Subject: Re: empty blocks vs. zero-length runs
From: Jesper Skov (jskov@redhat.com)
Date: Sun Jun 11 2000 - 13:48:07 CDT
>>>>> "Jesper" == Jesper Skov <jskov@redhat.com> writes:
>>>>> "Paul" == Paul Rohr <paul@abisource.com> writes:
Paul> At 02:01 PM 6/5/00 -0700, Paul Cubbage wrote: Could we prevent
Paul> all this trouble by guaranteeing that *no* level of the layout
Paul> hierarchy is ever totally empty?
>>> What then? Phony data in levels?
Paul> No.
Jesper> [snip]
Jesper> I haven't had much time to look at the code involved - and it
Jesper> scares me a bit :) But I think we'd want to add a new Run
Jesper> class which the Block code knows about - so when Runs are
Jesper> added to a line it always happens before the special
Jesper> Run. Special handling will also be needed for selection and
Jesper> point placement calculation.
Jesper> More as I dig into it - probably next weekend.
FYI, I did a very quick hack to the block code and was surprised to
see the end-of-paragraph block treated properly with regards to
insertion of Runs in the block (i.e., typed data ended up before the
special e-o-p Run). Still a ton of changes required for a proper
implementation, but I may actually have enough of a clue to implement
this - which I didn't quite expect, to be honest :)
However, I'll wait till we're past 0.7.10. I'll try to squash some
bugs until then for entertainment.
Cheers,
Jesper
This archive was generated by hypermail 2b25 : Sun Jun 11 2000 - 13:48:19 CDT