Author larry
Recipients amaury.forgeotdarc, eric.smith, larry, mhammond, paul.moore
Date 2009-05-05.21:35:09
SpamBayes Score 2.95266e-06
Marked as misclassified No
Message-id <1241559312.58.0.0992501922376.issue5799@psf.upfronthosting.co.za>
In-reply-to
Content
I've never understood what is the "philosophical problem" per se.  It's
clear from his implementation--Guido created "splitunc" when he removed
my patch--that he thinks these should be precise string operations. 
Whereas I propose making these slightly higher-level abstract operations
on "paths" with "drives" (really "mount points").  I think my approach
is more pleasant to use.

As for whether or not "\\foo" should be considered an absolute
path--perhaps my 1999 patch changed this behavior, but this patch does
not.  This:
  >>> import ntpath
  >>> ntpath.isabs("//foo")
  True
is unchanged by the attached patch.  (Or, at the very least, the latest
patch.)

At the time I guess I did a poor job of demonstrating the use case, but
I think the chorus of +1s shows there is one.

On the other hand, I'm not a Windows programmer anymore, and I barely
touch Windows boxes.  I'm doing this out of the kindness of my heart ;)
History
Date User Action Args
2009-05-05 21:35:12larrysetrecipients: + larry, mhammond, paul.moore, amaury.forgeotdarc, eric.smith
2009-05-05 21:35:12larrysetmessageid: <1241559312.58.0.0992501922376.issue5799@psf.upfronthosting.co.za>
2009-05-05 21:35:09larrylinkissue5799 messages
2009-05-05 21:35:09larrycreate