On Word's Formatting Toolbar, the drop-down list normally displays the font
names using the font itself for its name.  For Symbol and Wingdings fonts,
this would be unintelligible, so the name is spelled out in some standard
font.  I have added a symbol-type font to my system, but the drop-down
attempts to use this new font for the name.  How can I make it behave just
like Wingdings?  Running Word 2000 under W98SE.

"George B"  wrote:
> On Word's Formatting Toolbar, the drop-down list normally displays the 
> font
> names using the font itself for its name.  For Symbol and Wingdings fonts,
> this would be unintelligible, so the name is spelled out in some standard
> font.  I have added a symbol-type font to my system, but the drop-down
> attempts to use this new font for the name.  How can I make it behave just
> like Wingdings?  Running Word 2000 under W98SE.


Hi George,

Unless the font gives that information, Word can't know if it's dealing with 
a symbol font, or a "regular" font.
Likely your font doesn't contain that information. Especially freeware 
symbol fonts are often problematic in this respect.

You can check that:
Select some character in that font, and open "Insert > Symbol".
Now click on the button for defining a keyboard shortcut.

For a symbol font (like "Wingdings"), you should see under "Description" the 
name of the font, followed by a 5-digit number starting with "6".
For regular fonts, you'll likely see a shorter 3- or 4-digit number.
Word uses codes between 61472 and 61695 for characters from symbol fonts.

The only "solution" would be to patch the font file in a font editor. But 
that's probably too much trouble to go through, and maybe problematic 
legally.

Greetings,
Klaus

Thanks for your explanation.  I knew that the font name was not in the font
file, but never would have guessed that font numbers were used as
identification.

"George B"  wrote:
> Thanks for your explanation.  I knew that the font name was not in the
> font file, but never would have guessed that font numbers were used
> as identification.

Well, the font name certainly is in the font file.

The problem, as I understand it, that fonts should contain a Panose 
description of the font.
In that description, a symbol font should be marked as "decorative".
Font designers, on the other hand, usually call fonts that look ornamental 
"decorative".

So fonts with lots of flourishes often (erroneously) have the Panose 
description "decorative" in them and are treated as symbols by Word 
(resulting in text that's not hyphenated or spell-checked), while many 
symbol fonts don't have the proper "decorative" setting.

Klaus

Klaus Linke wrote:

> "George B"  wrote:
> 
>>Thanks for your explanation.  I knew that the font name was not in the
>>font file, but never would have guessed that font numbers were used
>>as identification.
> 
> 
> Well, the font name certainly is in the font file.
> 
> The problem, as I understand it, that fonts should contain a Panose 
> description of the font.
> In that description, a symbol font should be marked as "decorative".
> Font designers, on the other hand, usually call fonts that look ornamental 
> "decorative".
> 
> So fonts with lots of flourishes often (erroneously) have the Panose 
> description "decorative" in them and are treated as symbols by Word 
> (resulting in text that's not hyphenated or spell-checked), while many 
> symbol fonts don't have the proper "decorative" setting.
> 
> Klaus 

The Panose family classification of "Latin Decorative" is for text 
fonts that  are designed more for impact than readability. - so fonts 
with flourishes are properly "decorative". (Ornaments would not be). 
The "decorative" designation has nothing to do with "symbol" fonts, 
and I doubt that MS Word is so incorrect to use that particular flag 
to differentiate.

There are, however, a number of places where a font may be defined as 
being symbolic.  - I have no idea which ones are used by which 
applications or operating systems for what purposes!  These include:

IBM Class  (includes serif, script, ornamentals, symbolic)
PCL ID (a long list of font names, including PiFont)
VP-ID  (another long list of font names and classes, including symbol)
MS-ID (Roman, Swiss, Modern, Script, and Decorative)
Encoding and Unicode (includes "Symbol Character Set")
Unicode Ranges (Some specific sets of symbols have assigned ranges)
In the PCLT Table (One of the character set options is "Adobe Symbol")

There may be others!

Hi,

You seem to know a lot more about the technical specs than I do. The 
relevant point is that some fonts give wrong info.

I don't remember where I read about the "decorative" misunderstandings... a 
quick Google check seems to show it might not have been completely off the 
mark:
http://www.microsoft.com/typography/otspec/recom.htm
"If the font is a symbol font, the first byte of the PANOSE value must be 
set to 'decorative.' "

OTOH, maybe I did read something that wasn't really true, or misunderstood 
it.

