[Python-Dev] Update PEP 292
bac at OCF.Berkeley.EDU
Fri Aug 20 04:10:44 CEST 2004
Raymond Hettinger wrote:
>>Since few people have commented either way, and taking myself out of
>>picture, it comes down to a battle royale between you and Brett. :)
> Hmm. Brett is taller. I'm better looking.
Don't let the photos from PyCON fool you; I never have a good look on my
face and I tend to look stoned.
> He's almost a master of computer science. I'm just old.
I hope you are not trying to use some "wiser" argument with that. =)
> He's needs a summer job to eat. I've got a retirement account.
Got me there.
> Perhaps, I can win him over through guile:
> 1. Call up IDLE, then press File, Open Module, string. Look at the
> source. Now, install your package and try again. No source.
Yeah, but you will notice the directory and realize it is a package so
look in there.
> 2. Run "cvs ann string.py" or "cvs log string.py" and look at who did
> what and when. Now, install the package and try again. No history.
Eh. That module has been only changed to keep up with the built-ins or
tweak a docstring. Not exactly major stuff I feel bad about losing some
commit messages over.
> 3. Look at the new string docs, decide that you want to subclass
> string.Template and want to look at the existing pattern. Time how long
> it takes you to find the source (change to the string subdirectory, look
> in __init__.py, learn the template.py is the real source, yada yada
Only reason that could be considered an issue is because the __init__.py
file has the constants and everything else that is not deprecated in there.
Basically I can live with having a single string module, but I would
like to see some real deprecation happen. The nice thing about the
pacakge separation is it made it clear in the code what would be going
away. If we at least at PendingDeprecation to what is supposed to be
taken out I will be happy.
More information about the Python-Dev