology and
things like the tosmabru test, since nobody actually writes spaceless
text.
- mi'e .kreig.daniyl.
--
You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to loj...@googlegroups.com.
To unsubscribe from this group, send email to lojban+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.
--
mu'o mi'e .aionys.
.i.a'o.e'e ko klama le bende pe denpa bu
tijlan
There's actually one I tweaked a while ago (attached to this mail).
The glyphs are based on some supposedly free font I downloaded from
somewhere, but it looks identical to Verdana, so I'm not too sure
whether it's truly open-source.
I added the acute to every consonant in the capital section as well,
so you can type "LERfu" (if that's your preference) and get each
letter in that syllable marked.
mu'o mi'e tijlan
Jorge Llambías
>
> There's one place in non-spaceless text where you need stress: names.
It depends what we understand by "need".
"mu'a" and "mu'á" are the same word. You "need" the stress mark if for
some reason you feel it's important to mark the valid non-standard
pronunciation of that word.
"cinri" and "ci,n,ri" are the same word. You "need" the comma if for
some reason you feel it's important to mark the valid non-standard
pronunciation of that word.
"marian" and "mari,án" are the same word. You "need" the stress mark
and the comma if for some reason you feel it's important to mark the
valid non-standard pronunciation of that word.
Many people feel that valid non-standard pronunciations of cmevla are
more worth marking than valid non-standard pronunciations of cmavo or
brivla, possibly because of respect for the "owner" of the name. But
then again, some names have brivla forms, so if someone liked to be
called "la ci,n,ri" we would "need" the commas there too. And names
can include cmavo, so if someone liked to be called "la mu'á zei
mu'á", we would "need" the stress mark for those cmavo too.
I would say stress marks are never strictly "needed". they are just a
provided convenience to indicate preferred pronunciations, just like
the comma.
mu'o mi'e xorxes
Oleksii Melnyk
I'm wondering how closely tied people are to the idea that, if a hypothetical thing happened where Lojban got, say, tone, it should stick with ASCII-only characters?
Our keyboards already have too much buttons for 30 symbols, lojban needs and for the 10 fingers, most of us have. So, éntering a fúnny cháracters will require a lot of extra finger travel, therefore should be avoided, where possible. Unless we will get a better input devices or really urgent needs.
--
mu'o mi'e lex
Felix
Oleksii Melnyk <lame...@gmail.com> wrote:
> 2010/4/7 Christopher Doty <suomi...@gmail.com>
>
> Our keyboards already have too much buttons for 30 symbols, lojban needs and
> for the 10 fingers, most of us have. So, éntering a fúnny cháracters will
> require a lot of extra finger travel, therefore should be avoided, where
> possible. Unless we will get a better input devices or really urgent needs.
> ea
> --
> mu'o mi'e lex
>
>
Almost every keyboard outside of US English has simple ways to access accented keys, either directly or via a dead key, which causes as little effort as using Shift for typing a capital.
For fiddlers, Microsoft offers a free tool here:
http://msdn.microsoft.com/en-us/goglobal/bb964665.aspx
Linux users would know how to do that anyway. So, really no big deal i.m.o.
Muho mihe FelikS
Jonathan Jones
On Thu, 8 Apr 2010 09:29:45 +0300
Oleksii Melnyk <lame...@gmail.com> wrote:
> 2010/4/7 Christopher Doty <suomi...@gmail.com>
>
> Our keyboards already have too much buttons for 30 symbols, lojban needs and> ea
> for the 10 fingers, most of us have. So, éntering a fúnny cháracters will
> require a lot of extra finger travel, therefore should be avoided, where
> possible. Unless we will get a better input devices or really urgent needs.
> --Almost every keyboard outside of US English has simple ways to access accented keys, either directly or via a dead key, which causes as little effort as using Shift for typing a capital.
> mu'o mi'e lex
>
>
Just more proof that the U.S. as a whole is a country of conceited bastards. ;) (I have a standard U.S. keyboard- Linux lets me use the Caps Lock as a dead key for just that purpose- and while I don't remember the name of it, there's a Win program that let's me do that as well.)
For fiddlers, Microsoft offers a free tool here:
http://msdn.microsoft.com/en-us/goglobal/bb964665.aspx
Linux users would know how to do that anyway. So, really no big deal i.m.o.
Muho mihe FelikS
--
You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to loj...@googlegroups.com.
To unsubscribe from this group, send email to lojban+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.
