March 05, 2003


  • Syndirella now does RSS autodiscovery only if the address specified in the "Add RSS feed" dialog points to a HTML page. If you specify the address of a broken RSS feed, a non-RSS XML file or some totally unrelated file (for example, a .GIF image), you will now get an error message immediately.
  • Syndirella now animates its tray icon for 30 seconds after a feed update that brought new items has been completed. After that, the icon turns solid red, and remains red while you still have unreasd items. When there are no unread items, the icon is colored blue.
  • The primitive HTML parser used for RSS autodiscovery now also understands attribute values surrounded by single quotes, not only double quotes.
  • Fixed several errors reported by Syndirella users.

Download binaries, sources or installer.

Posted by yole at March 5, 2003 01:00 AM | TrackBack

Just for clarification, does it only autodiscover on HTML pages, or does it work with PHP, ASP, and/or CGI pages?

Posted by: Erik on March 5, 2003 01:17 AM

Erik: It autodiscovers on pages which have the opening <html> tag.

Posted by: Dmitry Jemerov on March 5, 2003 09:45 AM

Ruh-roh. A bit of trivia: An HTML document need not have an HTML tag.

In fact, neither the HEAD nor BODY elements seem to be required either.

You'd be better off looking at MIME types instead, or in addition.


Posted by: Jeremy Dunck on March 5, 2003 11:40 PM

Jeremy: Can you point me to an example of a weblog that doesn't have the html tag?

Many things are possible in theory, but since my time is limited, I consider it more practical to support only the things that are verified to exist.

Posted by: Dmitry Jemerov on March 6, 2003 12:01 AM

I'm sure I've seen some, but none comes to mind currently.

I agree that this need not be a priority, and in fact, would feel fine about not changing it until someone complains. :)

Posted by: Jeremy Dunck on March 6, 2003 10:52 PM

In an XHTML DTD document, using an XML parser, the HTML, HEAD, and BODY elements are all required, open and close.

In older HTML DTDs using SGML (or non-compliant implementations), the only thing that is not given start and/or end tag minimization and is required is the TITLE element.The HTML, HEAD, and BODY elements will be automatically assumed due to their use of SHORTREF.

In other words, *good* pages have doctype definitions at the top. *random* pages need to have the title tag. MIME type is how everything else detects it, more or less, and that fails sometimes too.


Posted by: Scott Swanson on March 24, 2003 04:10 AM
Post a comment