On Tue, Jul 15, 2008 at 5:13 PM, <brion(a)svn.wikimedia.org> wrote:
Log Message:
-----------
Revert r37663 for now:
"* (bug 13815) In the comment for page moves, use the colon-separator message
instead of a hardcoded colon."
"* So that this works properly, don't escape HTML entities in edit summaries. I
don't see any good reason for them to be escaped there. Of course, this may result in
old edit summaries displaying slightly differently if for some reason they included an
entity, but in that case there's at least a 50% chance that they intended it to not be
escaped in the first place."
This breaks the ability to easily discuss entities in summaries such as "add
".
. . . then does the fact that we interpret entities in article text
break the ability to easily discuss them there? As I said, as likely
as not, people who are talking about entities were already entering
&entityname; anyway.
I did a bit of quick research, in the form of SELECT * FROM
recentchanges WHERE rc_comment LIKE '%&%' LIMIT 100 on enwiki.
The results can be classified as follows:
* Three were, in fact, discussing entities in their summary, as you
say, and were correctly not escaped.[1-3]
* The other 97 were *incorrectly* escaped. These included a number of
cases where the failure to escape would have caused outright errors,
not just ugly appearance.
** In a bunch of cases (like [4]), the & occurred inside an
external URL, and the URL would fail if copied, since it was
double-escaped.
** In a few other cases (all deleted revisions adding AFD notices, so
I can't link to them) the & was in a wikilink. Although I can't
see the results, my testing[5] indicates that in edit summaries this
causes the link to fail, again incorrect.
** And the remaining few dozen were things like /* Wrexham &
Shropshire Route */ (e.g., [6]), which I'm not sure are affected by my
change, but if they are it's positively: they look ugly, and they
don't work as links either.
So in short, 97% of the time it's more correct -- or if you disqualify
the /* section name */ ones, if this doesn't affect them, it's still
more correct at least 80% or so of the time. *And* on top of this it
allows localization of the colon. Can I recommit this?
[1]
http://en.wikipedia.org/w/index.php?title=User_talk:Geestring&diff=2232…
[2]
http://en.wikipedia.org/w/index.php?title=Spam_(electronic)&diff=220390…
[3]
http://en.wikipedia.org/w/index.php?title=Template_talk:Frac&diff=21982…
[4]
http://en.wikipedia.org/w/index.php?title=User_talk:Atomadams13&action=…
[5]
http://en.wikipedia.org/w/index.php?title=User:Simetrical/Sandbox&diff=…
[6]
http://en.wikipedia.org/w/index.php?title=Wikipedia:Articles_for_deletion/W…