At 08:09 PM 1/3/2002 +0100, Tobias Burnus wrote:
>Hi Hans,
>
> > Hi impatient Tobias ...
>;-) Well at the moment I'm really under pressure, but you seem to be under
>pressure as well.
>(Here the question was asked: LaTeX is still more stable, isn't it? And I
>have the feeling that this is not completely wrong ...)
it depends on what features you use;
- latex has a relatively small kernel compared to context; on the other
hand there is quite some code in context that can be considered stable and
not touched frequently
- you are one of the users using recent code and new features; these can
have an 'under construction' tag;
- some features - like tricky footnote stuff - will always be fragile due
to the way tex works; take footnotes inside boxes; i use some tricks to
move them to the outer level, but testing (and improving) this is a matter
of time; another problem with footnotes is that one has to apply some magic
in order to 'reprocess' them under a different layout regime (which happens
in the case of local footnotes or text/par notes).
- other code is stable but incomplete, for instance transparency support:
by using it i find the border cases and can improve things; as with fonts,
color is sensitive for local / mvl usage but i think that currently it works ok
- occasionally i'm speeding up some code, and there i hope that
compatibility is not broken
i have no experience with latex apart from the versions around 1990 but if
i look at the amount of problems reported at the pdftex list ... you
probably should compare context with latex+manypackages
> > i fixed that (aproximation btw); there are a few places in context where
> > multiple passes over content are done, i such cases one can use
> > \iftrialtypesetting to catch unwanted behavior. I kind of built it into
> > TABLE now, so your case should work;
>It does :-)
that's another thing: preventing redundant processing; take a box that ends
up on more pages that has toc entries embedded (this is a problem that is
related to working on spreads and i will solve it some day soon; again this
would be easier with a couple of low level tex features)
> > >c) I use TABLE with nr and nc. nr works ok, but for nc the text cannot
> > >extend the width of one cell (the border is usually drawn correctly)
> >
> > for that i need a minimal example (and time)
>here comes one. By the way the time problem applies also to me ...
>but I hope that I can work around that problem.
ok, keep in mind that the TABLE support is tricky in the sense that one has to
- do wild guesses about the content
- do prerolls to determine dimensions
- permit color, spacing etc
- support backgrounds, frames etc
- apply decimal alignment etc
all kind of progrwammed from scratch without help from tex's unidirectional
alignment features.
>Tobias
>
>Example:
>--------------
>\bTABLE
>\bTR
> \bTH foo \eTH
> \bTH This is a very, very, very really long cell \eTH
>\eTR
>\bTR
> \bTD[nc=2] This is also a very, very, very really long combined cell\eTD
>\eTR
>\eTABLE
>\bye
>----------------
>The first column is very wide, but need to be if the text of the [nc=2]
>cell would also be shown under the second column.
i'll have a look; looks like the span is ignored (you can set
option=stretch to get it working but that's not meant for that purpose); i
think something got lost when i added some config options; because there
are some preferemces/decissions involved, i want to make things
configurable (maybe strategoes some day). I'll send you a fix (please let
me know if it breaks other code).
Hans
-------------------------------------------------------------------------
Hans Hagen | PRAGMA ADE | [EMAIL PROTECTED]
Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
tel: +31 (0)38 477 53 69 | fax: +31 (0)38 477 53 74 | www.pragma-ade.com
-------------------------------------------------------------------------
