Re: [NTG-context] bug or needs warning documentation

2013-10-13 Thread Hans Hagen
On 10/12/2013 2:15 PM, d.henman wrote: I found what I consider a bug in startcombination processing. it's actually a side effect of a tex property: \hbox {x} is the problem here. Makes me wonder if we should also ignore par tokens there (we already have an always \long variant in the

Re: [NTG-context] bug or needs warning documentation

2013-10-13 Thread Hans Hagen
On 10/13/2013 2:48 AM, Marco Patzer wrote: \unprotect \def\pack_combinations_pickup {\dostarttagged\t!combinationpair\empty \dostarttagged\t!combinationcontent\empty \dogotopar\pack_combinations_pickup_content_indeed} .. ok, patched ... but with a bit more efficient variant for

Re: [NTG-context] bug or needs warning documentation

2013-10-13 Thread Marco Patzer
On 2013–10–13 Hans Hagen wrote: On 10/13/2013 2:48 AM, Marco Patzer wrote: \unprotect \def\pack_combinations_pickup {\dostarttagged\t!combinationpair\empty \dostarttagged\t!combinationcontent\empty \dogotopar\pack_combinations_pickup_content_indeed} .. ok, patched ... but

Re: [NTG-context] bug or needs warning documentation

2013-10-13 Thread Hans Hagen
On 10/13/2013 11:26 PM, Marco Patzer wrote: How about switching paired boxes to \assumelongusagecs as well and getting rid of \dogotopar? Paired boxes are the only mechanism which uses \dogotopar. ok -

[NTG-context] bug or needs warning documentation

2013-10-12 Thread d.henman
I found what I consider a bug in startcombination processing. context version used: version: 2013.09.03 16:23 Examples: 1.) Good compilable source and 2.) source that fails The bad thing about this is that one, at least checks and checks syntax for errors and the error in fact is simply a

Re: [NTG-context] bug or needs warning documentation

2013-10-12 Thread Marco Patzer
On 2013–10–12 d.henman wrote: I found what I consider a bug in startcombination processing. […] The below fails \starttext \placefigure [here,none] [fig: reference tag] {citation text} { \startcombination[1*2]