Lojban versions and change scripts
{CODE(wrap="1")}jbocre: 18:17 <dbrock-> someone recently proposed we change the meaning of {y'y}
jbocre: 18:17 <vensa> it was I :)
jbocre: 18:18 <dbrock-> which is another interesting idea that becomes a complete waste of time when you actually suggest it for real
jbocre: 18:18 <vensa> dbrock: y?
jbocre: 18:18 <vensa> was there not a "big rafsi reallocation"?
jbocre: 18:18 <dbrock-> might as well run for president
jbocre: 18:18 <vensa> change happens!
jbocre: 18:18 <vensa> stop being such a stick in the mud
jbocre: 18:18 <dbrock-> yeah, I guess I'm a pessimist
jbocre: 18:19 <vensa> what do you think about my version-scripting system?
jbocre: 18:19 == Arla jbocre: [email protected] has joined #lojban
jbocre: 18:19 <dbrock-> �I haven't heard anything about it
jbocre: 18:20 <vensa> my proposition is that the current lojban be dubbed "lojban v1.0" and the lojban after the bpfk is done will be "lojban v2.0"
jbocre: 18:20 <vensa> lojban could keep being developed, and CHANGED, according to usage and what proves most useful as we gain more speakers and speaking experience
jbocre: 18:21 <UukGoblin> vensa, and pre-xorlo would be what? 0.1-beta3? ;-)
jbocre: 18:21 <vensa> the obvious problem with this is that documents written in "lojban 1.0" may have a totally different meaning under "lojban 3.0"
jbocre: 18:21 <vensa> uuk: I wanted to say that . yes.
jbocre: 18:22 <vensa> that is where the "script" idea comes in:
jbocre: 18:22 <UukGoblin> vensa, I have a solution
jbocre: 18:22 <vensa> every new standard version (which would only be released by the BPFK of course) will also have to come with a set of "scripts" or "algorithms" for converting the previous version of lojban into this one
jbocre: 18:23 <UukGoblin> add a word that misparses straight away, that means 'The following is lojban version x1'
jbocre: 18:23 <vensa> for example, in this case the script would be {paunai}=>{paucu'i}, and since {paucu'i} was previously undefined that is all
jbocre: 18:23 <UukGoblin> that way, old interpreters / parsers will stop at the very beginning
jbocre: 18:24 <vensa> uuk: yes, we also might want a cmavo that declares the version number
jbocre: 18:24 <tomoj> impossible in general, I think
jbocre: 18:24 <dbrock-> the conversion script sounds like a waste of time
jbocre: 18:24 <tomoj> at least, required conversion scripts restricts the kinds of changes you can make
jbocre: 18:24 <vensa> tomoj: challenge me
jbocre: 18:24 <tomoj> s/required/requiring/
jbocre: 18:24 <vensa> what cant be done in a script?
jbocre: 18:24 <dbrock-> but a cmavo that indicates dialect or version? sure, that would be useful
jbocre: 18:24 == kpreid jbocre: [email protected] has joined #lojban
jbocre: 18:24 == mode/#lojban +o kpreid by ChanServ
jbocre: 18:25 <tomoj> e.g. imagine trying to convert pre-xorlo to xorlo
jbocre: 18:25 <vensa> I thought about it already
jbocre: 18:25 <tomoj> you could, I guess, simply replace every {le} with {lo}, but..
jbocre: 18:26 <ksion> doi la vensa ma smuni lo'u paucu'i le'u
jbocre: 18:26 <vensa> but it probably requires a deep understanding of what xorlo does, which I dont :)
jbocre: 18:26 <dbrock-> you could just use {bau ko'a} at the start of your text
jbocre: 18:26 <vensa> ksion: the proposal is that {paucu'i} be the new "rhetorical question" and {paunai} changed to "answer follows"
jbocre: 18:26 <tomoj> what I'm saying is that deep understanding is required, but not of the change, of the text you're trying to convert
jbocre: 18:26 <dbrock-> {bau lo fadni} for standard Lojban, for example
jbocre: 18:27 <mathw> which standard? when? :)
jbocre: 18:27 <@xalbo> {le} is unchanged under xorlo, and {lo} only gains meaning. So previous text can be left unchanged.
jbocre: 18:27 <tomoj> certainly not?
