Обсуждение: Re: [HACKERS] switching documentation build to XSLT

Peter Eisentraut <peter.eisentraut@2ndquadrant.com> writes:
> If you find any changes in the output that bother you, let pgsql-docs know.

So, while I'm complaining, have a look at the Note near the bottom
of

https://www.postgresql.org/docs/devel/static/datatype-numeric.html

and compare to its appearance in

https://www.postgresql.org/docs/9.6/static/datatype-numeric.html

Now, to my taste, our past markup of note/tip/etc is already
overemphasizing them.  They're inset with narrower margins and
put on a different-color background, which makes it hard to avoid
the impression that they're the most important thing on the page
and you should barely bother to read anything else.  In almost no
case is that actually warranted; certainly not for stuff with
markup less than <warning>.  (Personally I'd argue that <note> means
it's *less* important than the surrounding text, not more so.)
But putting them in bold type with a separated heading is just over
the top.  What are we going to do for text that's actually critical
... make it bright red and blinking?

Just to add confusion, the note on the previous page is set
in normal type:

https://www.postgresql.org/docs/devel/static/datatype.html

which I don't understand since the sgml source is about the same.
But this one is still different from the way it was rendered by
the old toolchain.

            regards, tom lane

Hello,
It seems that there is some post-processing after `make html` that adds
class = "c1" or class = "c2" to html.
(It makes <div class="note c2"> or <div class="note c1">)
And docs.css on the website contains the rule:
.emphasis,
.c2 {
   font-weight: bold;
}
So I think, the issue is related to publishing on the website.

Best regards,
Alexander

23.11.2016 05:39, Tom Lane writes:
> Just to add confusion, the note on the previous page is set
> in normal type:
>
> https://www.postgresql.org/docs/devel/static/datatype.html
>
> which I don't understand since the sgml source is about the same.
> But this one is still different from the way it was rendered by
> the old toolchain.
>
>             regards, tom lane
>
>