MathGroup Archive 2001

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: Compound Expressions

  • To: mathgroup at smc.vnet.net
  • Subject: [mg29188] Re: [mg29178] Compound Expressions
  • From: Ben Jacobson <bjacobson at illumitech.com>
  • Date: Sat, 2 Jun 2001 05:54:42 -0400 (EDT)
  • Sender: owner-wri-mathgroup at wolfram.com

I get the same result with the input on the same line.  But then if I use
"Enter" to split the input into different lines, but still all a single
CompoundExpression, it works as expected:

Reproducing earlier result
---------------------------
In[1]:= a = 3; Remove["Global`*"]; a = 3; a

Out[1]= 3

In[2]:= a

Out[2]= a

Splitting input into different lines
-------------------------------------
In[3]:=a = 3; 
Remove["Global`*"]; 
a = 3; 
a

Out[6]=3

In[7]:=a

Out[7]=3

This is truly weird.

Ben Jacobson
Illumitech Inc.

At 04:15 AM 6/1/01 -0400, you wrote:
>I see that the following is what Mathematica does, so it's clearly what 
>Mathematica does:
>
>      In[147]:= a = 3; Remove["Global`*"]; a = 3; a
>
>      Out[147]= 3
>
>      In[148]:= a
>
>      Out[148]= a
>
>      In[149]:= a = 3; Remove["Global`*"]; Print[a]; a = 3; Print[a]
>
>      From In[149]:= Removed["a"]
>
>      From In[149]:= 3
>
>      In[150]:= a
>
>      Out[150]= a
>
>What I don't understand is, not what Mathematica does here -- I'm not 
>interested in tracing out the details; I believe it happens --but *why* 
>the language would have to be designed to function in this (utterly 
>bizarre and utterly confusing) way?
>
>To be more specific about this gripe:
>
>*  The "Remove" expression clearly removes "a" immediately when it is 
>encountered, as the output of  In[149] in particular confirms.
>
>*  The subsequent "a=3" expression then clearly recreates "a" and 
>assigns it a value *after* the "Remove" has been processed, as both 
>In[147] and In[149] confirm.
>
>So how does it make sense to have the "Remove" come back to life and 
>operate *again*, long after it should have been forgotten, so to speak?
>
>Does it *have* to be designed this way?
>
>


  • Prev by Date: Re: J/Link speed of accessing arrays
  • Next by Date: RE: I hate using Legend
  • Previous by thread: Compound Expressions
  • Next by thread: Re: Compound Expressions