I have failed to find a situation where ID4/Mac, namely the string ’Familyname StyleName’, is used. Where is it ever needed to be “nice and readable”.

Absolutely true.
If ID18/Mac is present in the name table, then application that I’ve tested will not show ID4/Mac nor identify a font by way of it. Exactly this is why — in another thread — I wondered if we couldn’t forget about ID18/Mac and abbreviate name ID4/Mac right away; the effect might be the same.
If however ID18/Mac is not present in the name table, then ID4/Mac will be shown. You will have noticed that the MakeOTF manual, p.6, also says that ’c1=’ (ID18/Mac) needs to be provided as soon as a ’c=’ (ID1/Win) was defined (to divide a larger family into a bunch of Windows-compatible 4-style families); this means: in some cases no ID18/Mac is, nor needs to be, present in the name table.

Even Adobe Brush Script Std which comes in just one weight defines ID1/Win and thus ID18/Mac (at least OTF Version 1.020). Would it not be a good standard to always provide ID1/Win, ID18/Mac, and let ID4/Mac identical to ID4/Win?

ID1/Win is always present in the name table. It is ID16/Win which will be added only if it differs from ID1/Win. (If you do not provide the compatible windows name in AFDKO, then the family name will go into ID1/Win rather than ID16/Win.)

Actually, to some degree assigning font names is also a matter of taste. What you suggest doesn’t look wrong, and there are fonts out there where ID4 is identical for Mac and Win. In general I would suggest to stick as close to common naming practices as possible, for the simple reason that they have proven to work.

If ID18/Mac is present in the name table, then application that I’ve tested will not show ID4/Mac nor identify a font by way of it.

I just tested more with XeTeX; I replaced ID4/Mac above with “familyname StyleName” (font selection is case sensitive) and managed to select the font with either “Familyname” or “familyname StyleName”; both ID1/Mac and ID4/Mac can be used to identify the above font in XeTeX.

Hi Michel — slightly off topic, but since you are here and you replied to me on the Font Forge list about the curled quotes in XeLaTeX, the situation is this: the main call of a font using fontspec produces, for the font that is called, the correct curled quotes. But suppose later in the preamble you call another face from a different font with ‘\newfontface\blah…’ that font will not have the quote marks curled correctly because this call does not apply the default language to this font. So lacking knowledge of the correct language it will leave the quotes as straight apostrophe quotes. Obviously there should be a way to tell the second font that is called what language it should be in: or if there is such a way, I haven’t found it.

Even more off-topic, but I am curious:

You write — I just tested more with XeTeX; ... and managed to select the font with either “Familyname” or “familyname StyleName”; both ID1/Mac and ID4/Mac can be used to identify the above font in XeTeX.

Since I never used TeX — how exactly did you select fonts in TeX syntax? I am asking because ID1 alone may not be sufficient to identify a font since all styles of the same family share the same string in name ID1.

Would you mind extending your test a bit? You could add “Win” to all Windows platform names, and “Mac” to all Mac platform names. (It’s most reliable to do this adjustment in TTX: names with platform ID3 are Windows, and with platform ID1 are Mac. But do not change ID6.) And then again try selecting fonts in XeTeX with name ID1 or ID4, once with the Mac platform strings, once with the Windows platform strings.
I suspect that TeX, like most other applications, actually uses the Windows platform name records rather than Mac platform name records, so that the latter is not relevant at least for TeX.

[...] the main call of a font using fontspec produces, for the font that is called, the correct curled quotes.

How do you do that? What feature are you using?

Question. Prove that if the n last posts are yours and if nobody posts in the meantime, you can always edit those n posts and restore them to their original order :)

I suspect that TeX, like most other applications, actually uses the Windows platform name records rather than Mac platform name records, so that the latter is not relevant at least for TeX.

Here are the values I gave to various IDs and yes means that XeTeX selected the font;

   ID1/Mac   Familyname               yes
   ID4/Mac   familyname StyleName     yes
   ID6       MinionPro-Capt           yes
   ID1/Win   WinMenuName              no
   ID4/Win   MinionPro-Capt Win       yes


For instance the input file containing the three lines

\font\test="MinionPro-Capt Win" at 48pt
\test Minion Pro Caption
\bye

gave me the expected pdf output, namely “Minion Pro Caption” in 48 points.

