Wikiforum threads backup

From Lojban
Jump to: navigation, search

1 lo vrici nu casnu 1 20130730132811 1 2147483647 1 0 0 Gleki
2 jboske / BPFK 2 20130730133048 1 2147483647 1 0 0

2 loi cnino stura ko'oi casnu lo cnino je se stidi stura be le bangu 2 1 2 2 9 20130801155740 20130730133331 1 0 0 0 0
3 Lo blogo pe La Gleki - La Gleki's blog lo vrici be mi se pensi - My various thoughts 1 1 13 18 1 20140112152246 20140107074512 1 20140405080642 1 0 0 1
5 lo vrici 1 2 -1 0 0 20141220135541 1 0 0 0 0

1 pe'a coi .ui lo lojbo munje .i mi'e la .ilmen. mu'o 20130730135251 8 0 0 0 1
2 test 20130730140837 0 2147483647 1 0 1
3 testing 20130730141423 0 0 0 0 1
4 .i coi mi'e la tsani ko mo la .estas. 20130801053508 9 0 0 0 1
5 I'd say that L4N should build upon previous chapters, since we already have the CLL for the other option. 20130801205654 53 0 0 0 3
6 First. I wanted to create such thread today only to discover that it has already been created. ki'e la tsani i u'i

Next. I don't know who will move and where (in case it's important for you I'm informing you that u r already an administrator, a bureaucrat, a forum administrator and a ... bot of LMW ;) ). Anyway I have nothing against placing this thread here. WikiForum extension has basic functionality that we need. I had no desire of installing another more powerful phpBB-like software that would require independent registering.

As for L4N I created it with the following goals. Create a full up-to-date tutorial for newbies that reflect "LST Group" Lojban (LST = latro'a, selpa'i, tsani). Those who don't like this dialect of Lojban are free to fork this project under a different page on this wiki. We won't touch your project, nu'e.

L4N has following tasks:

1. fix formatting (has been mostly done, e.g. all hyperlinks are fixed). Bold text is always for Lojban. Italics is for english examples and translations. Special notes are prefixed with ; in mediawiki syntax. Sentence examples are prefixed with ::

2. apply xorlo. Explain {le}.

3. apply dotside

4. cmevla = brivla. Add a short explanation of what it is (specifically, there is not problem here at all ;) ). I've been trying to add {cu} in all {la CMEVLA [cu] selbri} constructs. See if i missed it somewhere.

5. gije (again if you dont like it fork L4N). After gije-fying almost all sentences now I just HATE the old connective system. It's like ~3 times simpler. However, i havent fixed {gu'e} yet. GIhA is retained in the language for backward compatibility but is not mentioned in L4N. Also we need to carefully check all examples with connectives.

6. Subjunctives etc.

Also I placed comments inside < ! -- something - - > tags that have instructions on what to do next. Actually I read the whole L4B once again. So you'll find other tasks inside the text itself.

I mentioned ka as infinitive somewhere there but it's nice you are mentioning it here.

As for the corpus we also have Tatoeba. Doesn't matter if sentences there are bad Lojban. Sometimes sentences just test Lojban capabilities and if it's bad Lojban it's even better. It means that newbies don't understand exactly this part of Lojban!!! In such cases you ask me to change the sentence in Tatoeba, you add yourself the correct translation to L4N where needed.

Also may be we should teach zo'ei in {tu'a} section.

We have a decision to make: do we make L4N more "tutorial"-ish, or "textbook"-ish?

tutorial-ish as it was originally conceived. As for textbook I'm gonna fork CLL into an unofficial textbook again here in LMW. But before that I need to create a full two-way docbook <=> Mediawiki converter for docbook-CLL. Not that hard but I'm not sure I'll have enough time soon.

That if you want a Lojban name, it should be a selbri.

<-- Should be? I think can be. Not should be.

Except these last two points I completely approve of what you are suggesting, doi la tsani.

20130802042111 1 0 0 20130802042428 1 3
7 Well, we are lucky. Explaining logical connectives without resorting to EOR and similar shit IS possible :) 20140108083343 1 0 0 0 4
8 So now I suggest using mediawiki templates for indices and similar things in ELG (as an alternative to CLL):

There is nothing hard in converting docbook-CLL to this format. The tags I suggest (like {{g|gismu}}) are concise, easy to use for manual editing of ELG.

