NetSurf soon runs out of memory. This is because the memory claimed for the parsed HTML and box structures is not released when another page is requested. To get around this drag the Next bar in the Task Manager up to something ridiculous before running NetSurf. When NetSurf runs out of memory it will quit without warning.
Following a relative link does not work if the current URL does not end with a leafname or a slash. Eg: http://www.riscos.com would cause problems whereas http://www.riscos.com/ would not.
Multiple fetches (in different windows) don't seem to work yet.
This version includes new table code that should allow NetSurf to work with pages containing malformed tables.