View source for problems with Non-Lojban Quotes

<code>[12:38] <vensa> {la'oi} is super-used, but IMO b/c we're on IRC

[12:39] <Twey> {la'oi}'s semantics are not consistent, so even though it's used a lot, I wouldn't consider it anything more than slang

[12:39] <vensa> what do you mean "inconsisnet" Twey?

[12:40] <Twey> vensa: I'd swear your typing is getting worse :þ

[12:40] <Twey> Did you break a finger?

[12:40] <Twey> Erm

[12:40] <vensa> na go'i .iku'i mi ckana zvati .u'u

[12:40] <Twey> Right, ‘la'oi’ doesn't always indicate where it stops

[12:40] <lindar> selpa`i: I don't think we have a rating system, but if you do, let us know.

[12:41] <Twey> Or the indication may be different in speech than in text

[12:41] <vensa> Isn't it defined (in speech) to be "between two pauses"?

[12:41] <selpa`i> lindar: oO

[12:41] <lindar> Throw it into the mixing pot of ideas for jvs2. Not sure where those suggestions go.

[12:41] <Twey> vensa: Yes, but that's not the same as how it's defined in text (between spaces)

[12:41] <vensa> ki'a jvs2?

[12:41] <Twey> Depending on the contained script, there may be a pause indicated, but no space

[12:42] <vensa> Twey: you mean like {la'oi ra'anana} where ' is a glottal stop?

[12:42] <Twey> So the interpretation is different depending on whether it's written or spoken

[12:42] <Twey> Yes

[12:42] <lindar> >_>

[12:42] <lindar> vensa, JboVlaSte

[12:42] <vensa> o.k. so I think it should be redefined in writing "space or "speech-pause-indicating-character(s)"

[12:42] <vensa> ki'e lindar

[12:43] <vensa> of course no parser will be able to parse it

[12:43] <vensa> but no parser is parsing it now either :P

[12:43] <Twey> vensa: But what exactly those are depends on the interpretation of the script within it

[12:43] <Twey> vensa: Which is unknown

[12:43] <vensa> uhmm

[12:43] <Twey> Yeah, it's unparseable and always will be

[12:43] <vensa> I think the script IS know IN the context

[12:44] <Twey> Not necessarily

[12:44] <Twey> selpa`i: In general

[12:44] <vensa> selpa'i: depends who uses it

[12:44] <Twey> Unless there are good reasons it shouldn't (like la'oi)

[12:44] <vensa> if it's only lindar then: np :P

[12:44] <vensa> *no

[12:44] <Twey> Heheh.

[12:45] <vensa> Twey: how about redefine textual-la'oi so that it MUST use ONLY lojban-phonetic characters

[12:45] <Twey> vensa: Then it would just be ‘la’ :þ

[12:45] <vensa> not necessarily

[12:46] <vensa> cmevla must end in consonant or else they are understood as brivla

[12:46] <vensa> also, invalid consonant pairs will be allowed

[12:46] <vensa> and more diphtongs

[12:47] <vensa> I think that {la'oi} is used mainly in IRC for the purpose of "lighting up" someones name

[12:47] <vensa> if we all had lojbo names, there wouldnt be a problem

[12:47] <vensa> it's just the newbies that make it hard

[12:48] <vensa> basically, I see very little need for {la'oi} is speech anyway

[12:48] <vensa> (in

[12:48] <vensa> zo'oi however is more useful in speech too

[12:48] <vensa> and has the same "textual" problems

[12:49] <vensa> so, applying my "only lojban characters" to {zo'oi} might fix it

[12:49] <vensa> {la'oi} we can throw out

[12:52] <vensa> sorry for the monologue. now is where you either say I'm stupid or right

[12:52] <lindar> >_>

[12:52]  * vensa senses lindar is going for "stupid"

[12:52] <vensa> (he's just thinking of a good insult) :)

[12:52] <lindar> valsi zo'oi

[12:52] <valsi> zo'oi = quote next non-Lojban word only; quotes a single non-Lojban word delimited by pauses (in speech) or whitespace (in writing)

[12:53] <lindar> What exactly is your problem with it?

