relsize: What to do with "Failed to get list of font sizes."
Erik Quaeghebeur
I'm using relsize 3.1 with memoir/fourier/microtype and get the following
error:
Failed to get list of font sizes.
What can I do about this?
(Or, what further info do you need to diagnose the problem?)
Erik
Donald Arseneau
That means that all of the size-changing commands (like \large) were
of
unexpected format, and relsize could not determine what font size was
used for each. (As a result it can't perform the attempt to stay
close to
a ratio of 1.2 per "step".) But memoir has regular definitions for
the
size-changing commands, so you will have to track down what is the
actual conflicting package. (You could insert "\show\large" between
loading each package.)
Donald Arseneau as...@triumf.ca
Erik Quaeghebeur
>>
>> I'm using relsize 3.1 with memoir/fourier/microtype and get the
>> following error:
>>
>> Failed to get list of font sizes.
On Tue, 30 Sep 2008, Donald Arseneau wrote:
>
> That means that all of the size-changing commands (like \large) were of
> unexpected format, and relsize could not determine what font size was
> used for each. (As a result it can't perform the attempt to stay close
> to a ratio of 1.2 per "step".)
>> What can I do about this?
>> (Or, what further info do you need to diagnose the problem?)
> But memoir has regular definitions for the size-changing commands, so
> you will have to track down what is the actual conflicting package.
> (You could insert "\show\large" between loading each package.)
When I use \show\large, I always get:
> \large=macro:
->\@setfontsize \large \@xipt {13.6}.
which is memoir's 10pt definition... I have no idea how to proceed.
Erik
Erik Quaeghebeur
>>
>> I'm using relsize 3.1 with memoir/fourier/microtype and get the following
>> error:
>>
>> Failed to get list of font sizes.
>>
>> What can I do about this?
>> (Or, what further info do you need to diagnose the problem?)
On Tue, 30 Sep 2008, Donald Arseneau wrote:
>
> That means that all of the size-changing commands (like \large) were of
> unexpected format, and relsize could not determine what font size was
> used for each. (As a result it can't perform the attempt to stay close
> to a ratio of 1.2 per "step".) But memoir has regular definitions for
> the size-changing commands, so you will have to track down what is the
> actual conflicting package. (You could insert "\show\large" between
> loading each package.)
I found out that this message was due to the fact that I \RequirePackage'd
it inside some command for making a \sum-size/type cartesian product � I
took from some comp.text.tex posting.
Now I get: (multiple times)
Package relsize Warning: Size requested is too small.
(relsize) Using 6.0pt instead on input line 1413.
this follows closely after:
LaTeX Font Info: Font shape `FMX/futm/m/n' will be
(Font) scaled to size 15.89755pt on input line 1413.
LaTeX Font Info: Font shape `FMX/futm/m/n' will be
(Font) scaled to size 11.03998pt on input line 1413.
LaTeX Font Info: Font shape `U/futm/m/n' will be
(Font) scaled to size 15.89755pt on input line 1413.
LaTeX Font Info: Font shape `U/futm/m/n' will be
(Font) scaled to size 11.03998pt on input line 1413.
LaTeX Font Info: Font shape `FMX/futm/m/n' will be
(Font) scaled to size 13.24796pt on input line 1413.
LaTeX Font Info: Font shape `U/futm/m/n' will be
(Font) scaled to size 13.24796pt on input line 1413.
this seems to point to me using the commands that use relsize in a
sidenote in footnotesize. The resulting size of the relsized symbol is
somewhat big and could be shrunk without impeding legibility.
The (big cartesian product) commands that (indirectly) use relsize are
\makeatletter
\@ifundefined{@currsizeindex}
{\let\dolarger\relsize}
{\def\dolarger#1{\larger[#1]}}
\newcommand*\@b@gtimes[1]{%
\vcenter{\hbox{\dolarger{#1}$\m@th\mkern-2mu\times\mkern-2mu$}}}
\newcommand*\@bigtimes{%
\mathchoice{\@b@gtimes{4}} % display style
{\@b@gtimes{3}} % text style
{\@b@gtimes{0}} % script style
{\@b@gtimes{-2}} % script script style
}
\newcommand*{\Cart}{\hyperref{}{symbol}{Cart}{\mathop{\mathpalette\@bigtimes\relax}\displaylimits}}
\newcommand*{\IProd}{\hyperref{}{symbol}{IProd}{\mathop{\mathpalette\@bigtimes\relax}\displaylimits}}
\makeatother
\newcommand*{\tCart}{{\textstyle\Cart}}
\newcommand*{\tIProd}{{\textstyle\IProd}}
(this is not ideal, the whitespace around the � is not that good, I'll
probably need to fiddle with the \mkerns...)
