to Readtable:CL or not to Readtable:CL

[Alfred M. Szmidt]

In System 130 lots of effort was done to convert ZL code into CL --
but what I would call failing at the attempt.

Why I consider it failing is because everything would just get
littered with explicit ZL:foo calls, quoting would be done in CL
manner, and .. that is about it.  I realise that this transition was
never completed, but I don't think we will ever do it either ...

My idea for CL compatibility -- which I think we really do need (I
don't want it .. but I also don't want to port peoples code to ZL just
because :-) would be to keep "Lisp Machine" files in ZL, but then
extend the CL package, readtable, etc to support CL better.

For larger systems like DEFSTRUCT, pathname handling, LOOP .. the idea
would be then to make those conform (as much we can ...) to CL.

"Basic" functions that exist in both CL and ZL but not compatible with
each other would live in a CL package.


Why I'm asking this so early is that there is a bunch of CLism being
done in the cold load, and the "back and forth" between ZL and CL is
annoying me since those are files I cannot (easily) match to System
130 (there is too much "random" noise).  But also sometimes I've had
files that literally all they do is change ZL to CL, fix quoting
.. replace \ with CLI:REM .. and such, which just is "meh".