New submission from Alexander Belopolsky <belopol...@users.sourceforge.net>:

There are several issues that prevent

$ python3 Tools/pybench/pybench.py -s 27.bench -c 32.bench

from working.

1. There is a bug in _pickle that prevents 2.x pickles that contain classic 
class instances from loading in 3.x.  This is addressed in issue5180.

2. Once _pickle issue is addresses, 2.7 benchmark files still cannot be loaded 
because they contain references to the Unicode tests and Unicode.py is not 
loadable in python3.

3. Once Unicode.py gets 2to3 treatment, we are still left with 

* Benchmark versions differ: cannot compare this benchmark to "32.bench" !


The last issue is probably the simplest.  Version number was bumped up in 
r69411, and a log entry says that the intent was to keep pybench 2.1 compatible 
with 2.x.  Indeed, pybench 2.1 from py3k runs under trunk python and resulting 
benchmark file can be compared with 3.x generated benchmark once issues 1 and 2 
above are resolved.  It may, however be worth consideration to synchronize 
pybench versions shipped with 27 and 32.  I understand that this is likely to 
be too late for 2.7, but I am checking 2.7 version in case this is considered 
appropriate for 2.7.1.

The Unicode.py issue was discussed under issue5180 in msg108819 and msg108821.  
I like MAL's suggestion of Strings3 module, but I think it would be useful to 
be able to compare 2.x str performance to both 3.x str and 3.x bytes.  
Similarly, I think String3 should have from __future__ import unicode_literals 
and be loadable in 2.7/2.6 and server as means to compare 2.x and 3.x unicode 
performance.

For the _pickle issue, I am about to attach a patch to issue5180.

----------
components: Demos and Tools
messages: 108834
nosy: belopolsky, lemburg, pitrou
priority: normal
severity: normal
stage: needs patch
status: open
title: pybench: Cannot compare 2.x and 3.x benchmarks
type: behavior
versions: Python 2.7, Python 3.2

_______________________________________
Python tracker <rep...@bugs.python.org>
<http://bugs.python.org/issue9102>
_______________________________________
_______________________________________________
Python-bugs-list mailing list
Unsubscribe: 
http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com

Reply via email to