Am 13.04.16 um 12:22 schrieb Hans Hagen:
i guess float calculations like these will never be roundtrip (and
opening up the engine exposes some of these phenomena)
quite possible (or at least probably difficult to avoid without a lot of
overhead)
so, a next release will be a bit more accurate but still not exact
roundtrip
well, if it would be explitily known that one can be off by X sp as a
maximum or even if it would be known that one can be off and that the
resulting value is not necessarily the same that you would get from the
real packer that would be extremely helpful --- ie explicitly stating
that in the manual and adjusting the example given there:
------------
This calling method takes glue settings into account and is especially
useful for finding the actual width of a sublist of nodes that are
already boxed, for example in code like this, which prints the width of
the space in between the a and b as it would be if \box0 was used as-is:
\setbox0 = \hbox to 20pt {a b}
\directlua{print (node.dimensions(
tex.box[0].glue_set,
tex.box[0].glue_sign,
tex.box[0].glue_order,
tex.box[0].head.next,
node.tail(tex.box[0].head)
)) }
-------------
as written now, it really claims "actual width" ... and that isn't true
frank
