mediarss: less verbose handling of bad thumbnail tags

Description

http://www.woot.com/blog/rss.aspx has tags of the form:

<media:thumbnail xmlns="media">
http://www.woot.com/Images/Sale/Littermaid_LM920_Self-Cleaning_Litterbox838-thumbnail.jpg</media:thumbnail>

This doesn't seem to belegitimate according to the MediaRSS spec - the url
is supposed to be an attribute, not the contents of the element. But logging
an entire backtrace for this is annoying and unnecessary. This patch reduces
the logging to a single line.

(I turned the enclosing catch (Exception e) to a catch (RuntimeException e)
in case there were other problems that it was intended to catch. In my
own code, I'd probably have removed it entirely.)

Environment

None

Status

Assignee

ROME Jira Lead

Reporter

otaylor

Labels

None

Participants

None

Affects versions

current

Priority

Major