--
Bob LeChevalier
> On 7 Apr 2010, at 18:45, Christopher Doty wrote:
>>I'm wondering how closely tied people are to the idea that, if a hypothetical thing happened where Lojban got, say, tone, it should stick with ASCII-only characters?
ASCII-only is part of the baselined language, and I am unsure of what
hypothetical conditions might arise that would lead to tone, but I won't
be around for it.
> It already doesn't. People use the curly apostrophe for the
> apostrophe; it is outside of ASCII. Specifications discuss
> specifically the use of guillemets; these are outside of ASCII.
> Specifications also discuss specifically the use of á é í ó ú for
> anomalous stress; these are outside of ASCII.
Can you please cite where "specifications discuss" this. I'm not sure
what you are thinking of.
The only specification that discusses orthography is CLL (chapter 3).
It specifies ASCII characters, and capitalization of an entire syllable
only for non-standard stress in Lojbanized names. There is no mention
of diacritics (except as how they might be described using lerfu) or
guillemets (I've never even seen the latter word before, except as the
surname of one line of my French ancestors).
An alternate orthography is offered as part of the effort at
rapprochement with JCB's organization, and there is discussion of
Cyrillic and Tengwar as alternatives (but so far as I know, no one has
ever seriously considered using either of them).
As for the "curly apostrophe". I have no idea what other people use, but
if it isn't on my keyboard, I probably will never use it, and indeed
have no idea HOW to use it.
lojbab
Craig Daniel
>
> An alternate orthography is offered as part of the effort at rapprochement
> with JCB's organization, and there is discussion of Cyrillic and Tengwar as
> alternatives (but so far as I know, no one has ever seriously considered
> using either of them).
>
I used to favor Cyrillic for Lojban handwriting, because I found my
letterforms were more distinct from one another that way and I find I
can read Lojban in both about equally well. My normal handwriting has
since improved, so I don't use Cyrillic anymore. I wouldn't say that's
like seriously considering Cyrillic for anything other than personal
use, but it's not quite zero either.
Still, if my notes from that period were found by another jbopre, I
wouldn't expect them to be readable at a glance even by somebody
fluent; ergo, Cyrillic is not de facto a standard form of Lojban
writing.
Michael Everson
> It already doesn't. People use the curly apostrophe for the apostrophe; it is outside of ASCII.
The curly apostrophe is used in the Alice PDF and in "What is Lojban?" passim.
> Specifications discuss specifically the use of guillemets; these are outside of ASCII.
"What is Lojban?" p. 67.
>> Specifications also discuss specifically the use of á é í ó ú for anomalous stress; these are outside of ASCII.
I don't know if this is actually written down; it does not appear to be in "What is Lojban?".
> The only specification that discusses orthography is CLL (chapter 3). It specifies ASCII characters, and capitalization of an entire syllable only for non-standard stress in Lojbanized names. There is no mention of diacritics (except as how they might be described using lerfu) or guillemets (I've never even seen the latter word before, except as the surname of one line of my French ancestors).
"What is Lojban?" is a specification.
> As for the "curly apostrophe". I have no idea what other people use, but if it isn't on my keyboard, I probably will never use it, and indeed have no idea HOW to use it.
Please look up "smart quotes"; I guarantee that software you use offers this to you.
Michael.
Michael Everson
> Still, if my notes from that period were found by another jbopre, I wouldn't expect them to be readable at a glance even by somebody fluent; ergo, Cyrillic is not de facto a standard form of Lojban writing.
"Spaceless" Lojban is surely illegible "at a glance".
Michael
Craig Daniel
And, while technically valid, it's not part of the language as anybody
uses it. (Outside of examples about how breaking it up into normal
Lojban is done, that is.) Note the phrase "de facto" in there -
spaceless is the same way, since it's valid but not part of the
standards as actually used by actual Lojbanists.
Oleksii Melnyk
Almost every keyboard outside of US English has simple ways to access accented keys, either directly or via a dead key, which causes as little effort as using Shift for typing a capital.
IMO that "little effort" is still about 3 times as hard(in X11 "compose" mode, others may, probably, reach 2.x factor) as entering an "ordinary" letters.
Minimiscience
No, it's not. It's an introduction to Lojban for use in learning the language.
It is supposed to be a concise reflection of the fundamental parts of the CLL,
not an addendum to it. The only formal specifications of Lojban are *The
Complete Lojban Language* and (I think) the official {gismu} & {cmavo} lists in
the plain text Logflash file format.
mu'omi'e .kamymecraijun.