So I wonder what others will do after I create ELG for them. Probably nothing. As usual.

They say "BPFK is dead" but the question is there anybody who is alive? ju'oru'e they themselves won't start contributing too.

btw is anybody reading this blog?

20140109081103 1 0 0 20140109081126 1 4
9 Huh, i'll mention even old connective system in L4N. 20140109103648 1 0 0 0 4
10 Crash Course in Lojban:_The_Journey - this is the main place that stores all the info on future textbooks. 20140110164357 1 0 0 0 4
12 Also once I asked why eu is not a legal diphthong in Lojban. I think I just answered my own question. :D 20140112152246 1 0 0 0 5
13 i realized that it might be fi'i for greeting those who come to the channel, and fi'idai when you enter the channel yourself 20140117131659 1 0 0 20140117145511 1 6
15 Stimulus s1 causes reaction s2.
s1 = a1 cu aska a2
s2 = r1 cu repli r2

s1 stima s2 = s1 selzilfra s2

20140125155312 1 0 0 20140125155634 1 9
16 stimu = x1 (NU) stimulates x2 (NU)

lo nu mi cusku zo ju'i cu stimu lo nu do cusku zo re'i

mi jai stimu tu'a do ==(raising)==> mi stimu do

ko jai se stimu tu'a mi => ko se stimu mi (Answer to me!)

20140125163839 1 0 0 20140125164839 1 9
17 test 20140228111316 0 2147483647 1 0 11
18 А что такое ложбан вообще? 20140228111630 0 2147483647 1 0 13
19 df 20140304085605 1 2147483647 1 0 13
20 May be {faje}? 20140305051733 1 0 0 0 33
21 Another solution is just to use retsku - spuda - spusku triad.
mi spuda lo se retsku be ra lo ka spusku zo coi
20140305101705 1 0 0 20140507180033 1 9 Gleki
441 sei and le have similar grammar thus they need to be explained together. 20140624103744 1 0 0 0 4

1 ma do cmene coi i mi'e la gleki 0 20130730132925 1 2147483647 1 0 0 0 1 4 45 9 20130801053508
2 cumki munje saske / Subjunctives cfari 0 20130730133402 1 0 0 0 0 0 2 0 25 0 20130730133402
3 Developing the New Version of L4B (aka L4N) Moderators: please move this topic into a more appropriate section of the forums.

We should use this thread for discussing changes to L4B in order to make a version of the book that describes IRC Lojban as much as possible. I believe that a forum thread is more appropriate than the article's talk page, as we can keep the talk page for discussing the formatting, which is the standard use for talk pages. To be honest, I think it ridiculous to describe a language hardly spoken, and if so only by those few mailing list lojbanists who seldom even say anything at all in Lojban.

After quickly scanning the text and discussing briefly with la gleki, I know that xorlo and dotside have already been added to the text.

Furthermore, we should teach the following:

  • The modern interpretation of ka and of other abstractors.
  • That if you want a Lojban name, it should be a selbri. (Turns out L4B teaches that you should modify gismu to make cmevla if you want a Lojban name, e.g. mlatu -> .mlat. or .mlatus.)
  • The modern interpretation of ZAhO tags.
  • Common na, i.e. that selbri-tag na is equivalent to negation of the bridi-tail (which itself is frequently equivalent to in-place naku, but not always.)
  • That vi does not mean "at location ..." (that's bu'u), but rather that it tags a spatial distance, like zi tags a temporal one.

Also, I think that we should emphasize common useful constructs. The original L4B and to a lesser extent, the CLL, did this, but at the time of their publication, there was much less lojbanic "idiom" in existence, so doing so was understandable difficult. We should teach constructs such as zo'e ne, me SUMTI moi, and others. (I can't think of anymore off hand.)

Obviously, we should make more examples and exercises, ideally drawing examples from the corpus. The problem I've personally had when trying to take corpus examples is that frequently, any given sentence combines too many yet-unseen constructs for the purposes of the textbook. For instance, when explaining tags, I'll look up an example of mu'i in the corpus, but a great deal of those uses also use logical connectives, more complicated tenses, more advanced quantifiers, etc. Of course, we'll have to sift through the corpus and edit out these things before using them as examples in their own right in L4N.

