bug-texinfo
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: rethinking @def*


From: pertusus
Subject: Re: rethinking @def*
Date: Tue, 2 Aug 2022 15:18:22 +0200

On Tue, Aug 02, 2022 at 01:20:50PM +0100, Gavin Smith wrote:
> On Mon, Aug 01, 2022 at 12:56:23AM +0200, pertusus@free.fr wrote:
> > > What was the benefit of changing <em class="emph"> to
> > > <span class="r"><i class="slanted">?  Isn't the former much simpler?
> > 
> > It is not the same, <span class="r"> isolates from the surrounding
> > fonts, using <span class="r"><i class="slanted"> amounts to really doing
> > the same as in LaTeX (and, I believe, TeX).
> 
> I thought we still wanted a fairly simple HTML output without
> specifying exact details of formatting.  Simpler output would make
> it easier for users to customize the manual with CSS, as well as any
> other processing they wanted to do.
> 
> When would a @def* block be inheriting font styles that we would need to
> cancel?

There is @def* in @example and similar, though this is not a
very important use.  The idea, here, was to be more in line with
TeX/LaTeX formatting.  I like in particular your idea that we could say
that @def* (not @deftype) are @deftype {} {name} @r{@slanted{arg}}

> The 'emph' class on <em class="emph"> wasn't necessary as
> it wasn't coming from the @emph command (in the document,
> at least).  But I think a special class is useful.  I saw
> you added the "def-meta-var-arguments" class in a recent
> commit.

Indeed, in the latest code, there are specific classes (except for the
@slanted inside formatting).

> It seems like there are three choices for the tag, <em>, <i> and
> <var>.  Previously, we used <em>.  I don't have a strong opinion
> which one is best.

I think that <em> was actually a poor choice, not the right semantics,
no clear formatting, different from TeX/LaTeX.

> In the TeX output, we use slanted roman for the definition line
> and @var, not italics.

There is no slanted in HTML, probably because all the browsers render
italics as slanted, as you say below, so I simply used what we convert
@slanted to in HTML in general, which is <i class="slanted">.  An I
think that we dod not bother to define a CSS to ensure that it appears
slanted and not italics, as it is the default with most if not all
browsers.

> When I checked in a web browser (Chromium), <i> and others appeared to
> be slanted roman, not italics.  I couldn't get italics whatever I did
> (with a closed "a" and a tail on the "f").  (It probably depends on
> my default system fonts and likely other users would get true italics.)
> 
> I think it would be fine to use <em> or <var> and not worry if the browser
> does use a true italic for these in HTML output.

As I said above, we want slanted for more similar formatting with
LaTeX/TeX, so I used the same output as @slanted.  I do not think that
<em> is appropriate.  <var> would be right, in like with the
metasyntactic variable semantic of the @def* not @deftype* arguments.
I did not go that way to be more in line with what I thought we would
say, to make clear that the @var formatting is not used, only slanted,
mostly to avoid upper case in Info default.  But <var> in HTML would be
ok to me too.

Actually, as long as the formatting remains the same, to be as similar
as possible to printed output, I do not care much about the details.  I
tried to stick to a similar formatting as for printed output, but I do
not think that the deails of how it is implemented matters much.

-- 
Pat



reply via email to

[Prev in Thread] Current Thread [Next in Thread]