>>>>> "Peter" == Peter Denno <[EMAIL PROTECTED]> writes:
Peter> Removing ~& is not the ideal way to go about it, but if you remove the ~&
here
Peter> it prints one fewer empty line. And I am hoping that we agree that that is
Peter> good. If so, I leave it to the developers to do it right.
Probably confusion in the printer when *error-output* and
*standard-output* are the same? It's equally annoying if there's an
incomplete line on *error-output* and you print Restart at the end of
it. Easy to miss.
Peter> But garbage collection isn't new anymore. Many popular languages have it
and
Peter> most programmers (at least the ones who might be attracted to lisp) know
that
Peter> it takes place occassionally. Printing the message seems excessively
Peter> pedagogical. With regards to most of what we are talking about, my opinion
is
Peter> this: Let them focus on THEIR problem first, and if subsequently GC or some
Peter> lispy thing becomes their most immediate problem, they can go read about
what
Peter> they are doing wrong.
As a somewhat experienced Lisper, I usually leave GC messages on.
And, in fact, I even add an extra line with a gc hook. Ok, so I might
be weird.
And it's easy enough for newbies to turn off in their .cmucl-init.lisp
files. And without GC messages, how will newbies know that they've
done something wrong? It just seems to take much longer than expected
to do simple things. And the time to run something keeps changing
magically because GC is silently happening. How useful is that?
Peter> (defvar ext::*top-level-auto-declare* t) ; In ./code/eval.lisp
Peter> So you are saying that the PRESENCE of the warning means
Peter> you are safe, right?
Of course not.
Peter> When you get the warning, you are declaring a new special and therefore
will
Peter> not be affecting existing definitions. Things will behave as they had. When
Peter> you DON'T get the warning, you may be setting a special defined elsewhere,
Peter> and thing might not behave the same anymore.
Peter> I don't think newbies are going to follow that line of reasoning. More
likely,
Peter> when they see a warning, they are going to think they are doing something
Peter> wrong.
They did do something wrong. The variable has been declared special,
and they better know what they're doing. (I, however, set
*top-level-auto-declare* to T, and if I setf a variable at
top-level, I always use *'s around it. But then I just usually do a
defvar anyway.)
>> Anyway, those are my opinions.
Peter> my recent experience. -- I have discussed this with another
Peter> cmucl user who would also very much like to see cmucl
Peter> quieter out of the box.
I must be unusual. I've never usually desired to see less. The
warnings and notes have always been helpful to me. Oh well.
Peter> shocking, they might stick with it. If I were to come to
Peter> cmlcu from Java or Python I think I'd be bewildered. It was
Peter> pretty tough for me just coming from
I'd think they be bewildered anyway, without or without all the notes.
Peter> I'd be happy to contribute to the development of such a document, but I'd
like
Your contributions to such a document would be appreciated.
Ray
