A really bad idea.

[Daniel Dittmar]

sismex01 at hebmex.com wrote:
>>From: Daniel Dittmar [mailto:daniel.dittmar at sap.com]
>>zip, enumerate: readable, because you can search the docs for 
>>the words. It is also easy to implement variations: returning
>>other datatypes, returning an iterator to preserve memory. Thus
>>the whole program becomes easier to understand because similar
>>task are called in a similar way.
>>
>>list comprehension: not readable, because you don't know what to
>>search for.  You can't create dictionary comprehension or generator 
>>comprehension (both have been proposed on c.l.p), thus making the
>>resulting program less regular.


> But these are your *opinions*.  

No. These are criteria (or whatever the plural of criterion is), 
although of course there isn't a strict line between 'readable' and 'not 
readable'. Most Python programmers will agree that putting something 
into a library is generally a better choice than inventing a new syntax. 
Where some disagree is if building a list is so common that a new syntax 
makes the code actually more readable because the intentions are more 
explicit.

 > And while valid, others on the
> same list (or lists) have expressed nothing but rave reviews on
> the same things you don't like.

True. But the initial posters concern was that if more of these features 
make it into Python, than a specific quality and a specific user group 
(the casual Python programmer) may be lost.

Occasionally I have to change a Perl script, so I know what it is like 
if you can't look something up by a keyword.

> So, what to do?
> Macros, a-la Lisp, so anyone can have whatever syntax they want?
> A mishmash of features, like Perl?

Or perhaps nothing?

Daniel