As for editing L4B to make it easier to understand, it would obviously help if I had originally learned from L4B, but alas, I did not. I'll consult with la latro'a, as I know he has read L4B. He has also told me that L4B "gets weird" in the later chapters. Could it be that he's referring to the ones written by Nick Nicholas? If that's the case, we'll have to work on a unified style throughout the textbook and ensure that each chapter can be well understood.

We have a decision to make: do we make L4N more "tutorial"-ish, or "textbook"-ish? What I mean is this: should each chapter be self-contained in such a way as that it depends as little as possible on other chapters, or should each chapter truly build on the knowledge of the previous chapter. Both have their pros and cons.

If we make self-contained chapters, then the book can be used more readily as a reference with regards to a particular topic. This means that the chapter on tanru will explain *everything* that a tanru can do and all the interactions of tanru with other parts of speech, while at the same time limiting its references to other parts of the grammar. The CLL is constructed in such a way, and I'm not sure if it does a good job at it or not. Its chapter on tanru refers to logical connectives within tanru. Is this something that we'd want to do in our textbook, or should we wait until we fully introduce logical connectives before saying that they can be used in tanru?

.i mi'e la tsani mu'o

0 20130801155740 9 0 0 0 0 0 2 2 51 1 20130802042111
4 Lojbanistan: news from the battle-front L4B is sometimes hilarious.

First I thought I could update it to match the new Lojban. It appeared that L4B lacked too much.

  1. No mentioning of fi'o!
  2. As it was just mentioned in IRC, vo'a can be only long scope. In fact L4B mentioned everything else: {no'a} and {nei} but doesnt use it!

More to come...

0 20140107074916 1 2147483647 1 20140110164304 1 2147483647 1 3 73 9661 0 20140704190137 Gleki
5 Oddball orthography I don't like that x and h are distinct sounds in Lojban. It causes confusion for Speakers of English, Russian, Chinese.

Here is a solution.

  • " ' " - can be used as ",". Thus it can be pronounced as a hiatus.
  • ai, ei, oi, au, ia through iu and ua through uu are always pronounced as [aj], [ej]..., [wu] never as [a,i], [e,i], ... [u,u].
  • a'i, e'i, o'i and a'o, i'a through i'u and u'a through u'u are pronounced as [a,i], [e,i], ... or normally as [ahi], [ehi], ... It might be necessary to still use [h] in .a'o, .a'u to distinguish from .au
  • All other vowel pairs simply can be pronounced with [h] / [,] or without it as you wish. So you can pronounce e'u as [eu].
0 20140112152024 1 0 0 20140124113143 1 0 0 3 1 44 1 20140112152246
6 What do we need in Lojban? we need separate vocatives for mi cliva co'o and do cliva co'o, mi tolcanci coi and do tolcanci coi. 0 20140117100748 1 0 0 20140117145535 1 0 0 3 1 23 1 20140117131659
7 The new old concept. Autocorrection of places Moved to Auto-correction_of_places 0 20140125062323 1 0 0 20140624083821 1 0 0 3 0 103 0 20140125062323
8 Ci/uV. Our last hope. Free cmavo space is being depleted. Luckily we still have free experimantal space.

CiV and CuV, .iVV .uVV is our last hope and our last resort.

Let's look at possible cmavo carefully.

We have 16 consonants in Lojban and 5 basic vowels.

  1. Some complain that {siV} are hard to differentiate from {cV}. We can assume that {jiV}, {ciV}, {ziV}, {niV} will have the same problems. Thus we have to remove them.
  2. The same can be said about {buV}, {puV}, {vuV}, {fuV}, {muV}.
  3. It's also hard to pronounce {Cii} and {Cuu}.

Thus we have only 86 possible cmavo left in CiV and CuV space. Not too much. What meaning can we assign to them?

