ttang@puff.cs.wisc.edu (Theodore Tang) (02/19/88)
The index hole on the floppy disk is used only for formatting purposes so all tracks remain in sync. When the disk is read, the hole is not used. I'm not sure about writes. Theodore Tang University of Wisconsin at Madison USENET: ttang@puff.wisc.edu.UUCP FIDONET: 1:121/3 (Opus's Internat'l Archives BBS) BBS: Opus's Internat'l Archives BBS (608)251-4755 9600 USR HST MNP 5 "No, no, it wasn't me!" -anonymous
greg@gryphon.CTS.COM (Greg Laskin) (02/21/88)
In article <1412@puff.cs.wisc.edu> ttang@puff.cs.wisc.edu (Theodore Tang) writes: >The index hole on the floppy disk is used only for formatting purposes so all >tracks remain in sync. When the disk is read, the hole is not used. I'm not >sure about writes. > Don't do a term paper on index holes without some more research. The index hole to which you refer, as opposed to index holes that demarcate sectors on a hard sectored diskette, allows the controller to determine: 1) That the diskette is moving. 2) That the diskette has made a complete revolution (the hole went by twice.) In addition, it provides a convenient start-of-track reference marker when formatting the diskette. Diskette tracks are not in sync with anything, by the way. If, for example, you were looking for sector 4 and two index holes went by, it would be a good bet that you weren't going to find sector 4. Thanks to our friend, the hole, sectors don't have to be recorded sequentially on the diskette. If you see two index holes while formatting a track, you've written too much formatting on the track, although many controllers don't recognize this one. -- Greg Laskin "When everybody's talking and nobody's listening, how can we decide?" INTERNET: Greg.Laskin@gryphon.CTS.COM UUCP: ihnp4!scgvaxd!cadovax!gryphon!greg rutgers!marque!gryphon!greg codas!ddsw1!gryphon!greg
ttang@puff.cs.wisc.edu (Theodore Tang) (02/22/88)
In article <2660@gryphon.CTS.COM>, greg@gryphon.CTS.COM (Greg Laskin) writes: > In article <1412@puff.cs.wisc.edu> ttang@puff.cs.wisc.edu (Theodore Tang) writes: > >The index hole on the floppy disk is used only for formatting purposes so all > >tracks remain in sync. When the disk is read, the hole is not used. I'm not > >sure about writes. > > > Don't do a term paper on index holes without some more research. > > The index hole to which you refer, as opposed to index holes that > demarcate sectors on a hard sectored diskette, allows the controller > to determine: > 1) That the diskette is moving. > 2) That the diskette has made a complete revolution (the hole went by > twice.) > In addition, it provides a convenient start-of-track reference marker > when formatting the diskette. Diskette tracks are not in sync with > anything, by the way. If FORMAT uses the index hole as a reference to the start of track, I do believe that is syncronizing the tracks so they all start right after the appearance of the index hole. > If, for example, you were looking for sector 4 and two index holes > went by, it would be a good bet that you weren't going to find sector 4. > Thanks to our friend, the hole, sectors don't have to be recorded > sequentially on the diskette. First of all, try this out on your PC. Cover the index hole of any formatted disk. If it can boot, boot it. Surprise! Your PC is reading the disk quite well without the aid of the hole. Try DIR, surprise again! It also works and it didn't even have to use the index hole. Now try to create a file. Hey! That works too, and it didn't even complain with an error like INDEX HOLE COVERED STUPID, I CAN'T READ YOUR DISK. The ONLY time DOS uses that index hole is when it formats. At least we agree on that. Hole or no hole, sectors are scewed (scrambled) so access time is faster to find sectors. > If you see two index holes while formatting a track, you've written too > much formatting on the track, although many controllers don't recognize > this one. > > -- > Greg Laskin Theodore Tang University of Wisconsin at Madison USENET: ttang@puff.wisc.edu.UUCP FIDONET: 1:121/3 (Opus's Internat'l Archives BBS) BBS: Opus's Internat'l Archives BBS (608)251-4755 9600 USR HST MNP 5 "No, no, it wasn't me!" -anonymous
wheels@mks.UUCP (Gerry Wheeler) (02/23/88)
In article <1414@puff.cs.wisc.edu>, ttang@puff.cs.wisc.edu (Theodore Tang) writes: > The ONLY time DOS uses that index hole > is when it formats. Well, that is OK on a good disk, but what about a disk with a bad sector? I think the previous poster was right -- the disk controller uses the index hole as a timeout if it can't find the requested sector. If you request sector 2, and the controller hasn't found sector 2 after seeing a couple of index pulses, it has to assume sector 2 isn't there. -- Gerry Wheeler Phone: (519)884-2251 Mortice Kern Systems Inc. UUCP: uunet!watmath!mks!wheels 35 King St. North BIX: join mks Waterloo, Ontario N2J 2W9 CompuServe: 73260,1043
ttang@puff.cs.wisc.edu (Theodore Tang) (02/24/88)
In article <404@mks.UUCP>, wheels@mks.UUCP (Gerry Wheeler) writes: > In article <1414@puff.cs.wisc.edu>, > ttang@puff.cs.wisc.edu (Theodore Tang) writes: > > The ONLY time DOS uses that index hole > > is when it formats. > > Well, that is OK on a good disk, but what about a disk with a bad > sector? I think the previous poster was right -- the disk controller > uses the index hole as a timeout if it can't find the requested sector. > If you request sector 2, and the controller hasn't found sector 2 after > seeing a couple of index pulses, it has to assume sector 2 isn't there. > -- > Gerry Wheeler Phone: (519)884-2251 The keyword here is TIMEOUT. This is exactly what DOS does, it will time, not wait for the hole to reappear. It doesn't have to use the hole, just know that it take x usecs for a complete pass and if it hasn't found the desired sector by that time, bingo!, you get a TIMEOUT. This is kind of like when you don't close the drive door. Theodore Tang (Ted) @ University of Wisconsin at Madison TEL: (608) 251-4325 (voice) Bang: ...!uwvax!geowhiz!uwspan!circle!121!3!sysop UUCP: ttang@puff.wisc.edu.UUCP (uw puff) ttang@gosset.wisc.edu.UUCP (uw gosset) Fidonet: Sysop, 1:121/3 Opus International Archives HST (608) 251-4755