[12:53] <vensa> Twey's problem, and I agree is

[12:53] <lindar> I didn't see that Twey had a problem.

[12:53] <vensa> that the quoted word in text could have a character that isnt whitespace but still represents a glottal stop or a pause

[12:53] <vensa> e.g. {zo'oi ra'anana}

[12:54] <vensa> lindar, your big brother skills are decreasing

[12:55] <vensa> {zo'oi la'adan}

[12:56] <vensa> (not sure about that one)

[12:56] <selpa`i> Isnt that okay in writing?

[12:56] <vensa> zo'oi O'connor

[12:56] <vensa> selpa'i the problem is that lojban is supposed to have audio-visual isomorphism

[12:56] <selpa`i> Yes.

[12:56] <vensa> in speech it would break up into {zo'oi o    koner}

[12:57] <selpa`i> Right.

[12:57] <vensa> or something

[12:57] <vensa> so: problem

[12:57] <selpa`i> Well just remove zo'oi as well

[12:57] <selpa`i> :)

[12:57] <vensa> lindar: do U C

[12:58] <vensa> Twey: how is my proposal to scorch la'oi and fix zo'oi?

[12:59] <selpa`i> That fix makes it really similar to la

[12:59] <selpa`i> Only with diphtongs and stuff

[12:59] <vensa> selpa'i : I'm talking about fixing {zo'oi}. not {la'oi}

[13:00] <vensa> I'm proposing dropping {la'oi} altogether and if you want someones attention in IRC say: "EnglishName: lo lojbo jufra

[13:01] <selpa`i> Is zo'oi necessary to have?

[13:01] <selpa`i> I agree with ditching la'oi.

[13:01] <vensa> no. it's just for lazy ppl

[13:01] <selpa`i> I dont like either.

[13:01] <vensa> you can get away with {zoi gy bbdfasb gy}

[13:01] <selpa`i> Yes

[13:02] <vensa> but it could be handy in speech where you dont have to think about the lojbanization and can just utter a single non-lojban word without delimeters

[13:02] <vensa> i.e. the delimeter is a glottal stop

[13:02] <lindar> Stop proposing things.

[13:02] <selpa`i> But thats impossiblee isnt it?

[13:02] <lindar> Also, I wasn't looking because I was busy fixing real problems with Lojban.

[13:03] <vensa> xa'a'a

[13:03] <lindar> Nobody has a problem with la'oi or zo'oi or how they work, and I heard la'oi no less than 10 times on the phone last night with selkik.

[13:03] <vensa> la poi vajni ku'o lindar

[13:03] <lindar> It's helpful to figure out when a name is foreign or nonstandard jbo.

[13:04] <lindar> We're not changing a word because you don't understand how it works or why we have it.

[13:04] <vensa> well, I suppose we could still use {la'oi} to give the implication that the name is of foreign origin

[13:04] <lindar> Instead you should make some effort to understand why a word exists, rather than jump up and say we don't need it.

[13:04] <lindar> (Yes, I used to do that when I was a newbie.)

[13:04] <vensa> Twey: lindar is abusing me

[13:04] <vensa> pls put him in his place

[13:04] <lindar> =D

[13:05] <lindar> No more proposals.

[13:05] <selpa`i> Put the thing about O'Connor stands doesnt it?

[13:05] <Twey> lindar: It *doesn't* work.

[13:05] <Twey> That's the problem.

[13:05] <lindar> >_>

[13:05] <vensa> [12:39] <Twey> {la'oi}'s semantics are not consistent, so even though it's used a lot, I wouldn't consider it anything more than slang

[13:05] <Twey> A text using la'oi can be interpreted completely differently depending on whether it's read on paper or spoken

[13:05] <lindar> In what way does it not work?

[13:06] <lindar> Yeah, that's because it's non-jbo.

[13:06] <lindar> Nobody said it had to be pronounced, only parsed.

