80-Bus News |
July–August 1984 · Volume 3 · Issue 4 |
Page 3 of 43 |
---|
Adrian Perkins writes in the March-April 1984 issue of 80-BUS News about his defection from HS-1N to disk, and mentions a problem with ageing drives. As co-designer of the project, I have been using the same two drives regularly since 1980, and have found them remarkably reliable – hence I have not yet succumbed to the pressure to move to disks. They do, however, tend to drift away from the specified speed with age, and difficulties may arise when an old tape is edited.
HS-1N tapes are initialised by writing a blank catalog at the beginning of the tape; apart from this the tapes are not formatted in any way. At this stage all 28 remaining 2k blocks are available, and when a file is saved the blocks at the beginning of the tape are filled first, always back-spacing one block and running forwards before writing the next block (to allow for a check-read and to get the tape running up to a consistent speed).
If large files are used (say 20k source files, occupying 10 HS-1N blocks on tape) then it may be possible to fit only two on one side of a tape. When the first file is deleted, blocks 1 to 10 on the tape are released for re-use, and if the drive is running at a different speed corruption may take place.
The most likely problem to arise involves a slowing of the drive. In this case, when ten new blocks are saved there may be a residue of the original tenth block which is not overwritten by the new tenth block. Attempts to read the eleventh block (the first block of the original second file) may fail because of this, but you may be able to recover by advancing the tape over the garbage thus:–
If the drives are running faster than before, the tenth block of the new file may overwrite the first block of the previous one, and this block is therefore lost for ever. By using the above technique, and editing the HS-1N catalogue workspace, you may be able to recover the later blocks. Data loss from fast drives is less likely than difficulties with slow drives, as the new file must extend across the entire inter-block gap on the tape, before it can corrupt the next block.
I should stress that there is no need to worry about reading tapes from a slow drive on a fast one, or vice-versa, only about editing tapes on a drive with a different speed. It is best to use a blank (either new or deleted) tape when saving edited versions of old files.
Yours sincerely, Dr M.D. Hendry, Cupar, Fife.
Perhaps the following news/viewpoints would be of interest:–
a) There have been several recent letters suggesting that Lucas are not very helpful on some occasions. I can only say that over the last 6 months or so, I have written to them 3 times asking for help or information on various topics, and on each occasion I have received a courteous and helpful response.
Page 3 of 43 |
---|