colonel@sunybcs.UUCP (George Sicherman) (01/03/84)
For some reason, when tbl writes the actual table entries, it surrounds them with delimiters. Example: input: .TS l l. alpha (T) beta .TE output: ... \&\h'...'^B^Calpha^B\h'...'^Bbeta^C^B ... Does anybody know why? Obviously the delimiters should not be there. Troff badly misplaces the -me special characters because of these spurious delimiters. And of course, if your entries contain all the troff-acceptable control characters, tbl will choose graphics for delimiters. Try it! The fix ought to be to change t8.c to just not print the delimiters. But will that kill something? G. L. Sicherman ...seismo!rochester!rocksvax!sunybcs!colonel
jaap@haring.UUCP (Jaap Akkerhuis) (01/05/84)
<Well known line> \&\h'...'^B^Calpha^B\h'...'^Bbeta^C^B ... ... Obviously the delimiters should not be there. The fix ought to be to change t8.c to just not print the delimiters. But will that kill something? Well, that will kill tbl completely... Tbl uses the field capabilities of troff. The ^B and the ^C are just the padding indicators and field delimiters, so they have to be there. (See chapter 9 of N/Troff User's Manual, for an explanation of the field mechanism). Troff badly misplaces the -me special characters Do you mean the characters like u\*: (chapter 13 -me reference manual)? Well they work only properly if: 1) You have the Berkeley version of n/troff 2) The definitions of string like \*: is matched to your output device.