[MERGE] A hack to make urllib not call recv(1) lots and lots.

[SuperMMX]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Here is a more precise result. I use a VBS script to get CPU usage of 
bzr(python) every second, and get the average.

Bzr repo:

without the patch : 13.07%, 8m55.703s
with the patch    : 13.53%, 9m1.234s

2nd round:

without the patch : 12.74%, 8m44.421s
with the patch    : 12.57%, 8m42.156s

Emacs repo:

without the patch : 35.31%, 36m33.750s
with the patch    : 33.81%, 33m36.000s

So, it seems that the patch has no obvious improvement. If someone else
is interested, please post your testing result.


The script (downloaded somewhere via Google) :

strComputer = "."
Set objWMIService = GetObject("winmgmts:" _
    & "{impersonationLevel=impersonate}!\\" _
    & strComputer & "\root\cimv2")

set PerfProcess = objWMIService.Get(_
    "Win32_PerfFormattedData_PerfProc_Process.Name='python'")

While (True)
    PerfProcess.Refresh_
    Wscript.Echo PerfProcess.PercentProcessorTime
    Wscript.Sleep 1000
Wend

Save it as cpu.vbs, run it with "cscript cpu.vbs"

On Mon, 17 Mar 2008 11:13:04 +0800
SuperMMX <SuperMMX at gmail.com> wrote:

> Here is the rough number on Windows, without the patch, the CPU usage is
> about 45%, but with the patch, the CPU usage never exceed 10%
> 
> Any way to get the precise number ?

- -- 
A. Because it makes the logic of the discussion difficult to follow.
Q. Why shoudn't I top post?
A. No.
Q Should I top post?

A: Because it destroys the flow of the conversation
Q: Why is it bad?
A: No, it's bad.
Q: Should I top post in replies to mailing lists? 
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.8 (MingW32)

iEYEARECAAYFAkfeP8IACgkQYjhzyV/TMxvTIQCeNOXPa4td6XnMPwZEdGg0f5g+
AFIAnArPL+M5Fq25ze9pVUU8FSpzdxuK
=0ewo
-----END PGP SIGNATURE-----