Re: periodic hexadecimal reminder

[mark@kli.org]

--- In lojban@y..., Pierre Abbat <phma@o...> wrote:
> On Wednesday 26 September 2001 17:00, John Cowan wrote:
> > It's nonsense.  The reason ju'u is an operator rather than a
> > PA is to allow talking about variable bases:
> > (10 base b) - 1 = 10 base (b - 1), e.g.  But the left operand 
can't
> > be anything but a digit string:  (a+b) base c is just nonsense.
> 
> Which is why ju'u should be in its own selma'o.

Not necessarily.  selma'o often do things like this, including words 
that only make sense in certain situations.  {kau} doesn't really 
belong in UI, it's just handy because of where UI words can go (just 
about anywhere).  The various "shift" characters in BY aren't really 
lerfu.  {mi'e} is unique among COI in that it doesn't specify the 
addressee but the speaker.  I think words like {pi'u} only apply to 
sets, not masses (similarly, {lu'o} doesn't make sense applied to a 
mass, nor {lu'a} to an individual).  All kinds of PA words are very 
loosely jammed into "number" (like ci'i, fi'u, ma'u/ni'u... they all 
make sense in some situations, but what's one to make of the 
perfectly grammatical number {pa ci'i ci'i pa ma'u ma'u pa pa ni'u 
ma'u rau mo'a ka'o ka'o pai pai pi pi pi pi su'o}?) selma'o are 
syntactic groupings, and they can contain things that allow for 
semantic nonsense, like {li .abu ju'u vo}.  That happens, and it's 
not the grammar's fault, nor does it lie with the grammar to fix it.

~mark