mjg@ecsvax.UUCP (Michael Gingell) (03/02/88)
In article <1412@puff.cs.wisc.edu>, ttang@puff.cs.wisc.edu (Theodore Tang) writes: > The index hole on the floppy disk is used only for formatting purposes so all > tracks remain in sync. When the disk is read, the hole is not used. I'm not > sure about writes. > > Perfectly correct. The PC controller does not use the index hole if it is reading or writing unless you have a bad disk. In that case it may time out after a certain number of tries. On older disk drives you can actually cover the index hole with an opaque write protect tab and still read or write. However you need the index hole for formatting. However modern half height disk drives have more sophisticated electronics which times the intervals between successive index hole pulses. If it is not 200 millisecs +/- 5% then the drive says to itself "Oh - no disk in drive". It then lowers the drive ready line and disables read/write signal lines. Mike Gingell ...ecsvax!mjg
martyn@garfield.UUCP (Martyn Quigley) (03/05/88)
In article <4709@ecsvax.UUCP> mjg@ecsvax.UUCP (Michael Gingell) writes: >In article <1412@puff.cs.wisc.edu>, ttang@puff.cs.wisc.edu (Theodore Tang) writes: >> The index hole on the floppy disk is used only for formatting purposes so all >> tracks remain in sync. When the disk is read, the hole is not used. I'm not >> sure about writes. > >On older disk drives you can actually cover the index hole with an >opaque write protect tab and still read or write. However you need >the index hole for formatting. > Any comments about 3.5" discs? I have just formatted one, and I can't find any index hole in it.
ray@micomvax.UUCP (Ray Dunn) (03/09/88)
[Please excuse late news processing] To protect the rude and the ignorant, I shall leave off names... Mr A: >The index hole on the floppy disk is used only for formatting purposes so all >tracks remain in sync. When the disk is read, the hole is not used. I'm not >sure about writes. Mr B: > Don't do a term paper on index holes without some more research. > > The index hole to which you refer, as opposed to index holes that > demarcate sectors on a hard sectored diskette, allows the controller > to determine: > 1) That the diskette is moving. > 2) That the diskette has made a complete revolution (the hole went by > twice.) > In addition, it provides a convenient start-of-track reference marker > when formatting the diskette. Diskette tracks are not in sync with > anything, by the way. Mr A: >First of all, try this out on your PC. Cover the index hole of any formatted >disk. If it can boot, boot it. Surprise! Your PC is reading the disk quite >well without the aid of the hole. Try DIR, surprise again! It also works >and it didn't even have to use the index hole. Now try to create a file. > Hey! >That works too, and it didn't even complain with an error like INDEX HOLE >COVERED STUPID, I CAN'T READ YOUR DISK. >The ONLY time DOS uses that index hole >is when it formats. At least we agree on that. > If you guys would stop shouting at each other, you Mr A would learn something, and you Mr B would be able to communicate your ideas better and receive some respect from your peers. The **facts** are quite simple, and are as stated by Mr B: On an IBM type disk drive and controller: On a soft sectored diskette, the index hole is used as a start of track marker to indicate to the FORMAT process when it should start writing the soft sectoring information. Within the track, the sectors can be ordered in any way you like. I know of no disk controller which will abort the track format command if the index hole is seen again during the track format. During disk READING, the required sector is searched for by reading the sector headers written during the formatting, the controller does not need to wait for the index hole before it starts its search. If the sector is found *BY THE TIME THE INDEX HOLE IS SEEN FOR THE SECOND TIME* then the data is read. Otherwise an error is flagged, as the addressed sector cannot be found. A similar process takes place when the controller is searching for the sector header of the sector to be written. If you cover the index hole and then attempt to read an illegal sector, or one whose header has been corrupted, the search will continue ad pensionem! If I remember the distant past, the 1791 controller also used the index hole to determine whether a disk was loaded & spinning. Thus Mr A, it is only when things start to go *wrong* that the index hole has a bearing on reading or writing. But it is *IS* required!!! Stop shouting - start thinking!!! If you Mr B had not opened your response with an insult, Mr A might have listened to you!! Ray Dunn. ...{philabs, mnetor, musocs}!micomvax!ray
ray@micomvax.UUCP (Ray Dunn) (03/12/88)
In article <1424@puff.cs.wisc.edu> ttang@puff.cs.wisc.edu (Theodore Tang) writes: >The keyword here is TIMEOUT. This is exactly what DOS does, it will time, >not wait for the hole to reappear. It doesn't have to use the hole, just know >that it take x usecs for a complete pass and if it hasn't found the desired >sector by that time, bingo!, you get a TIMEOUT. This is kind of like when >you don't close the drive door. > >Theodore Tang (Ted) @ University of Wisconsin at Madison > RIGHT AND WRONG!! The *CONTROLLER* uses the index hole to know it isnt going to find the requested sector on a spinning disk. Under this circumstance, an error is returned to DOS which does *NOT* have to wait for a timeout to occur, the error will be reported within 2 revolution times. Under other circumstances, when the controller does not give an error return to DOS, *THEN* DOS will timeout! Ray Dunn. ..{philabs, mnetor, musocs)!micomvax!ray