jbocre: 18:27 <vensa> tomoj: I would ultimately replace every {le} with {lo}, unless it had a {bi'unai} after it, in which case I would make it {le}
jbocre: 18:27 <dbrock-> mathw: I dunno, CLL/BPFK/whatever?
jbocre: 18:28 <vensa> tomoj: you also need to think how you would translate pre-xorlo {lo}
jbocre: 18:28 <tomoj> pre-xorlo {re lo ci bakni} needs to be changed, right?
jbocre: 18:28 <vensa> and that also has a def under xorlo I beleve
jbocre: 18:28 <mathw> dbrock-: just pointing out that the standard changes, so you can't just say 'this is standard lojban', you'd have to say 'this is standard lojban as of the first of may 2011'
jbocre: 18:28 <mathw> or something
jbocre: 18:29 <vensa> yes
jbocre: 18:29 == jey jbocre: [email protected] has joined #lojban
jbocre: 18:29 <mathw> translating to xorlo would require careful attention to the numbers of things, wouldn't it?
jbocre: 18:29 <vensa> tomoj: {re lo ci bakni poi zasti po'o vi lo munje} :P
jbocre: 18:29 <mathw> I believe that changes a bit
jbocre: 18:30 <mathw> But I never understood non-xorlo quantities so I will refrain from further comment
jbocre: 18:30 <vensa> IMO it CAN be done
jbocre: 18:30 <mathw> oh it can sure
jbocre: 18:30 <vensa> (the version script)
jbocre: 18:31 <vensa> maybe there will be a small percent of innacuracy
jbocre: 18:31 <mathw> .u'i mu bakni
jbocre: 18:31 <vensa> but that's still better than nothing
jbocre: 18:31 <ksion> vensa: i'e I like it.
jbocre: 18:31 <soto> imo we should just leave it untranslated :p
jbocre: 18:31 <vensa> and it allows us to "grow" with the times
jbocre: 18:31 <vensa> and not be stuck in the mud becuz "someone 500 yrs ago decided it should be this and not that"
jbocre: 18:31 <vensa> ki'esai ksion
jbocre: 18:32 == rossi jbocre: [email protected] has joined #lojban
jbocre: 18:32 <vensa> soto: ppl like rlp who have written 60K word essays wont like that their writings are no longer supported...
jbocre: 18:32 <vensa> like windows95 :P
jbocre: 18:33 <vensa> xalbo: whats ur opinion on the version scripts?
jbocre: 18:34 <vensa> it seems to me fitting that lojban have version numbers. after all, it is mostly used by computer programmers :P
jbocre: 18:35 <@xalbo> think a lot of the work is figuring out what (if anything) old versions actually meant. And much of the changes are fixing that. So I don't think we can necessarily convert.
jbocre: 18:35 <@xalbo> version numbers seems useful
jbocre: 18:35 <@xalbo> requiring a script for all changes seems to assume that we can agree on what things were, which is often the problem in itself.
jbocre: 18:36 <mathw> And often a reason why a change is proposed in the first place from what I've seen
jbocre: 18:36 <vensa> xalbo: true
jbocre: 18:36 <vensa> so in places where there was no previous explicit meaning, we can maybe add a conversion comment
jbocre: 18:36 <vensa> {to'isa'a na se djuno toi} :)
jbocre: 18:37 <vensa> but think of all the rafsi that could be reallocated easily
jbocre: 18:37 <vensa> and the cultural gismus that can be abolished and turned into fu'ivla
jbocre: 18:38 <mathw> co'o
jbocre: 18:38 <vensa> co'o mat
jbocre: 18:38 <@xalbo> Those are both arguments *against* what you're attempting to do, in my mind.
jbocre: 18:39 <vensa> hehehe
jbocre: 18:39 <vensa> becuz YOU dont want to need to relearn stuff. right?
jbocre: 18:39 <@xalbo> "Who cares if we fuck over the people who learned the language earlier? They can just apply this 300 line sed script to their minds, and all is good."
jbocre: 18:39 <vensa> I knew youd say that :)
jbocre: 18:39 <vensa> <3
jbocre: 18:39 <vensa> xalbo: you are right
jbocre: 18:39 <vensa> but think of the other hand
jbocre: 18:40 <vensa> being stuck for ages with a bad choice of gismu or grammar, way after all those ancestral lojbaners have died
jbocre: 18:40 <UukGoblin> we could have 3-way handshakes to determine the version of lojban to use at start of a discussion! how cool would that be?