PS. This post has been edited (you can also see that from its comment number in the html file of this thread).

Word and textedit had their picture; to do justice, here is the one with TeXShop:

:)

Michel asked: How do you do that? What feature are you using?

\setromanfont[Mapping=tex-text,Numbers=OldStyle,Ligatures={Common,Rare}]{BemboBookMTPro-Regular}
\setsansfont[Mapping=tex-text]{Futura}
\setmonofont{Monaco}

…And of course for all these the quote marks are curled correctly.

Mapping=tex-text

Thanks! That’s great for using the same input files for pdflatex and xelatex. Maybe I was lucky but it seems to me that it is enough to add Mapping=tex-text in the call to \newfontface. And now, what requires more work is getting straight quotes. Do you know of any better solution than what follows?

PS. Edited.

Yup, Mapping=tex-text does the trick! :) *Slaps forehead!*

I think that probably is the best way to get the straight quotes.

And goddammit, I didn’t say “Michel AKSED”. :)

And goddammit, I didn’t say “Michel AKSED”.

I removed “Check spelling as you type” in TeXShop to get screen shots without red underlines and was not careful enough; with those nice spell checkers, we get less attentive, I am afraid. I swap letters much too often. I did it with AFDKO too, above.     s/aksed/asked/ and s/ADFKO/AFDKO/

There is now a generation of Australians who mispronounce ’asked’ as ’aksed’. I thought you were making a wry comment upon them. :) But no worries, as we are also wont to say!

According to Futurama, everyone pronounces it “aks” by the year 3000.

According to Futurama, everyone pronounces it “aks” by the year 3000.

Something to look forward to then. :(

I just heard a man give a talk where he pronounced the word ’facts’ as ’fak’. Noticeable, but not the end of the world. Now if it were in Australia, it might literally have been at the ’end of the world,’ so I might have gotten worried.

By the way, why would anyone use TeX? I understand that it is great for setting math text, but why for general usage? What benefits are there over either a commercial product or non-commercial such as Scribus?

I use TeX in my professional work as a Mathematician, but in my amateur roll as editor of a church newsletter, I’ve used every free-software program on Linux that’s remotely suitable, including Scribus, OpenOffice, KWord, etc. I’ve put at least a couple months into using so that I can get the hang of how it works, and I’ve finally settled on TeX for this newsletter as well. Since I’ve already gotten over the somewhat steep learning curve of TeX, I find the increased efficiency of doing all formatting from a text file where I never have to remove my fingers from the keyboard to be the best part about using TeX. I like being able to type \begin{twocolarticle}... \end{twocolarticle} (after defining this environment) and knowing that the text will be formatted correctly. Of course, styles in more visually-oriented programs do the same kind of thing, but I’ve realized that what really makes the difference for me is not having to use the mouse.

In projects such as a multiple-chapter church history, I like the relatively easy generation of structure: chapters, table of contents, footnotes, index, cross-references etc. I’m sure there are professional products that do this kind of thing, but the tools available in OpenOffice or Microsoft Word are just so much harder to use (in my opinion).

Finally, when I did a church photo directory, I liked the fact that it was easy to use a database to automatically generate the TeX code for each entry.

Hope this helps.

st

I also use LaTeX for typesetting mathematics. One great asset of LaTeX is the packages that people have contributed to it. If you need to write a multilingual text and hyphenate each language correctly, the babel package gives you simple macros you can call to do it. LaTeX is great when it (edit: i.e. the text to typeset) gets technical but sometimes even then I wonder: for instance, there is a package called ledmac (see also http://www.djdekker.net/ledmac/) for typesetting critical editions. Here is grab from an output using the sample file villon.tex; I used XeLaTeX with Adobe Garamond Pro and a custom single substitution to get Q.alt instead of Q (and threequartersemdash instead of emdash).

When I look at some of the .tex source examples, I wonder if there is no other tool able to handle footnotes and margin notes that would make life easier, but I confess I know nothing of critical editions.

For some mysterious reason, last night I browsed other sites and saw that my quotes for feet and inches above were not right; they should have been U+2032 and U+2033. I can’t get them with XeLaTeX. So, back to the old days with LaTeX, taking advantage of the math font (support for math fonts still seems incomplete in XeTeX).