Levi Stephen wrote:
> The annotation shall be the Statement
> of the CFJ and the judgement of that CFJ.
I prefer the feature of the present arrangement that the body of the
annotation is a well-formed sentence which is to be interpreted as true.
So for CFJs that are judged TRUE, the best annotation is simply the
statement. If you want to make the construction of annotations for
FALSE CFJs purely mechanical, how about:
If the CFJ was judged TRUE, the annotation shall be the
statement of the CFJ. If the CFJ was judged FALSE, the
annotation shall be a quotation of the statement of the CFJ
followed by "is false.".
But actually, even with the latitude to issue annotation orders for
FALSE CFJs, hardly any judges actually issue such orders. After more
than thirteen years of the annotation system, we have only four current
official annotations. They are, and as far as I recall almost always
have been, massively outnumbered by the unofficial annotations.
In fact, very few CFJs are worded such that the statement would make a
suitable annotation. Take CFJ 1631, for example: it concerns whether the
"Subject:" header contributes to the meaning of a public message, but its
statement is the utterly opaque "comex registered on or about Sat, 28 Apr
2007 22:00:01 -0400". I turned the CFJ into an unofficial annotation on
rule 478, expressed in the more useful form regarding the general case.
An order to annotate couldn't have mandated a useful annotation here.
So I think there are two sensible options for annotations here. Firstly,
allow the judge making the order to decide what sentence to annotate
the rule with. E can annotate the ruleset with the general principles
e has based eir findings on, not restricted to the statement of the CFJ.
Secondly, repeal rule 789 altogether, and leave annotations entirely to
the rulekeepor's discretion (as, in practice, most of them already are).
The official annotation mechanism isn't useful enough to justify the
space it takes up.
-zefram