jbocre: 18:40 <vensa> xalbo: we can put it a standard that a change may only happen once in X years
jbocre: 18:41 <vensa> uuk: lol
jbocre: 18:41 <ksion> UukGoblin: ACKsai
jbocre: 18:41 <vensa> xalbo: you could still talk in lojban1.0 with you lojban1.0 buddies :)
jbocre: 18:41 <vensa> kinda like old folks speak yiddish
jbocre: 18:41 <vensa> and dont know slang
jbocre: 18:42 <ksion> <?lojban version="1.0"?>
jbocre: 18:42 <vensa> :)
jbocre: 18:42 <@xalbo> I don't think a conversion script is necessary or sufficient to allow for unlimited changes to the language, and I am undecided on whether its utility outweighs its cost.
jbocre: 18:42 <vensa> xalbo: a conversion script will also make sure that it is harder to change stuff
jbocre: 18:42 <vensa> cuz you need to supply the script
jbocre: 18:42 <UukGoblin> agreed, major changes might require interpretation, not just mere transcription
jbocre: 18:43 <@xalbo> Right, and that's part of the cost.
jbocre: 18:43 <vensa> but you wanted things to not change
jbocre: 18:43 <soto> Having "number versions" for a language seems incredibly odd to me, but then I imagine a robot saying {coi do. I speak lojban v2.35. Beep.} and then I am tempted to change my mind because robots are so cool!
jbocre: 18:43 <vensa> make upo your mind
jbocre: 18:43 <@xalbo> And then arguing forever about whether your script is *right*, instead of just about the merits of the change.
jbocre: 18:43 <vensa> soto: lol
jbocre: 18:44 <vensa> xalbo: the scripting should be handled by a seperate "backoffice" department of the BPFK :)
jbocre: 18:46 <vensa> btw: you guys didnt address an open issue of "how do we convert audio recordings"
jbocre: 18:46 <vensa> but I am assuming that is equally as plausible, assuming we have a powerful speech-recognizer
jbocre: 18:46 <UukGoblin> loi
jbocre: 18:48 * vensa fantasizes about lojban3.0 where {tel} would be the rafsi or {te} and {go'i} would switch places with {goi} :)
jbocre: 18:49 == Wolvenreign jbocre: [email protected] has joined #lojban
jbocre: 18:49 <@Broca> What would you change {go'a} to?
jbocre: 18:49 <vensa> (and {a'y e'y i'y...} would replace the ugly {abu ebu ibu})
jbocre: 18:49 <vensa> valsi go'a
jbocre: 18:49 <valsi> go'a = pro-bridi: repeats a recent bridi (usually not the last 2).
jbocre: 18:49 == Wolvenreign has changed nick to labnytru
jbocre: 18:49 <labnytru> coi rodo
jbocre: 18:49 <vensa> go'a could stay go'a
jbocre: 18:49 <labnytru> I was wondering...
jbocre: 18:50 <vensa> I just think that since {go'i} is so frequently used, it should be reduced to 1 syllable
jbocre: 18:50 <labnytru> What would be a good Lojban translation for "infinite"?
jbocre: 18:50 <vensa> valsi cimni
jbocre: 18:50 <kribacr> I think ji'i?
jbocre: 18:50 <valsi> cimni = x1 is infinite/unending/eternal in property/dimension x2, to degree x3 (quantity)/of type x3.
jbocre: 18:50 <@Broca> Why do you think tel should replace ter?
jbocre: 18:50 <kribacr> There's a number for infinitity. I know that much.
jbocre: 18:50 <vensa> valsi ci'i
jbocre: 18:50 <valsi> ci'i = digit/number: infinity; followed by digits => aleph cardinality.
jbocre: 18:50 <kribacr> ci'i, sorry.
jbocre: 18:50 <kribacr> Yeah, based off of cimni. Makes sense.
jbocre: 18:50 <labnytru> Ah, thanks.
jbocre: 18:50 <kribacr> I knew it was Ci'i.
jbocre: 18:50 <vensa> broca: for aesthetic cohedrence: sel tel vel xel
jbocre: 18:51 == Jantaro jbocre: [email protected] has joined #lojban
jbocre: 18:51 <labnytru> Well, I've got good news.
