Why LispOS?

Mike McDonald mikemac@teleport.com
Wed, 25 Mar 1998 10:33:37 -0800 (PST)


>Date: Wed, 25 Mar 1998 17:38:02 +0100 (GMT+0100)
>From: Rodrigo Ventura <yoda@isr.ist.utl.pt>
>To: LispOS Mailling List <lispos@math.gatech.edu>
>Subject: Re: Why LispOS?

>        Hum, is that really needed? You can run UNIX for years without
>a reboot. I assume our LispOS will have that for granted, and event
>not require reboot, but just a (gc).

  Only Unix system I've ever seen that could do that was one that was
idle. (Although my old Sun 3/60 running SunOS4.3.1 came pretty
close. Primarily because Sun stopped "fixing" the OS!)

  As for doing a gc on a lispm, NO, NEVER! It was faster to reboot the
machine every couple of months than do a full gc. The only time one
would do a full gc was right before dumping a world load. Now, if you
make everything persistant, then I guess you'd have to do a full gc to
clean things up as much as possible. I'd hate to see how long that'd
take on a machine with a couple gig of memory.


>
>        That's a great feature. Can it be said that it reaches the
>level of changing code on-the-fly, or even build a program while
>running/debuging it. It could speed-up enormously development speed.
>
>        Regards,

  The patch system was for saving your changes away so they could be
shared with others. (Or in case you wanted to reboot for some reason.)
Modification of running programs was often done on the fly, including
the "OS".

  Mike McDonald
  mikemac@mikemac.com