[Fontinst] title graphic

Stephan Hennig mailing_list at arcor.de
Wed Jun 22 14:00:15 CEST 2005

Lars Hellström schrieb:
> At 20.51 +0200 2005-06-16, Stephan Hennig wrote:
>>Lars Hellström schrieb:
>>> At 19.52 +0200 2005-06-15, Stephan Hennig wrote:
>>>>(ii) Due to MetaObj
>>> Yes, that could be an issue. The MP code looks like it depends on a rather

>From your citation I thought you're critizizing the use of MetaObj.

>>So you prefer coding postscript then? :)
> No, but not because of the lack of high-level constructs. A couple of
> months ago I did some work on generating diagrams in PDF, and what bothered
> me most (in comparison with MetaPost) was the lack of equation-solving;
> that's such a wonderful feature! You apparently managed to do without it,
> however, so perhaps I should throw it back at you:
> So you prefer to micro-manage MetaPost then? :)

I think I now understand what you mean. When using MetaObj something like

input metaObj
test.e = test.c + 20bp*right;
test.c = origin;

isn't possible and MetaPost complains about an inconsistent equation
even with setObjectDefaultOption("Box")("fit")(false); .
So one has to build all those equally sized boxes in LaTeX (the other
thing that bothers me with MetaObj). Since I only worked with MetaObj
and never touched the boxes package before[1] I didn't know of boxes'
ability to control box dimensions. Reading all those north, south etc.
coordinates---yes; but it seemed natural to me that writing them is
forbidden. Hmm, I see I got slightly OT here but thanks for pointing me
at that direction!

You can find a new version of the graphic at
* It still uses LaTeX because MetaPost doesn't use kerning information
when printing strings.
* Self explanatory variable names are now used which makes drawing
connections more comfortable.

Stephan Hennig

[1] Since I know using MetaObj makes eps output bigger this graphic
seemed perfect for me to measure the bloat.

More information about the fontinst mailing list