Am 10.03.16 um 14:12 schrieb Hans Hagen:
On 3/10/2016 12:52 PM, Frank Mittelbach wrote:
whereas i think you always report
<actual looseness achieved>
basically a global variable is reported that kept track of the last
found looseness
so I thought :-)
only that is a variable that isn't set
it's more that it only got set when there is a check for looseness which
is not always happening (as it depends on other parameters as well) so
then you get the old value (so that state is/was preserved)
that's what I meant with not set (or rather reset) ... basically the
internal code is largely original Knuth only that you use the Knuthian
globals also for reporting but then you you need the extra step that Don
optimized out
well, the internal code is also a bit ugly in passing around info (and
by now i've gotten accustomed to extra housekeeping at the lua end)
as I wouldn't know ... starting from the internals of Don's "spaghetti"
programming (as he called it himself once :-) ) is not that easy
frank