ju'u

Since lojban allows the use of a variety of bases, ju'u is given to specify which is being used at the moment. The problem lies in specifying base 16...there is no cmavo for the digit 16. The most simple solution is to give the base as N-1. so decimal is base "so" (9), binary is base "pa" (1), and hexadecimal is base "vai" (15).

• If you are really that paranoid, set the base to 10 with dau and then to 16 with paxa.

Well, there are digits for 10 and 2, so we can specify decimal with ju'u dau and binary with ju'u re and dozenal with ju'u gai. If there's an interpretive convention to be proposed, I'd say allowing the only-slightly ambiguous ju'u su'ovai for hexadecimal. That wouldn't break existing usage or change the way it's used for other things. Won't help the die-hard octodecimalists out there (base-18), but then they have other problems. --mi'e .mark.

stidi lo'u vei vai su'i pa le'u (Certainly that's fine and correct. su'ovai is just a lot shorter and more Zipfean)

• Base-18 isn't represented by cmavo, so it is a special case. The problem is no matter how many cmavo you have, specifying that you want to us the highest base possible will be a problem. You could use "ju'u pano" for whatever highest base you have, but it's two syllables and still ambiguous. In any case, you can use "za'e" before the ju'u or PA cmavo to indicate you are trying this new method. By the way, shouldn't it be "su'evai", for "at most 15"?

• As I said, octodecimalists have other problems. Presumably you'd have to make nonce cmavo from experimental space for those rare occasions when you're dealing in such bases (I favor things like pa'ai / re'ai / etc for mnemonic value). Two syllables isn't a problem. Ambiguity is. pano is always the base I'm talking in at that instant! And you're exactly right: whatever our highest "primitive" number X is, using X-plus-one as a base (which is presumably why we have X as a digit) yields this problem. Hence the proposal for a convention: su'ovai, "at least 15", in the context of being specified as a base, being interpreted as "16" precisely, unless other considerations interfere. So we have "...in base 10" and "...in base 5" and so on meaning exactly what they say, and "...in base 15-or-more" conventionally taken to be "...in base 16." Get it? za'e is okay, and it's a word I definitely like, but I think it applies more to "this is an experimental word" than "interpret this construction funny."

Is there a method to set the base in a sticky fashion? That would be significantly more useful than eternal wanking about X is the one true base! --jay

• i thought i heard lojbab say something about this. i searched the lojban list for it but couldn't find it. the book does mention a need for a sticky order of operations, and i'd imagine you would use a similar mechanism for this. --tinkit

• • I, too, am too lazy to look things up, but as I recall, the book mumbles something about how there should be a mechanism for this, but it isn't worked out now, and it's something future users will have to work out.

• • • Actually, that's about operation priority. The general mechanism is there (ti'e), but exactly how to use is undefined.

• • I say, why get all bent out of syntax trying to come up with new grammar and cmavo, instead of just saying what you mean? If you want addition to have a higher precedence than multiplication in your discussion, say so with words and sentences and everything. Something like va'o lei dei mekso le nu jmina cu lidne le nu pilji ku zo'e... (mm, I don't like that. lidne isn't a good word for it. Maybe something with porsi or some such.) Whatever; the point being that you simply explain what changes in expected defaults are going on. That's how it works in contemporary scientific texts all the time. I often see things like "...in this book, we use log x to mean the natural logarithm and never the common logarithm" or "in this section, numbers are all in base-16," or "this book departs from IPA in using � for the open front unrounded vowel instead of a, which we use for an open mid unrounded vowel." It works fine in real life, why not in Lojban too? Say what you mean, and mean what you say. Don't be afraid of jufra and bridi to explicate your meaning. --mi'e mark

• • • Exactly so. --John Cowan

• stidi zo ki (Applied to what? ki can't just stand by itself: it attaches to a tense, not a MEX) mu'a lo'u co'e no'oxipa ju'u no'oxire ki le'u .ija pilno zo ju'a

When all else fails, say what you mean: mi kancu fo li dau (that a good gismu-place for this?) or li dau cu namcu befa mi jimcu or whatever. --mi'e .mark.

• Yes: this is definitive, but it seems to me plenty lobykai to assume once the base is stated, it remains in force until explicitly changed.)

• • Indeed. --John Cowan