  • PU has problems with PUbo. So use piu instead of pu and sepiu instead of ba but with no problems with bo.
  • {kia} (From Russian imperative clitic "-ка") is a UI of command. do kia klama mi - Come to me!
  • {kui} = sedu'u. I believe that {du'u} and {su'u} should mostly die and be replaced with simple {nu} or additional brivla (when one want to use {su'u2}). Although {mi cusku lo kui broda}={mi cusku lo se smuni be lo nu broda} or {mi cusku lo jufra be lo nu broda}
  • {mia} = mi jo'u lo drata be mi. Etymology: Bavarian. Corresponds to English "we". Before accepting the proposal of this topic is to be used in the form {mi'ai}


A careful of BAI-system and N-paradigm might ask for using a bunch of cmavo. We might wanna provide it with them.

A joke word would be

  • {kua} - the sound that frogs produce.
0 20140125075509 1 0 0 20140211103932 1 0 0 3 0 41 0 20140125075509
9 ask/reply brivla I want a "x1 asks x2 to x3 getting reply x4" brivla.

Or may be split it into two:

.aska = a1 asks a2 to a3
repli = r1=a3 answers r2 to r3=a1

a1 asks a2 (text/stimulus) to a3=r1, who makes reply/action r2 from question a2 to r3

Usually but not always r3=a1.

r1 replies not to the event encoded in a1,a2,a3 but r1 replies {go'i} only to a2.

r1 frati r2 a2 poi a1 reisku ke'a

0 20140125151908 1 0 0 20140125155121 1 0 0 3 3 50 1 20140305101705
10 mz in Lojban? I think {mz} should be allowed in Lojban BUT !!! made equal to {nz} cuz {djeins} in noisy enviroment would be undistinguishable from {djeimz}.

I wish {l}={r} and {x}={'} in Lojban too but that's too late.

0 20140126064143 1 0 0 20140604064043 1 0 0 3 0 14 0 20140126064143 Gleki
11 kairgismu and faurgismu Some gismu are in the form of
x1 broda lo ka brode. Let's call them kairgismu.

Some are

lo nu x1 brode cu broda. Let's call them faurgismu.

This is a serious problem for lojban (and probably for many natural languages).

mi kakne lo ka bajra => lo nu mi bajra cu faurkakne
mi kairfanza lo ka cladu tavla < = lo nu mi cladu tavla cu fanza

Some gismu were fixed from kairgismu to faurgismu (see L4B for the history of the place structure of {fanza} and {tcica}). Some havent been fixed (e.g. {bebna}).

Probably faurgismu are potentially better.

0 20140206081547 1 0 0 20140207160614 1 0 0 3 1 27 0 20140228111316
12 Your questions? (In English) Post any questions about Lojban or logical languages here. 0 20140228111454 1 0 0 20140314165832 1 0 0 1 387 1028 0 20140704101447 Dimnas
13 Ваши вопросы (по-русски) Задавайте здесь любые вопросы по языку ложбану (Lojban) или другим логическим языкам. 0 20140228111536 1 0 0 0 0 0 1 2 31 1 20140304085605
14 Vos questions (en français) S'il vous plaît, poser des questions sur Lojban ou langages logiques ici. 0 20140228112203 1 0 0 0 0 0 1 0 26 0 20140228112203
33 The problem of gi'e/vauje/gije and bridi tails ilmentufa now allows to say:
mi nelci do je citka lo plise

Yay! It works without any terminators.

Once again: it works. Next (broken) example:

*mi nelci lo ninmu je citka lo plise

Oops, that doesn't work. What to do?

mi nelci lo ninmu vau je citka lo plise might look awkward to some of us.

But remember that pronoun + je works. So what about mi nelci lo ninmu zo'e je citka lo plise

We don't have nelci3 but if it appears in future (this once happened to carna that got a new place) this example would be perfect. So I suggest that zo'e filling non-existing places is always okay.

So what we need here is a dummy word, possibly a pronoun. zo'e is nice but a two syllable one.

  • mi nelci lo ninmu da je citka lo plise won't work in all cases.
  • *mi nelci lo ninmu cu je citka lo plise would be a complete rework of cu
  • mi nelci lo ninmu ca je citka lo plise parses but isn't the same in meaning adding unnecessary information.

Do we have a good CV cmavo dummy enough to work as a "terminator"?