--
li'a .e'i ca vondei .i mi na'e pu'i kufra loi vondei
Michael Everson
> de'i li 08 pi'e 04 pi'e 2010 la'o fy. Michael Everson .fy. cusku zoi skamyxatra.
>> "What is Lojban?" is a specification.
> .skamyxatra
> No, it's not. It's an introduction to Lojban for use in learning the language.
> It is supposed to be a concise reflection of the fundamental parts of the CLL,
> not an addendum to it.
> The only formal specifications of Lojban are *The Complete Lojban Language* and (I think) the official {gismu} & {cmavo} lists in the plain text Logflash file format.
Michael
Jim Carter
> I'm wondering how closely tied people are to the idea that, if a
> hypothetical thing happened where Lojban got, say, tone, it should stick
> with ASCII-only characters? I realize there was a good reason for this in
> the past, but I have trouble seeing a reasoning for it now. Just wondering
> how people feel about that now that Unicode has become the standard? ASCII
> is synonymous with "old-fashioned computing" to me, so just curious.
I've had no end of trouble getting my desktop environment to load a compose
table that includes the accented characters I use and provides them to all
my applications. Thus I would prefer to minimize the use of non-ASCII
characters. But I like the aesthetics of the accents.
Someone else suggested a font in which upper case glyphs were similar to
lower case but with an accent. That sounds very nice for print media, but
it would take some work, and meta-tags in the documents, to make the font
appear on the fly for interspersed Lojban and non-Lojban messages or
documents. Charset:X-Lojban?
James F. Carter Voice 310 825 2897 FAX 310 206 6673
UCLA-Mathnet; 6115 MSA; 405 Hilgard Ave.; Los Angeles, CA, USA 90095-1555
Email: ji...@math.ucla.edu http://www.math.ucla.edu/~jimc (q.v. for PGP key)
Jonathan Jones
The CLL uses a different font for Lojban than for English. For example:
Chapter 6, section 1:
1.1) mi klama le zarciIf you wanted to write Lojban and some other language in the same document, I would suggest using that practice.
I go-to the market
That said, while I am not *against* using diacritic marking instead of capitals, that is pretty much the extent of my non-ASCII in Lojban acceptance.
--
You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to loj...@googlegroups.com.
To unsubscribe from this group, send email to lojban+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.
--
Craig Daniel
>
> That said, while I am not *against* using diacritic marking instead of
> capitals, that is pretty much the extent of my non-ASCII in Lojban
> acceptance.
>
I like it for handwriting, but where computers are involved I think
there are good reasons to stick to ASCII. Not because that restrictive
a set is required, but because keeping it there absent any compelling
reason not to (and Lojban, unlike some languages, has no need to do
otherwise) is good. But I also think capital letters in the middle of
words look ugly, and I think they draw your eye to them more than
accent marks do and can turn into feeling like that word is being
emphasized when it's not instead of like that word has odd stress
specified, so I'm also in favor of indicating stress by means of
diacritics. I don't like the idea of changing the language spec over
it, or of giving up on ASCII-only and making people have to edit all
software that cares, but in a vacuum I prefer it and if the language
were being created in this era of Unicode I'd be a strong proponent of
that as the standard instead of capitalization.
The reason I suggested an alternate font is that it lets you have your
cake and eat it too. The letters are all ASCII, no matter how you're
stressing the words, so for computers it represents no change in the
standards and it captures all the benefits of the restricted character
set. The computer doesn't care how the font you're using renders the
characters, just what character codes they map to; you could make
every glyph look like a dot in slightly different positions and while
it would be totally unreadable to humans it would be the same thing
from a machine's perspective. Ergo, using the diacritics-as-capitals
font means that from a machine-processing perspective all your Lojban
is the same as it's always been, and it's all ASCII-only. But for
humans, it shows up as the alternate convention, which I personally
consider vastly better from an aesthetic perspective and which I
suspect is also more readable. It's the best of both worlds.