jbocre: 18:51 <UukGoblin> vensa, there was a reason why ter is not tel
jbocre: 18:51 <@Broca> vensa: but you don't think go'a go'e go'i go'o go'u should be coherent?
jbocre: 18:51 <labnytru> I've found an official tutor, and am currently in the process of setting up a website and doing some SEO so that my income will be taken care of permanently.
jbocre: 18:51 <vensa> yes. because of stupid gismu for {stela}
jbocre: 18:51 <labnytru> What does this mean to the Lojban community?
jbocre: 18:51 <UukGoblin> you can't just go around changing everything for aesthetic reasons ;-]
jbocre: 18:51 <labnytru> It means I'm going to come here and stay here once I'm prepared.
jbocre: 18:52 <vensa> broca: that is a place where brevity trumps coherence IMO
jbocre: 18:52 <vensa> valsi go'u
jbocre: 18:52 <valsi> go'u = pro-bridi: repeats a remote past bridi.
jbocre: 18:52 <@xalbo> vensa: strongly disagree
jbocre: 18:52 <labnytru> I'll learn the entirety of the language and make this chatroom my "home".
jbocre: 18:52 <vensa> broca: and {go'u} is not even in its right place in the series
jbocre: 18:52 == MigoMipo jbocre: [email protected] has joined #lojban
jbocre: 18:52 <labnytru> brb
jbocre: 18:53 <vensa> <@xalbo> vensa: strongly disagree <-- about what?
jbocre: 18:53 <vensa> (there, I copied this time :]
jbocre: 18:53 <@xalbo> {go'i}/{go'a}/{go'u} follows the normal yow series. That leaves {go'e} and {go'o} for ad-hoc interpretation.
jbocre: 18:53 <vensa> (it's hard for me to copy cuz my mouse is laptop-internal :()
jbocre: 18:53 <@xalbo> vensa: I think breaking {go'i} out of the series is not justified by brevity considerations.
jbocre: 18:53 <@Broca> But why stop there? If what you need is brevity, why not just do {go'i} → {.a}?
jbocre: 18:53 <vensa> xalbo: cool. didnt think of it that way. thanks
jbocre: 18:54 <vensa> broca: that diff seems smaller than the diff between 1 and 2 syllables
jbocre: 18:55 <@Broca> So you seriously think swapping {goi} and {go'i} could be done?
jbocre: 18:55 <vensa> xalbo: but it is justified to break a series for stupid rafsi considerations???
jbocre: 18:55 <vensa> broca: why not? everything is possible
jbocre: 18:55 <@Broca> That is not funny. GDIAF.
jbocre: 18:56 <vensa> ki'a GDIAF?
jbocre: 18:56 <@xalbo> "everything is possible": Not in my Everett branch!
jbocre: 18:56 <vensa> huh?
jbocre: 18:56 <@Broca> http://www.google.com/search?q=gdiaf
jbocre: 18:56 <vensa> not familiar with everet
jbocre: 18:56 <@xalbo> http://wiki.lesswrong.com/wiki/Everett_branch
jbocre: 18:57 <vensa> broca: I dont understand the motive for your hostility. u'i
jbocre: 18:57 <vensa> si
jbocre: 18:57 <vensa> u'u
jbocre: 18:57 <vensa> (my finger slipped)
jbocre: 18:57 <vensa> xalbo: oh. multiverse
jbocre: 18:58 <vensa> you guys are like the two grumps in the muppets :)
jbocre: 18:58 <vensa> thats cute
jbocre: 18:58 <@xalbo> I think -tel- would be a better rafsi for {te} than -ter-, if starting from scratch. I just don't think the difference is sufficient to be worth changing.
jbocre: 18:58 <@Broca> xalbo: why do we always come here?
jbocre: 18:58 <vensa> .u'i
jbocre: 18:59 <@xalbo> Broca: I just enjoy seeing the curtain close at the end.
jbocre: 18:59 <vensa> xalbo: fair enough
jbocre: 18:59 <@Broca> Ha ha ha ha!
jbocre: 18:59 <vensa> but with my scripting system, it may be possible to change without upsetting the system too much
jbocre: 18:59 <@Broca> (Your line is “I guess we'll never know”, by the way)
jbocre: 18:59 <vensa> .u'isai doi mapets
{CODE}