It IS beta, so I guess bugs are to be expected. However, it's a nice improvement and we'd like to use it now. Here is our scenario:

We're using test 2.0 for a MC-test. 3 attempts. Best one counts (regardless of order). However, the progress report 2.0 reports "not passed" for students that have failed on a second (or third) attempt after passing their first (or second) attempt. Thus, it does not take into account that the "best attempt counts".

Hopefully you can fix this ASAP as we'd like to use the excel export provided with the PR 2.0.

In Test 2.0. If anything other than 'no assessment' is chosen, the students are prompted 'Grades will be given' before starting the test. This is most often wrong / misleading as we mostly use 'Accepted/Non accepted' as the preferred form of assessment. The students then wrongly expect to be graded, which every semester cause a lot of confusion / misunderstanding.

We know that the different forms of assessment are editable and thus it is difficult to design individual feedback based on the assessment chosen. However, we urge you to replace the current statement with something else. For instance: 'Assessment will be given'

In it's' mail system, HTML is used to format mails, defining images, links, fonts and more, however HTML also enpowers abusers to do things that the system shouldn't allow, such as using img against log_out.aspx or index.aspx (does the same thing when GETed without arguments), or even /, I threrefore suggest that you create a special language, in which the number of structures/functions/possibilities are reduced, making it more difficult to abuse.

When sending a message two posts are done, one to check recipients (javascript:ValidateRecipients()), and one to send the message itself (javascript:SendMessage()), this could have been accomplished by a single post in javascript:sendMessage, since most of the time recipients are correct. Changing this increases the speed in which the code is executed on the client side, (less waiting for server, less code executed), minimizes pressure on the servers (fewer packages sent), and also fixes a couple of security holesv(actually it makes it more difficult to abuse, since you would end up with several lines of javascript) since sending messages is reduced to a single function.

In your mail system, you seem to assume that all recipients of an mail exists, however recipients are chosen by a client, and checked in JS at the sender, but what about if the send button handler is changed from javascript:onSend() to javascript:SendMessage, that way recipients are never checked, and might include illegal recipients, which (since you seem to be missing an exception in the inbox), generates an unhandled exception, temporally paralizing the mailbox of that user.

Sometimes you get an "Runtime Error" for no reason. It would be nice with a better error message describing the problem.

Here is a page that provokes the message: http://www.itslearning.com/KURS/14901/holybible/

You will not be able to answer e-mails received by POP3/IMAP if the "Date" header is invalid. Please fix this.

The Java upload does not work with files over ~30 MB. Please fix this.