Re: [Groff] Re: Mysteries of .em request

2009-04-15 Thread Werner LEMBERG

> I've long been baffled by the .em macro, not knowing how to get it
> to emit more than one final page, disastrous for split tables,
> floats, indexes and so on.  I have long resorted to an explicit user
> end document macro, using .em only to ensure it is given.

My documentation is still not fine (actually, I have yet to debug why
`.pl +1v' `.ne 2' behaves differently to `.ne 1'), but my summary of
the whole issue is that I'll probably implement an `.em1' request
which doesn't have this silly restriction -- or can someone name any
reason for this bizarre behaviour of `.em'?

> Your exposition is very instructive, but who would have guessed
> this...

Well, it's mentioned in the original troff documentation.  I thought
I've read it very carefully, but this has completely missed my
attention.


Werner




Re: [Groff] Re: Mysteries of .em request

2009-04-15 Thread Larry Kollar


Werner LEMBERG wrote:


... my summary of
the whole issue is that I'll probably implement an `.em1' request
which doesn't have this silly restriction -- or can someone name any
reason for this bizarre behaviour of `.em'?


I'm guessing the original intent and assumption for .em was to make  
sure the last footer (and perhaps endnotes) got printed, and there  
was no incentive to have it do anything different.


The .em1 request sounds like a good compromise, just in case there  
*are* people who depend on the restrictions of the original.


-- Larry