On 27 Nov, 12:37, Jon Clements <jon...@googlemail.com> wrote: > On Nov 27, 11:26 am, FelixCatus <fram...@gmail.com> wrote: > > > Good morning to all, > > I have written a simple python script that extracts data from a lot > > (800Mb) of text files. > > Now... In Linux the extraction runs in more or less 1s in Windows Xp > > it takes 325 - 326 s. > > I find that really hard to believe; I don't think I'd even be > convinced 1s is correct if all the files were mounted under a ramfs, > with blazingly fast RAM and huge amounts of multi-core CPUs running in > parallel.
I was not clear in my message, I don't have to open all files, but just a smaller amount of them. > I'd guess your code has a subtle flaw that: > 1) under Linux causes it to not do some work > 2) under XP causes it to do too much work > > Can you confirm you get the same results, and maybe post some code? > > Jon. Just now I have found the issues. First of all in the win xp system I have discovered that there is the same processor of the linux station but much more less ram and so the system uses intensive swapping. With my personal computer wich has a smaller amount of ram and a not so good processor (not the one on the mounting station) it took more or less the same time of the mounting station in the first try but I have discovered that the problem was the AVG Anti-virus that makes some analysis each time you open a file. Stopping the antivirus the execution time is reduced down to 56s. On monday I can try to disable the antivirus also on the mounting station and watch how many time it spends to process the files. In Linux I don't have the antivirus nor a lot of processes that runs concurrently as it happens in windows. So... Sorry for the inconvenience, this is not a real python issue but just a computer problem. Thanks for your support :-) -- http://mail.python.org/mailman/listinfo/python-list
