Condition System Rewrite: restarts

[David A. Moon <Moon@STONY-BROOK.SCRC.Symbolics.COM>]

    Date: Sun, 4 Jun 89 19:47:08 PDT
    From: Richard P. Gabriel <rpg@lucid.com>

    There seems little need for restart-case, if you ask me.

Restarts are certainly a weak point of this proposal.  However,
something like restarts is definitely required.  Current practice shows
that this is one of the most important parts of the condition system.
Also restart-case is definitely used in preference to restart-bind.  Of
course restart-case could have been defined by a user as a macro that
expands into restart-bind, just as handler-case could have been defined
by a user as a macro that expands into handler-bind.

MLY had an interesting proposal to integrate restarts and handlers,
however he pre-assumed that no one would listen to him and never
developed the proposal far enough for it to be considered as an
alternative to Kent's proposal.  Too bad, if you ask me.  I would have
liked to see restarts done differently, and would like to see something
done with the cleanup issue CONDITION-RESTARTS (although version 1
contains some things that are unacceptable).  However, in my judgement
even what is currently in the ANSI CL language is better than no
restarts.  It was some weak points, especially in the convenience
features, but it is not bankrupt.