It also means that how it appears depends on the reader's preferences
rather than the writer's, since it means the two are just different
renderings of the same underlying standard. If the writer specs the
text as being Lojban or says that's the font it's in and capitalizes
only the vowel in the stress-specified syllable, then if the reader is
looking at it with a stylesheet that says Lojban is in that font and
has that font, it shows up with an accent on the vowel. If the reader
doesn't have the font installed, it gets rendered as whatever font
their software defaults to, and they see no change. So for those who
favor kaptalo over diacritics, it's trivial to keep it that way - just
don't install the font. Lojban will look like it always has. If you do
like diacritics, you have to install a new font to see them since
they're being sent to your machine as ASCII characters for uppercase
letters rather than lowercase letters with accents on them, but it's
not like it's hard to add new fonts.
I propose agreeing on a font modified from something verifiably open
(or built from scratch, but why bother?) rather than something that
might actually just be Verdana in a funny hat (mostly because I'd like
to see the Lojbanized version be open-source, and nobody has the right
to license it thus if it is derived from Verdana and we didn't realize
it); the first easy-to-read font of known provenance to come along
would make a good choice of standard. I'm happy to take a crack at
putting one together, but my life is fairly busy at present (between
part-time school and part-time work that add up to more than a
full-time commitment, plus Lojban not getting top billing among my
hobby projects right now) and I'd have to learn the basics of using a
font editor. The latter is something I kinda want to try anyhow, so at
some point I'll get to work on that, but it'll be a while and if
somebody else who actually knows what they're doing beats me to it, so
much the better.
Once we've got one distributed, I'd like Lojban web stuff to serve
Lojban text in that font wherever possible. (Maybe make the wiki have
a "Lojban text" tag of some sort that defaults to that?) Then adopt an
unofficial community standard that only the nucleus of a
stress-specified syllable should be capitalized; the other way is
still valid, of course (I don't see this as an issue it's worth
changing any part of the baselined language over!), but the
one-capital-per-syllable approach would be preferred. I personally
like that better anyhow, since syllabification varies a little from
person to person when consonant clusters are involved and the correct
way to specify it if you care is with slaka.bu rather than changes in
capitalization. That's an insignificant thing now, but if a lot of
people are looking at it in a diacritic font it's mildly important
since it makes the text not look like ass. The end result of this is
that folks who like seeing diacritics get Lojban that looks prettier
to them, with no change to the orthographic specifications of the
language.
- mi'e .kreig.daniyl.
who also really will get around to scanning exemplars of his Lojban
calligraphic hand and sharing his other musings on Lojbanic writing
sooner or later
Bob LeChevalier
> de'i li 08 pi'e 04 pi'e 2010 la'o fy. Michael Everson .fy. cusku zoi skamyxatra.
>
>>"What is Lojban?" is a specification.
>
> .skamyxatra
>
> No, it's not. It's an introduction to Lojban for use in learning the language.
> It is supposed to be a concise reflection of the fundamental parts of the CLL,
It's not even that. It was written as a book-form replacement for our
short brochure and language description that we used to send out to
people inquiring about the language. It is intended for new people
trying to get an idea of how things work. Most of it was written long
before CLL, at a time when there were no alternate orthographies at all.
> not an addendum to it. The only formal specifications of Lojban are *The
> Complete Lojban Language* and (I think) the official {gismu} & {cmavo} lists in
> the plain text Logflash file format.
Yes.
lojbab
MorphemeAddict
I'm wondering how closely tied people are to the idea that, if a hypothetical thing happened where Lojban got, say, tone, it should stick with ASCII-only characters? I realize there was a good reason for this in the past, but I have trouble seeing a reasoning for it now. Just wondering how people feel about that now that Unicode has become the standard? ASCII is synonymous with "old-fashioned computing" to me, so just curious.
(I think there are advantages both staying with ASCII or going by Unicode, and am not advocating for a change or anything; just curious what people think.)Chris
Ivo Doko
I believe that's why there are many different keyboard layouts.
Perhaps it's time to discuss what Lojban keyboard layout should be?
Has that ever been discussed or not because there was no need to
discuss it since all the used characters were ASCII characters, which
every keyboard layout has (should have)?
Bob LeChevalier
> I believe that's why there are many different keyboard layouts.
>
> Perhaps it's time to discuss what Lojban keyboard layout should be?
>
> Has that ever been discussed or not because there was no need to
> discuss it since all the used characters were ASCII characters, which
> every keyboard layout has (should have)?
We chose ASCII characters because of that reason.
There was never much discussion on orthography back then, except on ' vs
h, which still comes up once in a while.
In general, those who wanted something other than ASCII tended to want
something completely different, which is why Tenguar Lojban orthography
made it into CLL (it was the very first thread on Lojban List in 1988).
lojbab