Post & Fontographer 3.5
10 messages in this thread
Mike,
I've been trying to figure out why Fontographer versioin 3.5 generated
fonts won't work with Post 1.7. Have you noticed this problem, or heard
others complain about it? I've been comparing version 3.1, 3.3 and 3.5
Fontographer fonts and find a few differences but other font generating
programs use them too with success in POST. The diferences I've found are
using
/FontInfo 13 instead of
/FontInfo 9 dict dup begin
Encoding StandardEncoding def % where the older versions use
/Encoding 256 array
and the other difference I've noticed is this.
% ADL: 1100 150 0
I'm no Postscript guru, not even a good follower. I was wondering since
your in England, maybe you could ask Aylward to look over a Fontographer
v3.5 font and see what is causing the problem? If you need to find a font,
there are about fifteen font packages in Lib6 and I can tell you one to
choose that has 3.5, 3.3 and 3.2 and 3.1 versions, its called FONTS8.LZH.
Thanks,
Chris
Any line that begins with '%' is a comment and is ignored until end of
line, which can be CR or LF. Lines with '%%' are possibly structured
comments which can assist interpretation of the file, particularly to
programs that simply swish-and-spit EPS files, for example. The /FontInfo
sub-dictionary holds things like /version, /Notice, /FullName, /FamilyName,
/Weight, ItalicAngle, /isFixedPitch, /UnderlinePosition,
/UnderlineThickness. In one font, it looks like it's preparing to add 9
items, the other 13 items. That might be some of the problem… if one
font adds something the other doesn't/
John,
I'd sure like to figure out why Fontographer version 3.5 makes Post
spit out the font. I've prepared a list of all the Fontn.LZH files John has
uploaded so far. Fonts1.LZH thru Fonts15.LZH. I've listed the utility that
has generated each font. Maybe you could trap the culprit(s) in one of the
3.5 fonts? It might just be something simple a knowledgeable user could
change. Or it might be the actual binary is off?
Well look in Lib 6 for a file called FNTLST.LHA and check it out.
chris
Once upon a time, I decoded the method that Fontographer used to store
fonts, before the Type 1 spec was released. Are these fonts you're talking
about in Type 1 format, or the old Fontographer system of Type 3 with
sorta-hints?
John,
These fonts were apparently generated with version 3.5 of Fontographer.
That is AltSys' latest version. They're suppose to be Type 1 Fonts. They
work with Gold Disk's Fontmanager, and Soft-Logik's Pagestream 2.2HL which
both only use Type 1 version Postscript fonts to render outlines to screen.
The fonts in question can be found in Lib 6, named Fonts1.LZH thru
Fonts20.LZH, and there is a Font List in Lib 6 with the names of each font
and the utility that was used to generate them. It only lists up to archive
#15 but I'll update that list tonight or tomorrow.
Chris
John
These fonts are Type-1 format … and I have a Type-1 disassembler for the
purposes of examining the contents of the encrypted portion (I keep meaning
to upload it here, but keep forgetting … will upload just as soon as I've
finished this!!)
When you say you worked out how Fontographer stored fonts, do you mean you
cracked the Type-1 semi-compilation & double-encryption scheme? If so, I'm
significantly impressed!
Mike (from the UK)
No, I said I "cracked" the old Type 3 with Nimbus hinting that earlier,
pre-Black Book Typographer used. This feat was nothing more than
unravelling the code that was there; after all, in Type 3 you have no
protection from prying eyes, except for the unreadable obfuscation normally
possible in PostScript. 🙂
Chris
Well, the line beginning with % is simply a comment so should cause no
problems at all (lines beginning with %% are part of the document
structuring convention, and should have a specified syntax following the
%%, but they're still to be ignored by the interpreter)
The setting of the Encoding entry should also be perfectly okay since it's
a quick way of setting up an AdobeStandardEncoding without having to
specify it yourself. In other words … the /Encoding StandardEncoding def
does all the work that defining the array and then setting each entry one
at a time. I've looked at the source code for POST and it handles
StandardEncoding just fine.
The /FontInfo 13 thing is also fine … it's just the PostScript equivalent
of a DIM FontInfo() array definition (in effect) and all 3.5 is doing is
making a little more space.
However, I've examined the encrypted section of the Lilith font and I have
discovered that there is some unusual code in there. In _theory_ it should
still work, but it assumes the state of the stack and adds a comment. The
black-book specification for Type-1 fonts doesn't disallow this sort of
thing, but it does look very odd and it's possible that this is where POST
is falling over.
I'll continue investigating. Unfortunately, Aylward has stated he will no
longer support POST so I don't know if we'd get anywhere with him …
however, I may drop him a line.
Mike (from the UK)
Mike,
I'm glad you noticed some problems with the font. It sounds to me like
you may find a fix. I wonder if we could find out what changes were done to
version 3.5 of Fontographer? I'm going to try and call the publisher. Now
where'd I put my copy of MACwarehouse. They have a FAXFACTS no. that will
get press release type info sent to your FAX machine. Maybe I can find some
info on Fontographer. I have no idea who publishes that program.
Chris