Post by Steve Post by email@example.com Post by Steve Post by firstname.lastname@example.org
If it can print o-dieresis properly without the hollow attribute,
shouldn't it also be able to print it with the hollow attribute?
PostScript is a complete programming language, so it can be
very complex, and hard to debug. Your "hollow attribute" is
not a built-in feature of PostScript (not a language primative).
So you will have to see where it is defined in the document and
place some debugging code there.
"Hollow" is what DeScribe calls it. In the OS/2 Font Palette
editor, there is an emphasis called "Outline" that appear to
do the same thing.
Okay, there is a PaintType attribute that describes an outline
action. It says that a font that is designed to be "filled" can
be "outlined" instead. PaintType can be part of a font definition.
PaintType does appear in the PostScript file, several times. After
the first occurrence, a few lines later there are 211 lines of
hexadecimal, which could be defining things. The second occurrence
is embedded within a longer line. The third, fourth, and fifth
occurrence are also followed by extensive lines of hexadecimal.
Post by Steve Post by email@example.com
The only thing that is obvious in the Postscript file is that
the text with and without the hollow attribute is preceded with
478 /HelmetCondensed-Bold1 SF
478 /HelmetCondensed-Bold SF
In the "Blue Book", "PostScript Language Tutorial and Cookbook",
Adobe Systems Incorporated, there is a routine that creates an outline
font from a filled font (page 203). So the above makes sense If both
fonts are in both documents. Not that there are not fonts that are not
filled. To be filled, an outline must exist defining a glyph. The Blue
Book and the Language Reference can be downloaded as PDF's.
I had already found a file PLRM2.pdf, which I presume stands for
Postscript Language Reference Manual. It's 772 pages long, and I
haven't had time to read through all of it yet. Nor am I terribly
motivated to do so, just to track down a driver bug, because I
wouldn't know who could fix the driver itself.
Post by Steve
Some fonts are based on lines that are stroked. Think 'Courier' where
all the lines that make up a glyph have the same width. And some
fonts are made up of bitmaps. They cannot be filled or outlined either.
And there are a bunck of fonts that are repackaged True Type fonts.
The outline information may not be available? And there are/were
encrypted fonts that can't be modified.
So, examine your font and find out if it is the right type to have an
outline that can be accessed to create a new font. (The fact that you
are having problems implies...)
HelmetCondensed is a sans serif font, not too dissimilar to
Helvetica Narrow. Nothing fancy. The 7-bit ASCII characters
all work fine in their "hollow" or "outline" form, and o-dieresis
isn't all that different from an o or an i.
Post by Steve
Can you use a standard font to create your document? That may be
a way to get an acceptable result.
I'm going to experiment with UniversCondensed. If memory serves, way
back when I had an HP Laserjet III, it came with CG Times and Univers
built-in as the serif and sans serif fonts. DeScribe recognizes
Univers as both a printer and screen font. UniversCondensed is a
little bit wider than HelmetCondensed. I didn't like the aspect ratio
of Helvetica Narrow. And while DeScribe allows the user to control
the "Set Width" independently of the point size, thus having full
control over the aspect ratio, that capability is available only for
regular text, not a text graphic. The advantage of a text graphic is
that you have control over the color of the text, whereas document
text is restricted to black and white.