TMPLee@DOCKMASTER.NCSC.MIL (05/22/89)
I'll grant this is obscure and I beg the forgiveness of the 99% of you on the list who could care less, but I dont' have any better idea of how to get this to at least one person who has a clue as to what is going on. First, to recapitulate: I discovered recently (I think after I changed how I was using /ram5 and /ramaepc, the ram that comes with the AE PC Transporter) that the second time I try to catalog (cat) files on /ramapec from within kermit all the files disappear, i.e., the first and second cat's show everything fine, the third shows the volume as empty. This problem does not occur if I cat /ram5, for example, several times, from within kermit, OR, or if instead I cat or catalogue /ramaepc from BASIC.SYSTEM. I'm using ProDos8 V1.7. I didn't notice what version of BASIC.SYSTEM; kermit is 3.85. I used ProSel's block warden to try to see exactly what was happening. I initialized the volume from within GS/OS and also filled in block 1 (where SOS would be) with $FF. I copied a small (half dozen blocks or so) file into /ramapec with the "old" finder. After the two "cat's" from kermit here is what happened: blocks 0 AND 1 were completely nulled. the volume creation date ($1C-$1F) of block 2 was nulled. the version number, $20, was nulled. -- question: in the clean version (before the cat's) this was "5" -- why? I saw something recently that said GS/OS no longer used the version number -- what does the 5 mean then? the file count ($25) was changed from 1 to 0. ($2B-4E), which had been the file info for the file I had transferrred, was all nulled. the bit map (block 6) was changed to show that only the first seven (0-6) blocks were used. the actual file (block 8 and thereafter) was still there, unchanged. It looks to me like near the end of the second cat command, or just conceivably whenever whatever next tries to go at /ramaepc, something, (the /ramaepc driver?) decides to re-initialize the volume, but that's only a speculation. Any ideas oh guru's???? Can anyone else reproduce this behavior? TMPLee@Dockmaster.ncsc.mil