History: Problems with Non-Lojban Quotes
Preview of version: 1
- «
- »
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
12:41 <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 <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 <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 <selpa`i> So is it true that when a word gets used a lot its quite likely to become official?
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
12:48 == urandom__ ~user@p548A47C1.dip.t-dialin.net has joined #lojban
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 * vensa senses lindar is going for "stupid"
12:52 <vensa> (he's just thinking of a good insult) 
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
12:57 == Arla ~Niddler@h245n5c1o253.bredband.skanova.com has joined #lojban
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
13:00 == jeyk jey@69.59.129.28 has joined #lojban
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 <selpa`i> But thats impossiblee isnt it?
13:02 <lindar> Also, I wasn't looking because I was busy fixing real problems with Lojban.13:02 == donri ~dag@unaffiliated/dagodenhall has quit Read error: Connection reset by peer
13:03 <vensa> xa'a'a
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
13:05 <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 <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 == Arla ~Niddler@h245n5c1o253.bredband.skanova.com has quit Ping timeout: 276 seconds
13:05 <Twey> A text using la'oi can be interpreted completely differently depending on whether it's read on paper or spoken
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 <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 <vensa> la'o cy lindar sh be quite cy <-- example
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 == kpreid ~kpreid@128.153.214.195 has quit Quit: Offline
13:08 <vensa> no it doesnt
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 <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 <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 <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 <selpa`i> Good point. lol
13:11 <vensa> ehhh
13:11 <vensa> maybe
13:11 <vensa> what about {cy lindar shuh be quiet cy}?
13:11 <selpa`i> I like the "you're dumb"-part.
13:11 == donri ~dag@unaffiliated/dagodenhall has joined #lojban
13:11 <selpa`i> vensa: choose better delimiters
13:12 <vensa> but couldnt it be an innocent mistake?
13:12 <ksion> Lojban is not to protect the virtue of innocents
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 <selpa`i> <3
13:13 <vensa> where?
13:13 <vensa> valsi la'oi (notes)
13:13 <valsi> la'oi (notes) = See also {la'o}, {zo'oi}.
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 <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 <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 == lindar ~lindarthe@32.174.46.157 has quit Quit: leaving
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.
13:28 == Amie ~Amie@122.172.24.85 has quit Quit: You don't wanna know.
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
----------------------------
13:31 == Amie ~Amie@122.172.24.85 has joined #lojban
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.