Greetings,
Klaus



>> "George B"  wrote:
>>
>>>Thanks for your explanation.  I knew that the font name was not in the
>>>font file, but never would have guessed that font numbers were used
>>>as identification.
>>
>>
> Klaus Linke wrote:
>
>> Well, the font name certainly is in the font file.
>>
>> The problem, as I understand it, that fonts should contain a Panose 
>> description of the font.
>> In that description, a symbol font should be marked as "decorative".
>> Font designers, on the other hand, usually call fonts that look 
>> ornamental "decorative".
>>
>> So fonts with lots of flourishes often (erroneously) have the Panose 
>> description "decorative" in them and are treated as symbols by Word 
>> (resulting in text that's not hyphenated or spell-checked), while many 
>> symbol fonts don't have the proper "decorative" setting.
>>
>> Klaus
>

"Character"  wrote:
> The Panose family classification of "Latin Decorative" is for text fonts 
> that  are designed more for impact than readability. - so fonts with 
> flourishes are properly "decorative". (Ornaments would not be). The 
> "decorative" designation has nothing to do with "symbol" fonts, and I 
> doubt that MS Word is so incorrect to use that particular flag to 
> differentiate.
>
> There are, however, a number of places where a font may be defined as 
> being symbolic.  - I have no idea which ones are used by which 
> applications or operating systems for what purposes!  These include:
>
> IBM Class  (includes serif, script, ornamentals, symbolic)
> PCL ID (a long list of font names, including PiFont)
> VP-ID  (another long list of font names and classes, including symbol)
> MS-ID (Roman, Swiss, Modern, Script, and Decorative)
> Encoding and Unicode (includes "Symbol Character Set")
> Unicode Ranges (Some specific sets of symbols have assigned ranges)
> In the PCLT Table (One of the character set options is "Adobe Symbol")
>
> There may be others!
>
>
>
>
>
>
>
>
>

A lot of the problematic fonts seem to have been created with Fontographer.
They have a tech note for creating symbol fonts:
~~~~~~~~~~~~~~~~~~~~~~
#12312: Creating Symbol fonts for Windows
PC Symbol encoding is normally used for fonts which must work with 
application features such as
Microsoft Word's "Insert Symbol".
Follow these steps to create a PC Symbol encoded font:
1 If building from an existing font, open the font with original encoding. 
See Encoding Vectors FAQ for details.
2 Use the File > Generate Font Files, "Advanced" dialog to set the encoding 
to Custom.
3 Select "Decorative" under the PC Family pop-up menu on the right.
4 Click on the Truetype Options button and select the "Use Symbol Encoding" 
check box.
5 Select the "Character Mapping" pop-up menu and set it to "Sequential."
The font will now respond to applications which use "Insert Symbol."
This procedure is the same for both Macintosh and Windows Fontographer in 
versions 4.1 and above.
Additional Information
Be advised that PC Symbol encoding was not intended to work with Unicode 
keyboard drivers.
~~~~~~~~~~~~~~~~~~~~~~

Klaus Linke wrote:

> Hi,
> 
> You seem to know a lot more about the technical specs than I do.

That was just searching through FontLab's font property options.

 > The relevant point is that some fonts give wrong info.

Absolutely!!

> I don't remember where I read about the "decorative" misunderstandings... a 
> quick Google check seems to show it might not have been completely off the 
> mark:
> http://www.microsoft.com/typography/otspec/recom.htm
> "If the font is a symbol font, the first byte of the PANOSE value must be 
> set to 'decorative.' "

Interesting, and I think that Microsoft may be in error here (it 
wouldn't be the first time they've made a typographical mistake). In 
addition, the referenced URL in that paragraph (to the Panose spec at 
Fonts.com) is no longer valid. Here's a link to the official Panose spec:

http://www.panose.com/ProductsServices/pan5.aspx

where it states:
"Latin Symbol is where all the nonalphabetic fonts reside. These are 
fonts that can be loaded like normal text fonts, but do not contain 
readable characters. Dingbats and specialized symbol fonts are two 
examples."

"Decorative" is certainly better than 4 of the other six possible 
values (Any, No Fit, Latin Text, and Latin Handwritten"), but the 
first-byte value named "Latin Symbol" or in some listings "Pictorial" 
is pretty clearly the value intended for such fonts.

If MS' applications are actually using that value to determine how to 
process symbol fonts (and the paragraph you cite could indicate that 
they do), that would explain some of the historical difficulties 
people have with MS Word in working with symbols! Especially when 
combined with your first point about many, if not most, fonts having 
erroneous or missing information.

