Hi Alain, > > move data around to make clusters A ... B empty
> > - CHKDSK surface scan should try reading sectors N times. If that fails, > > mark the sector as broken and try copying the data elsewhere, > FWIK this is what scandisk does, am I wrong? No. You cannot tell it MANUALLY to make a range of clusters 1. empty and 2. marked as broken. Another problem with MS SCANDISK surface scan is that it tries very often to read sectors before finally marking them as broken. It should be more pessimistic. On the other hand, it should try harder (try reading more often) to read sectors which are about to be marked as broken, when attempting to copy the contents to fresh sectors. > > - IF DEFRAG or SCANDISK would allow me to free and mark-as-broken sectors > > in the neighbourhood of the actually broken sectors > I did that for floppie a long time ago, but I doubt that it would wor > today because of simulated geometry. I believe that SPINRITE is the tool... Even with simulated geometry sector N-1 will be physically close to sector N in many cases, so the idea still makes sense. And of course disks which break in 2004 can be made in 1994 or before, using a known physical geometry. For example you could KNOW that the disk actually has X sectors per track. Then you could tell SCANDISK/CHKDSK to mark X sectors before and X sectors after the broken place as bad, even if you access the disk in LBA mode (where DOS does not care about actual geometry), would still be useful. > > a lot. So let us do better than MS - let us introduce configurable XCOPY > > critical error handling (and error messages which tell you WHICH files are > > the victims of read/write errors) > for a start a flag for "continue on errors" (with on screen/file log) > would be a big help > have a look at XXCOPY. (if you don find it I should have a copy) but it > is shareware... Well it is shareware, but it would not hurt to mention the URL of a page which describes it :-). > > and configurable SURFACE SCAN 'confidence' > > (be more pessimistic, but try hard to rescue data; > Spinrite did that in in old days (Seagate STR238 could not live without it) ... which gives extra evidence that such a (simple to add) configurability of 'try how often before marking as broken' and 'try how often to rescue data to fresh sector' for CHKDSK / SCANDISK would be useful :-). By the way: It would be really great if CHKDSK /R had some kind of surface scan progress indicator. Suggestion: Update the display after every track (floppy) / every 'chunk read' (I assume you try to scan 32 or more sectors in one access). Display the affected cluster numbers when errors are found. Should be fairly straightforward with a printf(...\r...). Eric ------------------------------------------------------- This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 Project Admins to receive an Apple iPod Mini FREE for your judgement on who ports your project to Linux PPC the best. Sponsored by IBM. Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php _______________________________________________ Freedos-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/freedos-devel
