Re: [Maxima-users] What is special about the variable c ?

In case anyone cares, the origin of this misbehavior has
to do with the fact that Macsyma was written on a Lisp
system (Maclisp, c. 1966) which had only upper case. In
fact, the computer (PDP-6) did not really support upper and
lower case.

When it became obvious that the world outside MIT allowed
upper and lower case, MIT still fought for upper case only.
That is why the ANSI standard Common Lisp, by default, changes
all input to upper case.

Realizing eventually that the world outside MIT PREFERRED
upper and lower case, some effort was made to allow Macsyma
to use both cases (a global symbol bothcases  was introduced).
But the effort was half-hearted and incomplete.


RJF

Helge Hafting wrote:

> James Amundson wrote:
>
>> On Tue, 2004-08-17 at 07:46, Helge Hafting wrote:
>>  
>>
>>> The documentation states that capitalization matters, i.e. x and X 
>>> are separate.
>>>   
>>
>>
>> Actually, it's more complicated than that. Lowercase symbols are
>> converted to uppercase when they are first created if the uppercase
>> version already exists. There is general (if not universal) consensus
>> that this behavior needs to be changed. We plan to make case sensitivity
>> straightforward (x and X will always be different) in Maxima 5.9.2.
>>
>>  
>>
>>> Thats fine, but this seems to not apply to "c"?
>>>   
>>
>>
>> There are several symbols that get mysteriously upcased because of the
>> behavior defined above. "c" is one of them. (Actually, I think that "c"
>> will no longer be mysteriously upcased in 5.9.1, but it's really an
>> accident.)
>>  
>>
>>> (C1)  a+A+b+B+c+C+d+D;
>>> (D1)  d + D + 2 C + b + B + a + A
>>>   
>>
>>
>> Ouch. This is a particularly bad example.  
>>
> I discovered it while trying out some simple ax²+bx+c stuff.
>
>> It's partly bug, partly bad design. This sort of weird behavior should
>> be fixed once and for all in the 5.9.2 release of Maxima.
>>
>>  
>>
> Good to hear.  Not only known stuff, but also a planned fix. :-)
>
>> --Jim
>>
>> P.S. Even though there is a legacy maxima-users list on sourceforge, all
>> the real discussion goes on in maxima@www.ma.utexas.edu.
>>  
>>
> Thanks.  I found sourceforge via information from the debain package I 
> installed.
>
> Helge Hafting
>
> _______________________________________________
> Maxima mailing list
> Maxima@www.math.utexas.edu
> http://www.math.utexas.edu/mailman/listinfo/maxima