RE: [Moderator Action] litmus test, dangers of

Hi Frank,

> 1) Is this reluctance justified?  That is, would running the litmus
> test on a production server entail a risk of any significance?

There have been cases where running Litmus exposed an error that caused a
server to crash, and stop running. If Litmus is run against a collection
which does not contain production data, then there should be minimal danger
of data loss (the only risk would be Litmus exposing a bug that corrupted
the database).

So, I'd recommend running the test early some morning, when system use is
minimal.

> 2) If the reluctance is not justified, how might we allay their
> fears?  Is there anything from w3c (statements, etc.) that could be
> used in this endeavor?

The other approach would be to install WebCT on a test machine, and then run
Litmus against it.

> Using WebDAV in WebCT Vista is a little different from other systems.
> Courseware developers elect to use WebDAV at some point in a course
> under development by clicking on a button labeled "WebDAV."  The
> WebCT Vista system responds with a long and complex URI that the
> courseware developer copies for pasting into their WebDAV-compliant
> application or OS interface for WebDAV.
>
> The length of these URIs can be problematic for IE in Win2K and XP.
> Sometimes, it will generate a "too long" error message from "My
> Network Places" as well. MacOS X doesn't complain about the length
> but has other issues I won't go into here.  Here's a typical example
> of one of those WebCT Vista WebDAV URIs:

Litmus is unlikely to uncover a problem such as this, since this is more of
an interoperability problem, than a compliance problem. Litmus tests for
compliance.

Hope this helps.

- Jim

Received on Thursday, 18 September 2003 14:16:26 UTC