View Single Post
Old 20 March 2012, 23:28   #17
Global Moderator

prowler's Avatar
Join Date: Aug 2008
Location: Sidcup, England
Posts: 10,300
Well, I got the ReadADF program running today without much fuss, but it's MUCH slower than Disk2FDI Registered using the same hardware, and it's prone to errors where there are none using Disk2FDI.

The Disk2FDI cable uses two wires to carry signals between the onboard parallel port and the floppy disk ribbon cable, one of which is the same as the ReadADF wire, so it didn't take too long to prepare the machine for the test.

The Print1 program reported the wiring and lpt-port ok with one or both wires connected and it really didn't seem to affect the performance of the ReadADF program whether the additional wire was connected or not when I tried it. The program left the floppy drive activity light illuminated when it had finished, so I rebooted the machine to cancel that. However, the ReadADF program did not suffer from the same problem.

First, I tried the Disk2FDI program to make sure that the Amiga Computing coverdisk I had selected for the test gave no read errors, and then ran the ReadADF program with the default settings.

Personally, I found the program to be much too slow to be useful in anything but desperate circumstances, and I feel that the Disk2FDI registration fee would be money well spent if this were the only alternative.

Though, having said that, I'll admit that there are a number of features which give it some merit. I like the way the program displays an indication of its progress and which sectors it's having trouble with. Also, when a read error occurs, you have the option to (A)bort, (I)gnore, (R)etry or R(E)calibrate.

If you abort, you will have a partial ADF dump of all tracks imaged up to that point and, if you choose Retry, the program will persist with reading that track until it is successful. It seems to work in a similar way to the Catweasel ImageTool, in that the whole track is buffered in memory and only the bad sectors are updated, if read successfully, after the next retry. (The KryoFlux DTC could do with being rewritten to work the same way, instead of commencing each retry from scratch.)

I tried Racalibrate after one read error, but I wasn't sure whether it had a worthwhile effect (this will require more investigation), and I didn't try Ignore at all as that would have been counterproductive.

To sum up, the ReadADF program has two major drawbacks:
  1. It is sooooooo VERY slow (I haven't yet completed a dump), and
  2. The resulting ADF image is prone to errors, even where none are reported.
It might be possible to improve its speed and/or performance by using the manual command line switches provided and/or taking the parallel port signals directly from the motherboard header to shorten the wires and minimize any interference produced by the "antenna" effect, and I shall look at these possibilities next...

Last edited by prowler; 21 March 2012 at 00:19. Reason: The program is ReadADF! :banghead
prowler is offline  
Page generated in 0.03949 seconds with 10 queries