[Python-ideas] Default arguments in Python - the return

Terry Reedy tjreedy at udel.edu
Sun May 10 07:47:12 CEST 2009


Tennessee Leeuwenburg wrote:
> 
> > For me, Python 3k appears to be a natural place to do this. Python 3
> > still appears to be regarded as a work-in-progress by most people,
> > and I don't think that it's 'too late' to change for Python 3k.

Sorry, it *is* too late. The developers have been very careful about 
breaking 3.0 code in 3.1 only with strong justification.  3.1 is in 
feature freeze as of a few days ago.

>     Fortunately you're not Guido, and fortunately this isn't going to
>     happen. I recommend you either accept that this behaviour is here to
>     stay, or if you're *particularly* enamoured of late evaluation
>     behaviour of defaults, that you work on some sort of syntax to make it
>     optional.
> Thank you for the rest of the email, which was (by and large) 
> well-considered and (mostly) stuck to the points of the matter. I will 
> get to them in proper time when I have been able to add to the argument 
> in a considered way after fully understanding your points.
> 
> However, this last section really got under my skin. It seems completely 
> inappropriate to devolve any well-intentioned email discussion into an 
> appalling self-service ad-hominem attack.

I do not see any attack whatsoever, just advice which you took wrongly.

> ...se of Fortunately without a backing argument),

'Fortunately' as is clear from the context, was in respect to your 
expressed casual attitude toward breaking code.  Some people have a 
negative reaction to that.  In any case, it is a separate issue from 
'default arguments'.

 > attempt to bully me out of my position (recomment you accept this 
behaviour
 > is here to stay) are not appreciated.

He recommended that you not beat your head against a brick wall because 
of a misconception about what is currently socially possible.  He then 
suggested something that *might* be possible.  If that advice offends 
you, so be it.

Terry Jan Reedy





More information about the Python-ideas mailing list