[wxPython] wxPython, slowness, and Novell fileservers
This is a further comment on my earlier query about the slowness of
wxPython when accessed on our fileserver from public labs. Memory
consumption is much less on the workstations that run Python slowly
despite their having far more memory.
In our public labs the memory being used by Python with (wxPython
modules loaded) is far lower than when Python is being run locally
on a workstation or when it is run across our subnet. On my own
workstation (AMD K6 2 300 with 64 megabytes), where Python is local,
and on a more or less, equivalent Pentium 2 266 Mhz, where
Python is being accessed on our fileserver across our subnet, memory
consumption begins at about 7.4 megabytes and rises to about 8.2
megabytes (I'm not being very exact here, and there seemed to be a
difference between the two machines).
On the 400Mhz Pentium 2s with 128 megabytes in the public labs,
memory consumption begins at about 4.2 megabytes and goes up to
about 5.2 megabytes. This is similar to what I noticed on an old
machine in our subnet with only 32 megabytes of RAM. The old machine
had to do a lot of paging when it was accessing Python on our
fileserver, but it still did better than the much fast machines in
Thus it looks to me as if there is something about the set up on the
workstations in our public labs and/or bottlenecks in the network
that forces Python to behave as if available memory is restricted.
Has any user of this list got an idea about what it could be?
University of Dundee
[hidden email] tel 01382 344539