[13:06] <Twey> Or based on some unknown information possessed only by the speaker (the intended phonetic interpretation of the foreign script inside the la'oi-quote)

[13:06]  * vensa likes how newbies bring up innocent questions that turn into raged discussions. ki'e la selpa'i

[13:06] <lindar> If you have an issue with how it's parsed, that's one thing, but it's no worse in that regard than la'o.

[13:06] <selpa`i> Sry :P

[13:07] <vensa> I wasnt being sarcastic at all :)

[13:07] <vensa> hhmmmm

[13:07] <vensa> lindar has a point

[13:07] <vensa> something inside the la

[13:07] <vensa> *la'o

[13:07] <vensa> can be read as the delimiter

[13:08] <selpa`i> Stop!

[13:08] <selpa`i> lol

[13:08] <selpa`i> Not, if you are clever.

[13:08] <lindar> Seriously, I'm like a dark jedi up in this bitch. My power grows stronger, son!

[13:08] <ksion> I usually object to lindar nerdraging at everyone but this time I must reinforce his position ;P do bebna doi la vensa zo'o

[13:08] <selpa`i> la'o works-

[13:08] <vensa> no it doesnt

[13:08] <lindar> vensa: That works because "sh" is not "shuh".

[13:08] <ksion> vensa: It's up to you to get delimiter that doesn't fuck stuff up.

[13:08] <vensa> whatever, you get my point

[13:09] <lindar> Also, there's a rule that says the delimiter can't be in the quoted material.

[13:09] <selpa`i> It wokrds.

[13:09] <ksion> Similarly, it's up to you - when you use la'oi - to choose a work that doesn't have glottal stop in it.

[13:09] <selpa`i> Just choose a better delimiter.

[13:09] <ksion> If it has, use la'o. Simple.

[13:09] <vensa> ksion: it could be an innocent mistake, like the idea of {la} inside a cmevla

[13:09] <lindar> Yeah, if it has a glottal stop or is more than one word, you have to use la'o.

[13:09] <lindar> la'oi is just there for convenience.

[13:09] <ksion> And what's the difference between innocent mistake and guilty mistake? :P

[13:10] <vensa> heh

[13:10] <vensa> "the case against la" I suppose

[13:10]  * vensa likes to use things he has read to his advantage

[13:10] <lindar> >_>

[13:10] <lindar> What are you using to your advantage?

[13:10] <ksion> Nah but seriously, the fact that you can use something wrongly is poor argument. It works for {la}, maybe, but here it's pretty obvious that a glottal STOP is kinda unusual inside words.

[13:11] <vensa> the usage of dotside is backed by the idea that it is hard to avoid internal {la}

[13:11] <ksion> So if you quote a word with a stop using la'oi, you're dumb.

[13:11] <lindar> I'm going with ksion on this.

[13:11] <selpa`i> Good point. lol

[13:11] <vensa> ehhh

[13:11] <vensa> maybe

[13:11] <lindar> No, not maybe.

[13:11] <lindar> Definitely.

[13:11] <lindar> that's how it works

[13:11] <vensa> what about {cy lindar shuh be quiet cy}?

[13:11] <selpa`i> I like the "you're dumb"-part.

[13:11] <lindar> vensa: That's not allowed.

[13:11] <selpa`i> vensa: choose better delimiters

[13:12] <vensa> but couldnt it be an innocent mistake?

[13:12] <lindar> The grammar explicitly states that you can't use the delimiting word in the delimited material.

[13:12] <ksion> Lojban is not to protect the virtue of innocents :)

[13:12] <lindar> If it's an innocent mistake, it's an innocent mistake, but that's not a flaw in the grammar, that's you being retarded.

[13:12] <ksion> vensa: Noone frowns at XML standard that CDATA section cannot have]> inside of them, since it's their delimiter.

[13:13] <vensa> so, the conclusion is to change the def of {la'oi} and {zo'oi} to not allow glottal stops? (I can agree to that)

[13:13] <ksion> They already don't allow them.

[13:13] <lindar> That's already part of the bloody rule.

