When you generate a reference for a web page, our HTTP client needs to go out and fetch that web page so that we can gather required information such as the title of the page and the name of the authors. We found a minor bug in this HTTP client which resulted in it taking an unduly long time to fetch a small percentage of web pages. A smaller number of pages couldn’t be retrieved at all. In the first instance the output wasn’t effected (it just took upto a minute to produce the output). Unfortunately in the latter case, the generated reference did not contain the title of the webpage. This bug has now been fixed and all cached records have been purged.