Erik
Donald Arseneau
This posting sounds like you are making progress, but it leaves me
baffled.
> I found out that this message was due to the fact that I \RequirePackage'd
> it inside some command for making a \sum-size/type cartesian product � I
Characters corrupted in posting? \RequirePackage in a command?! or
in
some other package?
> Package relsize Warning: Size requested is too small.
> (relsize) Using 6.0pt instead on input line 1413.
You ask for something smaller than the smallest declared font size, so
the
warning is appropriate.
> LaTeX Font Info: Font shape `FMX/futm/m/n' will be
> (Font) scaled to size 15.89755pt on input line 1413.
So what's line 1413?
> \@ifundefined{@currsizeindex}
> {\let\dolarger\relsize}
> {\def\dolarger#1{\larger[#1]}}
What is this \@currsizeindex command, and what is this choice between
\relsize and \larger?
I don't think I can diagnose what's happening without a short
(minimal)
demonstration document.
Donald Arseneau
as...@triumf.ca
Enrico Gregorio
> On Oct 1, 2:25 pm, Erik Quaeghebeur <equae...@nospammail.net> wrote:
>
> This posting sounds like you are making progress, but it leaves me
> baffled.
>
> > I found out that this message was due to the fact that I \RequirePackage'd
> > it inside some command for making a \sum-size/type cartesian product ÔøΩ8 I
>
> Characters corrupted in posting? \RequirePackage in a command?! or
> in
> some other package?
>
> > Package relsize Warning: Size requested is too small.
> > (relsize) Using 6.0pt instead on input line 1413.
>
> You ask for something smaller than the smallest declared font size, so
> the
> warning is appropriate.
>
> > LaTeX Font Info: Font shape `FMX/futm/m/n' will be
> > (Font) scaled to size 15.89755pt on input line 1413.
>
> So what's line 1413?
>
>
>
> > \@ifundefined{@currsizeindex}
> > {\let\dolarger\relsize}
> > {\def\dolarger#1{\larger[#1]}}
>
> What is this \@currsizeindex command, and what is this choice between
> \relsize and \larger?
The origin of this code is related to the fact that both AMS classes
and relsize provide the command \@currsizeindex, so the definition has
to cope with this. The code should be
\@ifundefined{@currsizeindex}
{\RequirePackage{relsize}\let\dolarger\relsize}
{\def\dolarger#1{\larger[#1]}}
in the preamble or in a package. If the user is not using an AMS class,
relsize will be loaded. In any event, the command \dolarger will do
the right thing.
I proposed this method for a \bigtimes command in a discussion with
Lars Madsen. In the end he decided to put in his dlfltxbcodetips
package another definition.
> I don't think I can diagnose what's happening without a short
> (minimal) demonstration document.
Of course.
Ciao
Enrico
Donald Arseneau
> Donald Arseneau <a...@triumf.ca> wrote:
> > What is this \@currsizeindex command, and what is this choice between
> > \relsize and \larger?
>
> The origin of this code is related to the fact that both AMS classes
> and relsize provide the command \@currsizeindex,
Not relsize.
I see that the AMS classes use a different list of font-size commands
(\SMALL instead of
\scriptsize) and that the definitions of such would produce the
problem noted
in the first message: "Failed to get list of font sizes." But they
also provide their
own \larger and \smaller commands, so no need for relsize.
> \@ifundefined{@currsizeindex}
> {\RequirePackage{relsize}\let\dolarger\relsize}
> {\def\dolarger#1{\larger[#1]}}
But relsize allows \larger[#1] in the same way as ams-classes. So
better:
\def\dolarger#1{\larger[#1]}
@ifundefined{larger}{\RequirePackage{relsize}}{}
I wonder if Erik is using ams classes rather than memoir.
Donald Arseneau as...@triumf.ca
Enrico Gregorio
> On Oct 3, 4:02 am, Enrico Gregorio <grego...@math.unipd.it> wrote:
> > Donald Arseneau <a...@triumf.ca> wrote:
>
> > > What is this \@currsizeindex command, and what is this choice between
> > > \relsize and \larger?
> >
> > The origin of this code is related to the fact that both AMS classes
> > and relsize provide the command \@currsizeindex,
>
> Not relsize.
I didn't remember the discussion very well, it seems. What I wanted
was to check if \@currsizeindex was defined.
