It's got my vote too.
Both the RB and the ModSqueak tests require it to run properly.

David Farber <dfarber at numenor.com> wrote:
> 
> yes! please, please, please include sUnit3.0 in the new release!!
> 
> david
> 
> At 08:10 AM 10/2/01 -0400, you wrote:
> >Dan and the Harvesters,  (it sounds like oldies but goodies!  ;-)
> >
> >There's no doubt about it, it's great seeing all of those community changes 
> >go though.  It looks like there were some nice fixes in there.  In light of 
> >the beta, following by such slicing and dicing, please consider putting the 
> >following changesets into the 3.1alpha stream.
> >
> >[3.0 SUnit]
> >I have attached the SUnit 3.0 changesets, with changes to remove the 
> >existing, deprecated Classes and methods from the 2.8 version in the 
> >current image.  All SUnit tests pass.
> >
> >[NuBlu Events code plus image repairs]
> >The second set of changesets are the new event code, from the NuBlu 
> >book.  I changed the top level actionMap dictionary to be a 
> >WeakKeyIdentityDictionary, like the EventsFields class var used to hold.  I 
> >also went through the image and changed all of the references to the old 
> >events code to this set.  There wasn't a lot of use of this mechanism, but 
> >some in Morphic.    Please consider putting these into the pre 3.2 stream, 
> >also.    This framework is used by ModSqueak, which is still being looked 
> >at, especially for the Declarative model it introduces.  This changeset 
> >would help ease that effort, and perhaps help us combine the two Module 
> >mechanisms.
> >
> >cheers,
> >Rob
> >
> >PS.  I recently switched from Outlook to Eudora, so I fervently hope that 
> >there isn't any HTML nonsense going on in this email.  I told eudora three 
> >times!   Please let me know if I;ve mangled it once again.
> 
> --
> David Farber
> dfarber at numenor.com