http://www.encrypted.net/~jh/img/roach.gif Back to MoinMoinTodo | Things that are MoinMoinNotBugs | MoinMoinTests contains tests for many features

Tables broken by trailing spaces

Tables don't work right if there are trailing spaces.

Yes, by design, just like headings don't accept trailing spaces. In the case of headings, it is important because "= x =" is somewhat ambiguous. For tables, the restriction could be removed, though.

This row

is ok

|| this row || isn't ||

OK

again

Cookies

And as a fun sidenote, the UserPreferences cookie doesn't seem to work. The cookie is in my ~/.netscape/cookies file still. My EfnetCeeWiki and XmlWiki cookies are duplicated!? It is kind of like the bug you found when I, ChristianSunesson, tried this feature with netscape just after you deployed it. Hehe, and my changes are not logged to the RecentChanges database, I'll hope you find these changes this year. :) BUG BUG BUG!

Well, Netscape suxx. I send the cookies with the CGI'S path, w/o a hostname, which makes it unique enough. Apparently not for netscape. I'll look into adding the domain, too.

Hmmm...I use NetScape, MsIe, MoZilla and Galeon. I haven't had a problem but some other's using MoinMoin 0.8 have intermittently lost cookies. Any ideas? -- BruceDurling

Misc

Unicode issues

With 0.3, TitleIndex is broken if first letter of Japanese WikiName is multibyte character. This patch works well for me but need to be fixed for other charsets.

*** moin.cgi.sav        Sat Oct 28 13:26:26 2000
--- moin.cgi    Sat Oct 28 13:39:04 2000
***************
*** 623,628 ****
--- 623,630 ----
          letter = name[0]
          if isUnicodeName(name):
              letter = "~"
+         if letter not in (string.letters + string.digits):
+             letter = "~"
          if letter not in index_letters:
              index_letters.append(letter)
          if letter <> current_letter:

AtsuoIshimoto Differently broken. :) I think we can live with the current situation, the worst edges are removed (before, chopping the first byte out of an unicode string lead to broken HTML markup!). It will stay that way until I buy the Unicode 3.0 book.

paragraph bug redux

They render identically with IE5. Provide some info on your browser & OS. And the HTML is identical:

Some text. <p>
<b>window</b> <ul>
 The main, rectangular background, control and data area of an application. <p></ul>
<b>dialog boxes</b> <ul>
 A temporary, pop-up window created by the application, where the user can 
enter information and provide commands. <p></ul>

Please note that they aren't actually identical: the P that precedes dialog boxes is interrupted by the closing /UL tag, whereas the P preceding windw is not followed by a closing /UL.

Bug in ASCII encoding

The page PrettyPrintXslt seems to have a nasty bug in it ...

That is a known bug in 4XSLT.