Python-list Digest, Vol 18, Issue 208

[Jeff Shannon]

Martin v. Löwis wrote:

> Jeff Shannon wrote:
> 
>> I'd be in favor of that, unless someone can come up with a compelling 
>> current use-case for octal literals.
> 
> Existing code. It may use octal numbers, and it would break if they 
> suddenly changed to decimal. 

Right, which was my original point -- it was only in the context of 
Python 3.0 / 3K, when backwards compatibility is *already* being 
deliberately discarded, that getting rid of octal constants would be 
worth considering.

(I had specified *current* use-case to specifically indicate "other 
than backwards-compatibility and historical reasons" -- I know those 
are overriding until Py3K, but if the whole point of Py3K is to ditch 
all the "features" that exist only for backwards-compatibility and 
historical reasons, then...)

Jeff Shannon