From: Alan G. <ala...@gm...> - 2008-09-27 22:47:35
|
Hello, Why would an app compiled with py2exe run about six times slower than it does under the python interpreter? I've got a bunch of trace points throughout my python code, and have measured the average time between successive trace points. There are no real outlyers in the datasets for both the py2exe and python runs, just a 6-fold average increase in the time deltas for the py2exe run. This leads me to believe that there's no single point of slowdown under py2exe that needs investigation, just a sort of overall slowdown. Any ideas? Thanks, -Alan |
From: Grzegorz A. H. <gr...@ti...> - 2008-09-28 08:00:02
|
El 28/09/2008, a las 0:47, Alan Grow escribió: > Hello, > > Why would an app compiled with py2exe run about six times slower > than it does under the python interpreter? > > I've got a bunch of trace points throughout my python code, and have > measured the average time between successive trace points. There are > no real outlyers in the datasets for both the py2exe and python > runs, just a 6-fold average increase in the time deltas for the > py2exe run. This leads me to believe that there's no single point of > slowdown under py2exe that needs investigation, just a sort of > overall slowdown. Would it be possible for you to generate the smalles possible source code example that shows this behaviour on your computer so we can try reproduce it? |
From: alan g. <ala...@gm...> - 2008-10-01 03:55:06
|
On Sun, Sep 28, 2008 at 3:59 AM, Grzegorz Adam Hankiewicz <gr...@ti...> wrote: > Would it be possible for you to generate the smalles possible source > code example that shows this behaviour on your computer so we can try > reproduce it? Still working on this. A question: does py2exe add any overhead to C shared library calls? -Alan |
From: Alexander B. <bi...@uk...> - 2008-10-02 07:18:19
|
Alan Grow пишет: > Why would an app compiled with py2exe run about six times slower than it does under the python interpreter? > > I've got a bunch of trace points throughout my python code, and have measured the average time between successive trace points. There are no real outlyers in the datasets for both the py2exe and python runs, just a 6-fold average increase in the time deltas for the py2exe run. This leads me to believe that there's no single point of slowdown under py2exe that needs investigation, just a sort of overall slowdown. > > Any ideas? Try to turn off any antivirus software on your computer, and check again. |
From: Alexander B. <bi...@uk...> - 2008-09-28 06:50:54
|
Alan Grow пишет: > Hello, > > Why would an app compiled with py2exe run about six times slower than it does under the python interpreter? > > I've got a bunch of trace points throughout my python code, and have measured the average time between successive trace points. There are no real outlyers in the datasets for both the py2exe and python runs, just a 6-fold average increase in the time deltas for the py2exe run. This leads me to believe that there's no single point of slowdown under py2exe that needs investigation, just a sort of overall slowdown. > > Any ideas? Try to turn off antivirus software on your machine. |