[13:13] <selpa`i> <3

[13:13] <vensa> where?

[13:13] <lindar> -_-

[13:13] <vensa> valsi la'oi (notes)

[13:13] <valsi> la'oi (notes) = See also {la'o}, {zo'oi}.

[13:13] <lindar> Dude, just shut up. You're not helping, you're actually distracting me.

[13:13] <ksion> u'i

[13:13] <lindar> This is a stupid proposal.

[13:13] <vensa> then dont listen

[13:13] <lindar> You're wrong on every count.

[13:14] <vensa> fine so go away

[13:14] <lindar> Seriously, stop. Go read CLL five more times.

[13:14] <vensa> I'm talking to ksion

[13:14] <lindar> Help with the byfy work if you really want to do something.

[13:14] <vensa> ksion: where in the def of la'oi does it say "no glottal stop"?

[13:14] <ksion> valsi la'oi

[13:14] <vensa> fine, goodbye lindar

[13:14] <valsi> la'oi = single-word non-Lojban name; quotes a single non-Lojban word delimited by pauses (in speech) or whitespace (in writing) and treats it as a name

[13:14] <selpa`i> vensa: It kinda says it. "pauses"

[13:14] <vensa> selpa'i: IN SPEECH

[13:14] <ksion> The word is delimited by pauses. It cannot have pauses in a word itself, since that would delimit it.

[13:14] <ksion> QED

[13:15] <vensa> an apostrophe, which usually marks a glottal stop IN WRITING isnt mentioned

[13:15] <selpa`i> right ksion

[13:15] <vensa> ksion: IN SPEECH

[13:15] <vensa> for text the only requirement is no whitespace

[13:15] <vensa> I think this deserves a clarification in the notes at least

[13:15] <selpa`i> Uhm..

[13:16] <selpa`i> I think I agree.

[13:16] <ksion> A word with apostrophe for glottal stop can be still written with la'oi but cannot be exactly pronounced. That's no problem. la'oi etc. break isomorphism *by design*.

[13:16] <vensa> wha?!

[13:16] <vensa> I think it should be disallowed

[13:16] <vensa> use {la'o} for that

[13:16] <lindar> Vensa, let's stop the bitching for a second here.

[13:16] <lindar> http://groups.google.com/group/lojban-announcements/browse_thread/thread/7bc6a77ae385603d

[13:16] <lindar> Please read this link.

[13:16] <vensa> why break the isomorphism by design?

[13:16] <lindar> Please follow these instructions as this is explicitly what Robin wants.

[13:16] <vensa> ok. Ill try

[13:17] <ksion> vensa: Of course it would be more reasonable to use la'o for such words. I'm just saying what happens if you use la'oi anyway.

[13:17] <vensa> lindar: oh I already read that

[13:17] <lindar> We will seriously consider any concerns you may have and clarify the definition.

[13:17] <vensa> what do you care what I talk about with my lojban busddies?

[13:17] <vensa> if it doesnt interest you, dont listen

[13:17] <vensa> if I get to the conclusion that it is a serious issue of content, Ill post it

[13:17] <ksion> Geez, could you children play in other sandbox? ;p

[13:18] <vensa> ksion: if you use la'oi anyway you get an ungramatical result in speech

[13:18] <ksion> Yes.

[13:18] <vensa> fine

[13:18] <vensa> so the def should state not to do stuff that could render a gramatical error

[13:18] <ksion> But only if you pronounce the word exactly according to foreign language it comes from.

[13:19] <ksion> Lojban does not cover it.

[13:19] <vensa> in any case, {la'o} and {la'oi} arent parseable

[13:19] <vensa> *machine parseable, in text

[13:19] <vensa> :(

[13:19] <ksion> Additionally, a spoken utterance is not tied to a written utterance in a anyway. You can alter the pronuncation of the la'oi-ed word if necessary to have spoken version be gendra. It's non-Lojban fragment.

[13:20] <vensa> hmmm

[13:20] <ksion> Yes they are, the word has to not contain spaces for that.

