I'm not proposing to remove the EV command. It is just that
some simple uses of it, say to replace x by 3, done by
ev(expression, x=3)
or
expression, x=3
are probably more clearly thought of, and computed as, substitutions.
ev(expression, somethingflag=true) is not a substitution, but
as you say, sometehing having to do with an environment.
Every system makes a hash of this. Rules in Mathematica,
random flags in Maple and Macsyma.
RJf
C Y wrote:
> --- Richard Fateman wrote:
>
>
>>I also suggest that we might add, in the documentation, notes that
>>give some advice, like "This command exists for compatibility with
>
> old
>
>>programs. We suggest you use ..... for new programs. {e.g. ev vs.
>>subst}" or "Some people expect this command to do ... but this is not
>>its objective. {e.g. factor}"
>
>
> I guess in my still-newbie mind I still find it a bit odd that subst is
> a replacement for ev. I guess I really shouldn't, but if we are really
> going to do away with ev I'd rather make it an inoperative command
> altogether, since (to my mind at least) it is rather seductive. When I
> think of the ev command, I think "defining my own local specialized
> environment within which this expression will be evaluated." What that
> implies I probably haven't thought through properly, and I'm not sure
> anyone has (or even if they really can). But I don't immediately think
> of "subst", which my mind parses as "use this to substitute something
> in for something else" as a substitute for ev. So one of these
> releases (not this one though) I'd like to see us decide some final
> resolution to this. Clearly the strongest proponent of ev was the old
> Maxima documentation, since it has little current support. Given
> another development cycle or so Maximabook will need the part relevant
> to ev revamped somewhat anyway, once I bring it back to life again, so
> I'm no longer worried about that. But I think the whole business of
> ev, subst, ratsubst, whatever the part commands are, etc. should be
> written up (probably as its own chapter) and clearly explained,
> defined, measured, whatever. Kill whatever isn't good, and add
> robustness where needed (in docs if no other place.) I think this
> concept will be an important one to hammer out with finalty early on,
> since a) a lot of other behaviors will depend on it and b) it's one
> even new users are likely to start using early on.
>
> Maybe it won't be so hard to sort out, but if ev is not a good command
> to use than IMHO it should die.
>
> CY
>
>
>
> __________________________________
> Do you Yahoo!?
> Yahoo! Mail - Find what you need with new enhanced search.
> http://info.mail.yahoo.com/mail_250
>
> _______________________________________________
> Maxima mailing list
> Maxima@www.math.utexas.edu
> http://www.math.utexas.edu/mailman/listinfo/maxima