On Sat, Nov 30, 2002 at 12:23:08PM -0800, Brion VIBBER wrote:
Tomasz Wegrzanowski wrote:
Please restore old meaning of <pre> which
suppressed interpretation
of wiki markup ! (for Polish Wikipedia at least, I don't care much
whether others will have broken markup)
I don't remember it ever working differently; I was not aware that the
behavior had been changed. Changing it again now would of course
probably break other code...
It always worked differently.
If you don't want to mess with code too much, just insert something like:
if ($language eq "pl")
{
$text =~ s/<pre>/<pre><nowiki>/g;
$text =~ s/<\/pre>/<\/nowiki><\/pre>/g;
}
It broke
virtually all code examples on Polish Wikipedia !!!
If you want that, use:
<pre><nowiki>
...
</nowiki></pre>
Some 90% of <pre> would need such change.
First, it would mean lot of wasted effort.
Second, <pre> on wiki has always been meant for listing code,
so it makes no sense at all to use wiki markup inside <pre>.
And third, <nowiki> is extremely low level markup. Introducing
more low level markup is not right direction.
In future, some higher level mechanism like <code
name="hello.cpp"></code>
may be implemented, to get rid of <pre> and <nowiki> and give some extras
like downloading examples listed in arcticle, searching for code etc.
But as for now let <pre> keep its old meaning.
And don't
do any more changes to markup in future without telling
people about that.
I'll just point out we had a test server running for *months* with the
new software and a copy of the Polish wiki precisely to check for things
that might need to be fixed or changed.
Human testing is always unreliable - people take most efford to check things
they suspect might be broken, like translation. They don't check things they
don't expect to break.