I have noticed in the unstable repository that there are many places that URLs are constructed, and they seem to be gaining & entities across the board.
I would just like to say that is likely to lead you into a whole lot of trouble further down the line, which will include double-encoding and having to decode and re-encode strings all over the place. The format of a URL involves separating GET parameters with ampersands (&s), not & entities. Keep the XML-encoding at the level it belongs - close to where the URL is output in XML format, and not in the non-template code that puts these strings together.
I've seen this happen in other projects, and it stems from a misunderstanding of what a URL is, and where encoding needs to take place for a particular output format (which may *not* even be HTML).
Just a general thing anyway, on the table for you to think about.
0 • •