[13:20] <vensa> hmm

[13:20] <vensa> what about {la'o cy hey shuh you cy}?

[13:20] <ksion> There is a very fishy pratice of quoting URLs with la'oi. Works in text, but breaks up in speech.

[13:20] <vensa> are you saying a machine would parse it from cy to cy

[13:21] <vensa> and in speech I would have to say it differently to avoid the middle delimieter?

[13:21] <ksion> Or choose different delimiter.

[13:21] <vensa> but it's already written

[13:21] <vensa> I am reading off a book, e.g.

[13:21] <ksion> I don't care. It's not Lojban. If you want to incorporate non-Lojban into Lojban, that's your responsibility.

[13:21] <vensa> so, if I wanted to read out a url quoted with {la'oi}, that would sound fucking wierd

[13:22] <vensa> imagine somebody wrote a lojban document and {la'oi} quoted a url. now I have to read it aloud. who's responsibility is it and why do I have to get screwed?

[13:23] <ksion> Yours. The speaker's. You are speaking Lojban, you should speak correct Lojban.

[13:23] <vensa> but I'm obliged to read what he wrote

[13:23] <ksion> Analogy: Try to read 'gfywufvew4hfei0nuvereiufehwui' in English.

[13:23] <vensa> well, maybe I'd just change it to {la'o url}

[13:23] <ksion> Yes, exactly what you need to do.

[13:24] <vensa> ksion: I would read it letter by letter

[13:24] <vensa> cool

[13:24] <vensa> ok

[13:24] <ksion> You'd read it letter-by-letter, but it might have been inteded to be read as one word.

[13:24] <ksion> Note that it's not Lojban's fault at all. It's because natlangs are weird and incosintent, but we want to cooperate with them anyway.

[13:24] <vensa> fine, so the conclusion is that NOTHING needs to be changed with the defs, and just that speakers of la'o and la'oi and zo'oi need to be aware and change delimiters and pronounciation accordinglyt. yea?

[13:25] <ksion> Requires some compromises.

[13:25] <ksion> This should be added to the notes, yes.

[13:25] <vensa> agreed! :)

[13:25] <vensa> Twey?

[13:25] <ksion> Also for zoi and la'o.

[13:25] <vensa> .ie

[13:26] <ksion> Bottom line is to choose delimeters that work in both cases for la'o and zoi, and avoid words with either written space or spoken pause for la'oi and zo'oi. That's best pratice.

[13:27] <vensa> .ie

[13:27] <vensa> so always {la'o url}, not with {la'oi}

[13:27] <ksion> Note that you cannot control that by grammar, since we have no idea about the text2speech relation for foreign text.

[13:27] <ksion> Yes, that's best.

[13:28] <ksion> URL doesn't feel like one word even in text, anyway

[13:28] <vensa> yeah

[13:28] <selpa`i> {url} ?

[13:28] <ksion> I'd use {ky.} though (kibro) as {url} is not easy to pronounce.

[13:29] <vensa> ksion: what?

[13:29] <ksion> la'o ky. www.google.com ky.

[13:29] <vensa> {ky} OR {kibro}?

[13:30] <ksion> {ky.}, since it's based on {kibro}.

[13:30] <vensa> oh ok

[13:30] <selpa`i> It doesnt really matter what one uses, because it usually obvious that it is a url

[13:30] <ksion> Sure.

[13:30] <ksion> As long as it doesn't occur in the URL itself.

[13:30] <vensa> I'm afraid {ky} might be too short in some instances

[13:30] <vensa> isnt 

[13:30] <vensa> isnt ".ke" a suffix of some state?

[13:31] <ksion> la'o kibrystuzi. www.google.com kibrystuzi. u'i

[13:31] <selpa`i> Does {uburyly} work?

[13:31] <vensa> .u'i clani bacru

--------------------------------------------------------------------------------

[13:32] <ksion> selpa`i: Without pauses, yes.

[13:32] <ksion> {ubu ry. ly.} does not work since it's more than one Lojban word.</code>