Printable View
I have a VB function to read a file and create a list of objects. The program runs in 67 seconds on 98 with a P233 (32RAM), and in 56 seconds on a P200Pro running 98. When I run the program on NT with 128 RAM and PII333 it runs in 160 seconds. Any clues?
If the same machine this is the order for speed:
1) Windows 95
2) Windows 98
3) Windows NT
4) Windows 2000
Generally: NTFS is much slower than FAT32.
Michael Vlastos
Automation Engineer
Company SouthGate Hellas SA
Development Department
Athens, Greece
Not sure if that's always the order, since NT executes loops faster then win95 (98 i don't know...). For filereading I don't know though.
Crazy D @ Work :-)
I was talking by average. And I think, by experience, that I am right.
This forum is for Vb programming, and if we try to be more specific, concentrating on speed of execution of vb programs, I think again that 98 are more efficient (or even 95 if they support it).
Anyone's else experience???
Michael Vlastos
Automation Engineer
Company SouthGate Hellas SA
Development Department
Athens, Greece
I've never really noticed that much difference between them (although I tend to mainly use NT4) - it could be because the NT machine is running lots of other processes in the background though.
Chris Eastwood
CodeGuru - the website for developers
http://codeguru.developer.com/vb
Well I have done a few tests based on Michael's 95->98->NT->2000 proposition. My program timings show that a disproportionate amount of time is spent in a class module which reads line of text from a file on NT when compared with the same object in 98. I suspect that the readline function of VBs TextStream object is not efficient for NT and/or NTFS is very slow in relation to FAT32.
There are so many variables (especially if you are testing on different machines). However file I/O would be the most significant factor.
Questions:
The status of Find Fast on all machines.
The level of disk fragmentation on all machines.
The size of the disk drives on all machines.
The amount of free disk space on all machines.
The disk speeds (seek, rotation, latency, etc) on all machines.
The type and size of disk cache on the controllers.
The type of disk controller on each machine.
The list of disk I/O related factors just goes on and on.
Best bet. Run all your tests on the same machine.
Test under FAT16 on all OS's, Test under FAT32 on Win9x OS's. Test under NTSF on WinNT. Make sure background tasks and disk tasks (i.e., Virus scanning and Find Fast) are disabled if possible.
My gut feeling is the NT would be only slightly slower with all other factors the same.