The Panose value is theoretically used to find "similar" fonts so the 
system can make an appropriate substitution, and is also used to group 
fonts - for instance XP allows you to "list fonts by similarity" in 
the installed fonts directory.

In FontLab, the panose values are cross-linked to other font 
properties - if you change the property value, the panose value 
changes, and if you manually change the panose value, the property 
name is changed in the appropriate field.

> OTOH, maybe I did read something that wasn't really true, or misunderstood 
> it.
> 
> Greetings,
> Klaus
> 
> 
> 
> 
>>>"George B"  wrote:
>>>
>>>
>>>>Thanks for your explanation.  I knew that the font name was not in the
>>>>font file, but never would have guessed that font numbers were used
>>>>as identification.
>>>
>>>
>>Klaus Linke wrote:
>>
>>
>>>Well, the font name certainly is in the font file.
>>>
>>>The problem, as I understand it, that fonts should contain a Panose 
>>>description of the font.
>>>In that description, a symbol font should be marked as "decorative".
>>>Font designers, on the other hand, usually call fonts that look 
>>>ornamental "decorative".
>>>
>>>So fonts with lots of flourishes often (erroneously) have the Panose 
>>>description "decorative" in them and are treated as symbols by Word 
>>>(resulting in text that's not hyphenated or spell-checked), while many 
>>>symbol fonts don't have the proper "decorative" setting.
>>>
>>>Klaus
>>
> 
> "Character"  wrote:
> 
>>The Panose family classification of "Latin Decorative" is for text fonts 
>>that  are designed more for impact than readability. - so fonts with 
>>flourishes are properly "decorative". (Ornaments would not be). The 
>>"decorative" designation has nothing to do with "symbol" fonts, and I 
>>doubt that MS Word is so incorrect to use that particular flag to 
>>differentiate.
>>
>>There are, however, a number of places where a font may be defined as 
>>being symbolic.  - I have no idea which ones are used by which 
>>applications or operating systems for what purposes!  These include:
>>
>>IBM Class  (includes serif, script, ornamentals, symbolic)
>>PCL ID (a long list of font names, including PiFont)
>>VP-ID  (another long list of font names and classes, including symbol)
>>MS-ID (Roman, Swiss, Modern, Script, and Decorative)
>>Encoding and Unicode (includes "Symbol Character Set")
>>Unicode Ranges (Some specific sets of symbols have assigned ranges)
>>In the PCLT Table (One of the character set options is "Adobe Symbol")
>>
>>There may be others!
>>
>>
>>
>>
>>
>>
>>
>>
>>
> 
> 
>

Klaus Linke wrote:
Sorry, didn't see this note until I responded to the previous one.

It seems to validate the point that MS products use the value as you 
first indicated they do.  Erroneously, in my opinion, but after 15 
years it's probably cast in concrete :(


> A lot of the problematic fonts seem to have been created with Fontographer.
> They have a tech note for creating symbol fonts:
> ~~~~~~~~~~~~~~~~~~~~~~
> #12312: Creating Symbol fonts for Windows
> PC Symbol encoding is normally used for fonts which must work with 
> application features such as
> Microsoft Word's "Insert Symbol".
> Follow these steps to create a PC Symbol encoded font:
> 1 If building from an existing font, open the font with original encoding. 
> See Encoding Vectors FAQ for details.
> 2 Use the File > Generate Font Files, "Advanced" dialog to set the encoding 
> to Custom.
> 3 Select "Decorative" under the PC Family pop-up menu on the right.
> 4 Click on the Truetype Options button and select the "Use Symbol Encoding" 
> check box.
> 5 Select the "Character Mapping" pop-up menu and set it to "Sequential."
> The font will now respond to applications which use "Insert Symbol."
> This procedure is the same for both Macintosh and Windows Fontographer in 
> versions 4.1 and above.
> Additional Information
> Be advised that PC Symbol encoding was not intended to work with Unicode 
> keyboard drivers.
> ~~~~~~~~~~~~~~~~~~~~~~ 
> 
>

The other day I was reviewing my article at
http://word.mvps.org/FAQs/General/InsertSpecChars.htm and cringing at the
amount of rewriting needed. The article was written when I was using Word 97
and then-Webmaster/editor Dave Rado was using Word 2000, so we cobbled it
together between us, and I think perhaps it's been updated in some places
since. I'm not sure where I got the information about "decorative" fonts,
but regardless of what bit it is that Word uses, it is the bit that used to
determine whether a font was listed in the Insert Symbol dialog or not, and
it was clear at the time that a lot of fonts were coded incorrectly because
many non-symbol fonts appeared in the Symbol dialog and some symbol fonts
did not appear. That problem was "corrected" in a later version (possibly
even Word 2000) by just listing them all.

