So both this effort and the exobox were vaporized. I see a trend...

Tim, in your opinion, would this be worth trying one more time?

If so, would it have to be started from scratch or is anything out there
that could be reused/revived?

Bob

Tim Rowledge wrote:
> Brent Vukmer <brent.vukmer at gmail.com> wrote:
> 
> 
>>>We made a valiant attempt to tackle this at Interval for the WebPad by
> 
> making
> 
>>>an OS that let Squeak handle realtime events and processor scheduling all
> 
> the
> 
>>>way down. Killed by billg.
>>>
>>
>>Dammit.  How far did you guys get with this?
> 
> Quite a long way. We had a working piece of handheld hardware that we showed at
> OOPSLA 99(?) which scheduled all tasks via Squeak. This included all the
> machine code written tasks for networky stuff etc. It also had a Slang to
> object-code translator that was on the way to being able to generate prims on
> the fly (for device drivers etc). 
> 
> 
>>How did Squeak do at
>>handling realtime events & processor schedules?
> 
> Quite well. It certainly wasn't finished in any production sense but it ran a
> really rather big test suite without problems. The realtime requirements were
> 'moderately hard' rather than 'very hard' but it could manage many thousands of
> interrupts per second IIRC.