Message 270251 - Python tracker

➜

This issue tracker has been migrated to GitHub , and is currently read-only.
For more information, see the GitHub FAQs in the Python's Developer Guide.

In-reply-to
Author	gvanrossum
Recipients	gvanrossum
Date	2016年07月12日.15:54:01
SpamBayes Score	-1.0
Marked as misclassified	Yes
Message-id	<1468338842.06.0.843714859349.issue27498@psf.upfronthosting.co.za>

Content
In Python 3.6, the repr() of a class includes its memory address. This is going to cause a lot of problems for tests in 3rd party code that expects the nice and clean <class '__main__.C'> instead of <class '__main__.C' at 0x7feb0360f568>. I understand the desire to provide more clarity in case somehow two class objects created at different times have the same name -- but I'm not sure I like to rub the hex address in the user's face all the time. Can we please roll this back or think about a better way to reveal this that won't break so many 3rd party tests?

Content

In Python 3.6, the repr() of a class includes its memory address.
This is going to cause a lot of problems for tests in 3rd party code that expects the nice and clean <class '__main__.C'> instead of <class '__main__.C' at 0x7feb0360f568>.
I understand the desire to provide more clarity in case somehow two class objects created at different times have the same name -- but I'm not sure I like to rub the hex address in the user's face all the time.
Can we please roll this back or think about a better way to reveal this that won't break so many 3rd party tests?

History
Date	User	Action	Args
2016年07月12日 15:54:02	gvanrossum	set	recipients: + gvanrossum
2016年07月12日 15:54:02	gvanrossum	set	messageid: <1468338842.06.0.843714859349.issue27498@psf.upfronthosting.co.za>
2016年07月12日 15:54:02	gvanrossum	link	issue27498 messages
2016年07月12日 15:54:01	gvanrossum	create

homepage