> I see that the AMS classes use a different list of font-size commands
> (\SMALL instead of
> \scriptsize) and that the definitions of such would produce the
> problem noted
> in the first message: "Failed to get list of font sizes." But they
> also provide their
> own \larger and \smaller commands, so no need for relsize.
>
> > \@ifundefined{@currsizeindex}
> > {\RequirePackage{relsize}\let\dolarger\relsize}
> > {\def\dolarger#1{\larger[#1]}}
>
> But relsize allows \larger[#1] in the same way as ams-classes. So
> better:
>
> \def\dolarger#1{\larger[#1]}
> @ifundefined{larger}{\RequirePackage{relsize}}{}
Of course, probably I didn't realize the presence of that functionality.
> I wonder if Erik is using ams classes rather than memoir.
We can't know, because he's not telling us.
Ciao
Enrico
Erik Quaeghebeur
>
> > I wonder if Erik is using ams classes rather than memoir.
On Mon, 6 Oct 2008, Enrico Gregorio wrote:
>
> We can't know, because he's not telling us.
Memoir; no ams class (as I said in the OP).
I do use amsmath/amssymb, but that is probably not relevant.
The reason I have stayed silent is:
* my attempts to produce a minimal example from the bottom-up failed,
* I am very busy wrapping up my doctorate[#], so I did not start a
reduce-stepwise minimal example generation procedure...
* a previous reply of mine to a previous message in the thread did not get
delivered because I (stupidly) left a "Control: cancel" header in all my
usenet postings...
I modified my \bigtimes to
\newcommand*{\scaledtimes}[1]{\vcenter{\hbox{\relsize{#1}$\mkern-2mu\times\mkern-2mu$}}}
\newcommand*{\bigtimes}{\mathop{
\mathchoice{\scaledtimes{4}}
{\scaledtimes{3}}
{\scaledtimes{0}}
{\scaledtimes{-2}}}\displaylimits
}
but still get "Size requested is too small"-warnings; although less. I
should test some more to be able to really diagnose it. However, the final
output is good, so currently I'm not going to spend time on that. I
probably will when preparing the published version.
BTW, thanks to both of you, Donald & Enrico, for looking at this issue.
Erik
# On the latex-side I'm now fighting with the indexes, toc, and some
spurious blank pages... (so if you see other vague, minimal-example-less
posts of mine, don't judge me too harshly).
Donald Arseneau
>\newcommand*{\scaledtimes}[1]{\vcenter{\hbox{\relsize{#1}$\mkern-2mu\times\mkern-2mu$}}}
> \newcommand*{\bigtimes}{\mathop{
> \mathchoice{\scaledtimes{4}}
> {\scaledtimes{3}}
> {\scaledtimes{0}}
> {\scaledtimes{-2}}}\displaylimits
> }
>
> but still get "Size requested is too small"-warnings;
OK, they were the trivial ones. It was the "Failed to get list of
font sizes"
that was the big problem and couldn't be diagnosed.
In a \mathchoice, all versions are formatted in advance, but at the
end
just one is chosen for typesetting. But you can get warnings from
all four. If the current font size is \scriptsize, you always get the
warning
because there are not two steps smaller. You will get similar
warnings
when you use this in large title fonts without 3 steps larger.
> output is good, so currently I'm not going to spend time on that.
It is just a warning, so it is fine to ignore it.
DA
Erik Quaeghebeur
>
>> \newcommand*{\scaledtimes}[1]{\vcenter{\hbox{\relsize{#1}$\mkern-2mu\times\mkern-2mu$}}}
>> \newcommand*{\bigtimes}{\mathop{
>> \mathchoice{\scaledtimes{4}}
>> {\scaledtimes{3}}
>> {\scaledtimes{0}}
>> {\scaledtimes{-2}}}\displaylimits
>> }
>>
>> but still get "Size requested is too small"-warnings;
On Mon, 6 Oct 2008, Donald Arseneau wrote:
>
> OK, they were the trivial ones. It was the "Failed to get list of font
> sizes" that was the big problem and couldn't be diagnosed.
>
> In a \mathchoice, all versions are formatted in advance, but at the end
> just one is chosen for typesetting. But you can get warnings from all
> four. If the current font size is \scriptsize, you always get the
> warning because there are not two steps smaller. You will get similar
> warnings when you use this in large title fonts without 3 steps larger.
Interesting to know (but it seems somewhat strange).
>> output is good, so currently I'm not going to spend time on that.
>
> It is just a warning, so it is fine to ignore it.
Ok, thanks!
Erik