What is more in order to put e.g. a UI in the end of the sentence one has to use vau vau to attach it to the whole sentence:

mi nelci lo ninmu vau je citka lo plise vau vau .ui.

thus we again get baby talk with vauvauvauvau which loglan was famous of. This is another problem to solve.

0 20140304060814 1 0 0 20140304071146 1 0 0 3 1 36 1 20140305051733
34 A gismu for "arrive"? I think we've got too many requests for such gismu.

The most obvious choices are Anglo-Romance arriv-/ariv-/arib-, Chinese "dau" ("dao").

But again they will mutilate the form of the word making it hard to memorize for everyone.

After looking at translations to different major languages I noticed Japanese {tsuku}.

.i pei

0 20140305070348 1 2147483647 1 20140305071329 1 0 0 3 0 12 0 20140305070348 Gleki
35 Your ideas about le Start. 0 20140307060021 1 2147483647 1 20140507174628 1 0 0 3 0 15 0 20140307060021
36 A gismu for "cause"? A gismu for {se jalge}.

Candidates are Chinese "dao zhi" => {dadji}.

Also consider Anglo-Romance "causa". {kozda}?

We'd also need a BAI and a rafsi for it.

0 20140311070020 1 0 0 0 0 0 3 0 8 0 20140311070020
37 No uniformity in fu'ivla (copied from;natlangword=0;commentid=993;definition=52907)

krtisfranks wrote: > I would prefer a fu'ivla. Nylon has become a generic word for the material > and it could be useful to have such a brivla (similar to silka and its > ilk). Also, the compound/substance "nylonase" could benefit in its > translation if we have such a brivla.

The same question as for Higgs bozone. Do you have a complete algorithm of how to Lojbanize physico-chemical nomenclature?

The attempts by far have been mostly cunso, not just mutilating the form of the original world-wide accepted terms but mutilating them in a new direction with every new contributor.

Specifically for this case X-ase is a non-polysemous suffix denoting a protein that can destruct X. Will the form of both "X" and "-ase" be retained the same throughout all enzymes?

Or it will deviate from this rule here and there always bowing its head to self-segregating morphology thus making Lojban terminology less cleancut than the international one

0 20140319161334 1 0 0 0 0 0 3 0 20 0 20140319161334
38 0 20140320071939 1 2147483647 1 20140505071620 1 0 0 3 0 20 0 20140320071939 Gleki
39 0 20140320175917 1 2147483647 1 20140505071617 1 0 0 3 0 5 0 20140320175917 Gleki
40 0 20140324141006 1 2147483647 1 20140505071605 1 0 0 3 0 7 0 20140324141006 Gleki
41 0 20140326103735 1 2147483647 1 20140505071629 1 0 0 3 0 4 0 20140326103735 Gleki
42 zo'e, zu'i, zi'o words,_zu%27i,_zi%27o_words 0 20140329115853 1 0 0 20140529152732 1 0 0 3 0 66 0 20140329115853 Gleki
43 "Simple English" as the first swallow of the summer.  ;gleki:We have a new language in jvs: Simple English. I will use it for new gismu definitions. Definitions will be simple. Notes will have a more precise explanation. Final places can be ignored (like "by standard").
Export will be trickier: all words that haven't been defined in en-simple will be taken from English direction.
is this simple english, like on wikipedia?
No, ... I'll reuse this direction for something really simple and useful.
because currently it's not that simple as it could be.
just abridged definitions.
so it's simple english, just not Simple English
yes, i just didnt know what to call it.
the same will go for UI:!topic/bpfk-list/BQP2FONjbi4
ah yes those definitions are confusing for nintadni
0 20140330074122 1 0 0 20140504080140 1 0 0 3 0 17 0 20140330074122
44 zi'o and zo'e Omitted places should be filled not with zo'e but with no'e zi'o.

We have

  • zi'o - no place
  • no'e zi'o - I don't know if this place exists
  • to'e zi'o - this place is filled.

zo'e has two meanings:

  1. something obvious from context
  2. to'e zi'o (this place is filled)

Lojban needs a clear tool to specify ternary logic.

0 20140529133437 1 0 0 0 0 0 3 0 22 0 20140529133437 Gleki