… as many of you know, TRS-80 disks were never meant to hold data for 30+ years, and one of the things I do for people is read their disks into DMK images. Much of the time, people are trying to get their old documents and records off the disks, so getting the best possible read off the disks can mean the difference between a corrupt file or a good one.

To do the disk reading I have a dedicated system set up with a Catweasel, a floppy drive, DOS boot, and Tim Mann’s wonderful Catweasel utilities.

As any of you who have a Catweasel may know, Tim’s utilities allow for a command line option to give a specific number of retries for any track which reads with errors BUT it outputs the track as read on the LAST read rather than the track as read on the BEST read. You could have 19 reads of “9 good sectors, 1 bad sector” but if the 20th retry is “6 good sectors, 4 bad sectors” that is what you get, and 3 otherwise good sectors are lost (you can restart the process again, but it is all going to be a roulette game). Those sectors being good could easily mean the difference between a good file and a corrupted one.

The reason for this, however (and as Tim has previously told me), is that the Catweasel works only on a Track level, which is the rub.

I am wondering if anyone out there has the C++ know how to modify Tim’s CW2DMK.c program so that when a track is read (as currently written in the program), an array is populated with the actual sectors (which I guess might be one entry for any DMK header, and the remaining array equal to the track length divided by the sector size; but what do I know), and then if a successive retry produces a good sector where there was previously a bad one, the (bad) sector in the array is replaced with the newly successfully read one; and once the retry count is done, the array is assembled into a track; at which point the track is written out (as currently written in the program).

Tim has been busy with other projects (he has been well aware of the request for quite some time), and the track vs. sector nature of the Catweasel probably makes this a project he does not wish to tackle, but a recent set of disks where the person was desperate for their data has just triggered me to see if anyone out there can do it. Modifying the source code appears, on my quick read, to be permitted under the license under which it was issued, provided certain obligations are met.

However, it is Tim’s project, Tim’s source, and whether this is actually made public should be entirely up to Tim in my book. Certainly anyone who writes in the functionality is encouraged to send it to Tim; perhaps he will integrate it.

Links:

10 REM     PROGRAM TO FIND VERSION OF MODEL I ROM
20 CLS
30 FOR I=11264 TO 12287:V=PEEK(I):S=S+V:NEXT I: X=S/16
40 A=(X-FIX(X))*16:Y=FIX(X)/16:B=(Y-FIX(Y))*256
50 V=(A+B)
60 IF V=176 PRINT"ROM 1.0": GOTO 100
70 IF V=142 PRINT"ROM 1.1": GOTO 100
80 IF V=10 PRINT"ROM 1.2": GOTO 100
90 IF V=162 PRINT"ROM 1.3" ELSE PRINT"VERSION OF ROM UNKNOWN"
100 END
						

The 1st annual Vintage Computer Festival Southwest will be held August 7th and 8th, in the Rio Grande Ballroom, in the E.H. Hereford University Center, University of Texas Arlington. The event is sponsored by the Museum of Information Technology at Arlington, the UTA College of Engineering, the UTA Office of Information Technology and Vintage Tech. The VCF is an international event that celebrates the history of computing. There have been VCFs at several locations since 1997.