Being listed in the Symbol dialog or not was the big issue then, and I don't
think we had fully discovered the horrors of the problem as it affects line
breaks, hyphenation, and other behavior of the fonts.

-- 
Suzanne S. Barnhill
Microsoft MVP (Word)
Words into Type
Fairhope, Alabama USA
Word MVP FAQ site: http://word.mvps.org
Email cannot be acknowledged; please post all follow-ups to the newsgroup so
all may benefit.

"Character"  wrote in message
news:jnK4g.293951$iS4.10868@fe05.news.easynews.com...
> Klaus Linke wrote:
>
> > Hi,
> >
> > You seem to know a lot more about the technical specs than I do.
>
> That was just searching through FontLab's font property options.
>
>  > The relevant point is that some fonts give wrong info.
>
> Absolutely!!
>
> > I don't remember where I read about the "decorative"
misunderstandings... a
> > quick Google check seems to show it might not have been completely off
the
> > mark:
> > http://www.microsoft.com/typography/otspec/recom.htm
> > "If the font is a symbol font, the first byte of the PANOSE value must
be
> > set to 'decorative.' "
>
> Interesting, and I think that Microsoft may be in error here (it
> wouldn't be the first time they've made a typographical mistake). In
> addition, the referenced URL in that paragraph (to the Panose spec at
> Fonts.com) is no longer valid. Here's a link to the official Panose spec:
>
> http://www.panose.com/ProductsServices/pan5.aspx
>
> where it states:
> "Latin Symbol is where all the nonalphabetic fonts reside. These are
> fonts that can be loaded like normal text fonts, but do not contain
> readable characters. Dingbats and specialized symbol fonts are two
> examples."
>
> "Decorative" is certainly better than 4 of the other six possible
> values (Any, No Fit, Latin Text, and Latin Handwritten"), but the
> first-byte value named "Latin Symbol" or in some listings "Pictorial"
> is pretty clearly the value intended for such fonts.
>
> If MS' applications are actually using that value to determine how to
> process symbol fonts (and the paragraph you cite could indicate that
> they do), that would explain some of the historical difficulties
> people have with MS Word in working with symbols! Especially when
> combined with your first point about many, if not most, fonts having
> erroneous or missing information.
>
> The Panose value is theoretically used to find "similar" fonts so the
> system can make an appropriate substitution, and is also used to group
> fonts - for instance XP allows you to "list fonts by similarity" in
> the installed fonts directory.
>
> In FontLab, the panose values are cross-linked to other font
> properties - if you change the property value, the panose value
> changes, and if you manually change the panose value, the property
> name is changed in the appropriate field.
>
> > OTOH, maybe I did read something that wasn't really true, or
misunderstood
> > it.
> >
> > Greetings,
> > Klaus
> >
> >
> >
> >
> >>>"George B"  wrote:
> >>>
> >>>
> >>>>Thanks for your explanation.  I knew that the font name was not in the
> >>>>font file, but never would have guessed that font numbers were used
> >>>>as identification.
> >>>
> >>>
> >>Klaus Linke wrote:
> >>
> >>
> >>>Well, the font name certainly is in the font file.
> >>>
> >>>The problem, as I understand it, that fonts should contain a Panose
> >>>description of the font.
> >>>In that description, a symbol font should be marked as "decorative".
> >>>Font designers, on the other hand, usually call fonts that look
> >>>ornamental "decorative".
> >>>
> >>>So fonts with lots of flourishes often (erroneously) have the Panose
> >>>description "decorative" in them and are treated as symbols by Word
> >>>(resulting in text that's not hyphenated or spell-checked), while many
> >>>symbol fonts don't have the proper "decorative" setting.
> >>>
> >>>Klaus
> >>
> >
> > "Character"  wrote:
> >
> >>The Panose family classification of "Latin Decorative" is for text fonts
> >>that  are designed more for impact than readability. - so fonts with
> >>flourishes are properly "decorative". (Ornaments would not be). The
> >>"decorative" designation has nothing to do with "symbol" fonts, and I
> >>doubt that MS Word is so incorrect to use that particular flag to
> >>differentiate.
> >>
> >>There are, however, a number of places where a font may be defined as
> >>being symbolic.  - I have no idea which ones are used by which
> >>applications or operating systems for what purposes!  These include:
> >>
> >>IBM Class  (includes serif, script, ornamentals, symbolic)
> >>PCL ID (a long list of font names, including PiFont)
> >>VP-ID  (another long list of font names and classes, including symbol)
> >>MS-ID (Roman, Swiss, Modern, Script, and Decorative)
> >>Encoding and Unicode (includes "Symbol Character Set")
> >>Unicode Ranges (Some specific sets of symbols have assigned ranges)
> >>In the PCLT Table (One of the character set options is "Adobe Symbol")
> >>
> >>There may be others!
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >
> >
> >

