I said:

>>Does NetBSD really not have sigsetjmp?

and Chris Torek replied:

>It was broken in 4.4-Lite (at least the generic version -- a generic
>sigsetjmp is not possible, as it has to use the same frame as the
>setjmp).  I am not sure if it is fixed in NetBSD
>in general, though I think I fixed the sparc version before Lite shipped.

So what's the bottom line? Even though FreeBSD 2.x has a sigsetjmp 
declaration in setjmp.h, is it broken in FreeBSD too? What about the 
4.4Lite version of BSD/OS? Does it work on Intel and not on SPARC?

Like I said, this information needs to be captured in an X Consortium 
bug-report and sent to xbugs. If I said I would do it it runs the risk 
that I get some or all of it wrong. It would be much better if the people
who know and understand all the issues filed the bug-report.

>[lex.l changes for twm]
>>Your patch didn't apply or something. This was fixed in fix-11. Did
>>XFree86 3.1.1 undo it?

>I just looked at the `official' lex.l for twm.  It uses

>	#if BSD >= 199103 || ...

>This really ought to just be

>	#ifdef FLEX_SCANNER

>since it seems to be trying to detect systems that use flex.

Not sure what is meant by `official'. Fix-11 to X11R6 introduces the
symbol CSRG_BASED for the Net/2 and 4.4Lite based operating systems,
all of which are presumed to be using flex. If you have any questions
look at http://www.x.org/consortium/R6src/programs/twm/lex.l

--

Kaleb