Talk:Group code recording
From Wikipedia, the free encyclopedia
[edit] 1541's pioneering scheme of CAV + varying data rate
Perhaps it's worth noting that while the 1541 used constant angular velocity, it did vary its data rate to keep the data density near the maximum across all tracks. Compare this with the PC, which really did space bits further out on outside tracks, and CD-ROM drives, which started out Mac-style and changed to 1541-style in the mid/late 90s. --Anonymous
- GCR Didn't start with CBM. GCR was being developed before Commodore deployed it on their drives. The Durango Systems F-85 (introduced in late 1976) boasted 500K per 100 tpi floppy using 4/5 group code; eventually expanded to a double-sided option boasting almost 1MB per diskette. A few years before this, Sperry ISS was working on large hard drives for the mainframe business using group coding. --CG
GCR being thought of as originally a floppy format says something about which generation many Wiki editors (including me) belong to. It was around as early as 1973 (with the IBM 3420/3803 models 4,6, and 8), as a tape format for 6250 BPI drives. Wozniak re-invented a much simpler form for the Apple II with 5-and-3 and later 6-and-2 codes. I don't know where CBM got their code, which had 5 disk bits for 4 data bits; perhaps they didn't have an 8-bit data latch as a constraint.
Jackpot-- I found one of IBM's papers on the subject: http://www.research.ibm.com/journal/rd/186/ibmrd1806Q.pdf. IBMs GCR is way more complex than the CBM or Apple versions, as it not only satisfies physical constraints but provides error correction as well.
Nybbler 23:20, 25 February 2006 (UTC)
