Re: [Groff] Failed to pick a low-hanging (?) fruit

[Sigfrid Lundberg, NetLab]

On Tue, 5 Mar 2002, Werner LEMBERG wrote:

...

> It should be rather easy to to integrate your dvi-specific stuff into
> the current version of pspic.tmac

I'd love to see them there.

> -- the only thing I don't understand are the following two lines:
>
>   .ds hoffset hoffset=-58 \"\n[ps-offset]u
>   .ds voffset voffset=-48 \"\n[ps-offset]u
>
> what are those hard-coded values good for?

They are no good... I've always had problems with the arithmetics and
scaling in *roff -- and still have. I guess that something went wrong, and
I solved it by hard-coding.

Actually, I've been experimenting with moving this into pspic.tmac
The solution should contain something like (where I just cannot figure out
what something[1..3] should be

.ds boundingbox llx=\\n[llx] lly=\\n[lly] urx=\\n[urx] ury=\\n[ury]
.ds hoffset  hoffset=\\n[something1]
.ds voffset  voffset=\\n[something2]
.ds width    hsize=\\n[something3]
.ds height   vsize=\\n[something4]
.
\h'\\n[ps-offset]u'\
\v'\\n[ps-desht]z'\
\X'psfile=\\$1 \\*[boundingbox]' \\*[heigth] \\*[width] \\*[hoffset] 
\\*[voffset] clip'

cf http://www.tug.org/dvipsk/dvips_32.html

Although the dvips documentation doesn't say anything about llx et al,
but they are present in the tex 'specials' I have on my computer, and
dvips complains without them.

I came as far today, that dvips actually included my eps, but it didn't
show up on ghostscript :/


Sigge


>     Werner
>