I have to agree with aionys here. =A0As you correctly poin= t, the CLL says "that an already-filled place is skipped". = =A0"fi ku" is not "filled", however. =A0It is empty. = =A0It's merely mentioned that a slot exists, but it does not fill it. = =A0Hence "FA KU" are =A0basically semantically meaningless struct= ures, unlike FA zo'e=A0

=A0 There are, however, of course,= lotsa bugs with jboski (why, oh why, did genrei have to go?), such as not = liking LI structures at the end of an utterance that are not terminate with= lo'o (i.e. "mi mitre li re" will bomb)

=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 --gejyspa

On = Sun, Apr 7, 2013 at 7:00 AM, guskant <gusni.kantu@gmail.com> wrote:

Le dimanche 7 avril 2013 19:34:31 UT= C+9, aionys a =E9crit=A0:

On Sun, Apr 7, 2013 at 4:2= 1 AM, guskant <gusni...@gmail.com> wr= ote:

Le dimanche 7 avril 2013 16:10:25 UTC+9, la = gleki a =E9crit=A0:
On Sunday, April 7, 2013 10:49:03 AM UTC+4, aionys wrote:
On = Sun, Apr 7, 2013 at 12:32 AM, la gleki <gleki.is...= @gmail.com> wrote:

On Sunday, April 7, = 2013 1:29:55 AM UTC+4, guskant wrote:
Le dimanche 7 avril 2013 06:12:29 UTC+9, LordDraqo a =E9crit=A0:
Did you try "f= i la lojban ku tavla"?

I don't need to try it,= but {fi la lojban ku tavla} means "talk about Lojban". I meant {= fiku} with no sumti should be x3, and a terbri next to it should be x4, but= jbofi'e doesn't parse it as I expected.

I was wondering about that in past a= s well. I can't find that thread now but the funny thing is that=A0
{fizo'e la lojban tavla} does what you need.

so does

{fo la.lojban. tavla}

<= /div>and

{la.lojban. ve tavla}

I guess guskant is searching for stylistic div= ersity which is indeed present here.

Like
in {fizo'e la lojban tavla} {la lojban} is tavla4.<= /div>
But in {fizo'e tavla la lojban} {la lojban} is now tavla2 !!!=

BTW, here is the old thread=A0= https://groups.google.com/d/topic/lojban/ZBpL4xoUwgQ/<= u>discussion

ki'e gleki for the info. M= y problem is however a bit different from yours.

From your problem, = I see jbofi'e forgets the places assigned by FA when a selbri appears, = and after the selbri, the default place structure is applied. I would like = to say it bothers mi'a who are not familiar with European languages and= should be a bug, but my actual point is something other.

According to jbofi'e, {fi ku la lojban tavla} means "Lojban is= talker."
Even if {fi ku} doesn't mean /anything/, it is not a = valid reason for ignoring the usual rule "that after a FA-tagged sumti= , any sumti following it occupy the places numerically succeeding it, subje= ct to the proviso that an already-filled place is skipped" described i= n CLL9.3.
Even {fi zi'o la lojban tavla} is parsed as "talk in Lojban" = that I expected. Why not {fi ku}?

I would s= ay, because {fi ku} is /not/ a FA-tagged sumti. Which it isn't. It isn&= #39;t a sumti at all.