I very frequently have my page titles rejected by Confluence because it contains an illegal character, typically a colon or a slash. This drives me and my team batty! There are old issues regarding this problem that have been closed (CONF-253 and CONF-4238) and I couldn't find a way to reopen them, so I'm making this new one.

This is especially bad for blogs. I use Confluence to write my internal blog, and I'm constantly wanting to use a title like this one (from Jonathan Nolen's blog):

Confluence Massive: Clustered Wiki
http://blogs.atlassian.com/news/2007/01/confluence_mass.html

I'm not sure what restrictions (if any) blog titles have, but it would be ideal if Confluence wasn't more restrictive. For example, we were considering migrating a blog into Confluence, but a number of the titles would not have translated well.

Having said this, page titles still cause a problem. Some examples of page titles I couldn't create if I wanted to:

AC/DC
2001: A Space Odyssey
Meeting Notes: Team Meeting 10/20/2007

I'm sure I don't understand all the issues, but here are some thoughts:

There already is logic where certain pages use an id URL when invalid characters are included, so that means that there should never be restrictions on characters that are restricted in URL syntax. Primarily this means that '/' should be straightforward to support in titles. Of course, I may be missing something!

I imagine the real issue is that the Wiki syntax makes special use of these characters. I wonder if titles could be escaped in those sitations. This wouldn't be a problem for people using the link button, but it would make life more annoying for someone just typing in a name. Since slash is not part of the Wiki syntax, that shouldn't be an issue. However, I see that colon is definitely more problematic. Depending upon how strict the parsing is, you could special case links to interpret the colon in both senses.

e.g. [Confluence Massive:Clustered Wiki] could interpret as a link first, but if such a page does not exist then it would interpret it as a space directive. Of course, there are many macros that interpret links that would have to be fixed (including my Metadata Plugin) so maybe escaping is a better solution. For blogs it is less of an issue, as it is almost impossible for a normal user to link to one!

Supporting [ and ] in titles is less important for me, and I could see would have similar parser problems. However, if a general purpose escaping scheme was implemented, then I would suggest that titles should be completely unrestricted. A wiki is meant to be issue to use, and it is very irritating when you can't create the title you need.