> http://word.mvps.org/FAQs/General/InsertSpecChars.htm


Hi Suzanne,

Yes, that may well be where I picked up that stuff about decorative fonts 
originally.
I can't get my head around the Panose specs (or at least gave up for the 
time being).

Most fonts making problems (either latin text fonts acting like symbol 
fonts, or the other way round) really seem to be the work of amateurs (in 
the good sense of the word) using Fontographer. It's been many years since 
I've seen a font by a big commercial typography foundry that showed such 
problems.

Regards,
Klaus

Suzanne S. Barnhill wrote:
> The other day I was reviewing my article at
> http://word.mvps.org/FAQs/General/InsertSpecChars.htm and cringing at the
> amount of rewriting needed. The article was written when I was using Word 97
> and then-Webmaster/editor Dave Rado was using Word 2000, so we cobbled it
> together between us, and I think perhaps it's been updated in some places
> since. I'm not sure where I got the information about "decorative" fonts,
> but regardless of what bit it is that Word uses, it is the bit that used to
> determine whether a font was listed in the Insert Symbol dialog or not, and
> it was clear at the time that a lot of fonts were coded incorrectly because
> many non-symbol fonts appeared in the Symbol dialog and some symbol fonts
> did not appear. That problem was "corrected" in a later version (possibly
> even Word 2000) by just listing them all.
> 
> Being listed in the Symbol dialog or not was the big issue then, and I don't
> think we had fully discovered the horrors of the problem as it affects line
> breaks, hyphenation, and other behavior of the fonts.
> 

As Mr. Spock would say, "Fascinating"! And by the way, that's an 
excellent article!

This raises a purely academic question: If a symbol/pictorial font is 
created according to the Panose spec with a "5" as the first byte, 
indicating Latin Symbol, instead of the recommended "4" for 
Decorative, would those older versions of Word recognize it as a 
symbol font?  It depends on how they tested that byte - if they looked 
for exactly "4" the answer would be no; if they looked for "greater 
than 3" or just looked for a bit in the 4 position then the answer 
would be yes.

> "Decorative" is certainly better than 4 of the other six possible values 
> (Any, No Fit, Latin Text, and Latin Handwritten"), but the first-byte 
> value named "Latin Symbol" or in some listings "Pictorial" is pretty 
> clearly the value intended for such fonts.

Symbol fonts all do seem to have a "5" ("Pictorial"), at least in current 
Word (XML) docs.
Not sure whether the OS/2 table mentioned is something apart from the Panose 
number, or if that has been a previous misunderstanding that has been 
remedied.

Regards,
Klaus

Just a quick note. I have _not_ read the entire thread so please don't 
take this as a comment on anything in particular. It is simply some 
fair-comment given that the OS/2 table was mentioned by Klaus.

The OS/2 table is often misunderstood. Because of its name it is 
sometimes assumed to be of no relevance to Windows systems. However, it 
is a compulsory table under the TrueType specification. Nevertheless, it 
is sometimes not included in some freeware/shareware/cloned fonts. 
Windows XP will not use the font if the table is missing.

As a matter of history, the name arises because of early Microsoft 
involvement in TrueType development before IBM and MS parted ways 
resulting in Windows 3.x as separate from OS/2. Note that memory dims. 
However, I think think I am substantially correct.

http://www.microsoft.com/typography/otspec/os2.htm

Tom
MSMVP
Windows Shell/User


"Klaus Linke"  wrote in message 
news:%23Ks1oL7aGHA.4564@TK2MSFTNGP03.phx.gbl...
:> "Decorative" is certainly better than 4 of the other six possible 
values
: > (Any, No Fit, Latin Text, and Latin Handwritten"), but the first-byte
: > value named "Latin Symbol" or in some listings "Pictorial" is pretty
: > clearly the value intended for such fonts.
:
: Symbol fonts all do seem to have a "5" ("Pictorial"), at least in 
current
: Word (XML) docs.
: Not sure whether the OS/2 table mentioned is something apart from the 
Panose
: number, or if that has been a previous misunderstanding that has been
: remedied.
:
: Regards,
: Klaus
:
: