zgel05@apctrc.trc.amoco.com (George E. Lehmann) (08/22/89)
I've enabled my thermal-retard shields, because I know this is probably going
to stir some flames from somewhere, but:
DON'T USE 3.5" DS/DD DISKETTES FOR HD (1.44MB) USAGE!!!!!
In my dealings with a number of customer sites, I have to read backups of
databases sent through the mail on floppy disks. Invariably, if the disk is
a 5.25 DS/DD formatted at 1.2MB, or a 3.5 DS/DD formatted at 1.44MB, I will
not be able to read the entire disk without one or more sector errors. (The
stupid IBM PS/2 machines will even format a DS/DD disk at 1.44 MB *without*
the extra hole, and all other machines in the world won't read it, cause if
it doesn't have that extra hole, it *can't* be formatted at 1.44, right?!?!?)
The bottom line is, you will probably get away with it for a while, maybe for
a long time. This seems to be more of a problem between machines, so I would
be especially cautious if these diskettes were being used for backups which
might have to be read on another machine. In the final analysis, I wouldn't
put *anything* on one of these "user-enhanced" diskettes that I would even
slightly miss if the disk went bad.
-----------------------------------------------------------------------------
To live and not fly is not living...
George Lehmann PP/ASEL/IA ...!uunet!apctrc!zgel05 -or- zgel05@trc.amoco.com
Amoco Production Co. PO BOX 3385, Tulsa, Ok 74102 Voice:918-660-4066gjp1@mtuxo.att.com (XMRK4-G.PATEMAN) (08/22/89)
In article <1001@apctrc.UUCP> zgel05@apctrc.trc.amoco.com (George E. Lehmann) writes: >DON'T USE 3.5" DS/DD DISKETTES FOR HD (1.44MB) USAGE!!!!! AND DON'T USE 5.25" DS/DD DISKETTES FOR HD (1.2MB) USAGE!!!! OK, once more for those who still believe it's a cost-effective approach: When you format a low density (LD) disk in a high density (HD) drive, a large number of sectors will be marked as "bad". They're not actually bad, of course. It's just that the 300 Oersted magnetic coating of an LD disk can't hold the signal the same way an HD's 600 Oe coating can. Look at the two types side-by-side. The HD disk has a noticeably darker coating. Alright, so you've formatted the 360K at 1.2M and all the "bad" sectors have been marked. You don't quite have 1.2M but you have a lot more than 360K. The problem is that many of the marginal sectors which read back successfully THIS TIME won't necessarily be readable later on when the disk contains your valuable data. How likely is that, you ask? VERY! An associate of mine used LDs exclusively on his HD machine and found that more than half of them contained unreadable files. When he tried to read them on another machine, the situation was even worse. If the read error occurs in a text file you might, with the help of the Norton Utilities, Mace Utilities, etc, be able to piece it back together. If it occurs in a binary file, you probably won't. If it occurs in a critical area like a directory or the file allocation table, you can probably kiss most or all of the disk goodbye. Is it really worth the difference in price between HD and LD media? ------------------------------------------------------------------------------ George J. Pateman CAHTP Media Lab AT&T / Bell Labs Room 4D-333 200 Laurel Ave. Middletown, NJ 07748
bill@bilver.UUCP (Bill Vermillion) (08/23/89)
In article <5221@mtuxo.att.com> gjp1@mtuxo.UUCP (XMRK4-G.PATEMAN) writes: > > ... It's just that the 300 Oersted magnetic coating of an LD >disk can't hold the signal the same way an HD's 600 Oe coating can. Actually it is the higher current causing self-erasure on the inner tracks because of the lower coercivity of the DD media. Because the HD has higher coercivity we can resolve a "tighter pattern", but require the higher current to write. It's not the retentivity that makes the real difference but that we have to pound on it much harder to make it accept the signal in the first place. -- Bill Vermillion - UUCP: {uiucuxc,hoptoad,petsd}!peora!tarpit!bilver!bill : bill@bilver.UUCP Please use either of the above in replies. The Reply To: may be bogus!
pete@Octopus.COM (Pete Holzmann) (08/23/89)
In article <5221@mtuxo.att.com> gjp1@mtuxo.UUCP (XMRK4-G.PATEMAN) writes: >(George E. Lehmann) writes: >>DON'T USE 3.5" DS/DD DISKETTES FOR HD (1.44MB) USAGE!!!!! > >AND DON'T USE 5.25" DS/DD DISKETTES FOR HD (1.2MB) USAGE!!!! > >OK, once more for those who still believe it's a cost-effective approach: Let's pause for breath here, please. It is fully proven that 360K floppies are not reliable when written in a 1.2M drive. Not even at 360K, let alone at 1.2M. I have never, ever, seen a 360K floppy formatted at 1.2M come even *close* to getting the full 1.2M available, and I've personally done extensive testing to verify that a 360K diskette written (at 360K density) on a 1.2M drive is simply not *always* going to be readable on another drive. It works much of the time, but not always. But 3.5" drives are a completely different matter. There is far more compatibility! A 1.44M drive can reliably read and write the correct disk at the correct density. There is no question of that. There's absolutely no need to buy both a 1.44 and a 720K drive. In addition, whenever a 720K floppy is formatted at 1.44M, you don't just get a little more than 720K out of it, you get the whole 1.44M, or very close to it (same as you do when formatting a "real" 1.44M diskette). So the real-world situation as we see it so far is: 5.25" disks: no *reliable* compatibility between 360K and 1.2M-- o Never get full 1.2M when formatting a 360K at 1.2 o 360K diskettes written on a 1.2 in any mode (360 or 1.2) are not guaranteed to be readable on other drives 3.5" disks: complete compatibility in the short view (what we've seen so far: o Almost always get full 1.44 when formatting 720 at 1.44 o 720K diskettes written on a 1.44 drive are always readable when written at 720K, and the majority of responses I've seen so far indicate that they are readable when written at 1.44M. So. Please let's limit the discussion to the remaining unknown, which can be verified by experiment (manufacturer's experiment is fine with me, but perhaps different diskette manufacturers have different quality diskettes!) The only unknown about 720-on-1.44 is: - What is the reliability of a 720K diskette formatted at 1.44M? I would personally love to see postings from people who have experience with this setup. We've heard from a few people who've just started down that road, and are very positive. How about people who've done it longer than a month or two? Please, postings like the following only confuse the issue: >When you format a low density (LD) disk in a high density (HD) drive, >a large number of sectors will be marked as "bad". They're not actually >bad, of course. It's just that the 300 Oersted magnetic coating of an LD >disk can't hold the signal the same way an HD's 600 Oe coating can. Look >at the two types side-by-side. The HD disk has a noticeably darker coating. [The last sentence is true. But the first one is demonstrably false. Does this guy really talk from experience, or from theory?] >Alright, so you've formatted the 360K at 1.2M and all the "bad" sectors have >been marked. You don't quite have 1.2M but you have a lot more than 360K. >The problem is that many of the marginal sectors which read back successfully >THIS TIME won't necessarily be readable later on when the disk contains your >valuable data. How likely is that, you ask? VERY! An associate of mine >used LDs exclusively on his HD machine and found that more than half of them >contained unreadable files. When he tried to read them on another machine, >the situation was even worse. [If this 'associate' was working in the 5.25" world, of COURSE he had trouble. 360K and 1.2M diskettes simply don't mix reliable on the same drive at all. That has nothing at all to do with 3.5" drives...] I'm sorry this turned into such a long posting, especially since I really have little knowledge to add to the discussion. I've only started using 720K in 1.44M drives recently myself, and don't have time to do my usual experiments- to-find-the-real-answer. [Somebody ought to get a box of each of several brands of 720K disks, do an archival backup on each box at 1.44M, then report on the readability of the disks over the next year (or more!)...] Pete -- Peter Holzmann, Octopus Enterprises |(if you're a techie Christian & are 19611 La Mar Ct., Cupertino, CA 95014 |interested in helping w/ the Great UUCP: {hpda,pyramid}!octopus!pete |Commission, email dsa-contact@octopus) DSA office ans mach=408/996-7746;Work (SLP) voice=408/985-7400,FAX=408/985-0859
efv@mbunix.mitre.org (Eugene F. Vogt) (08/24/89)
In article <1989Aug23.151302.2622@Octopus.COM> pete@octopus.UUCP (Pete Holzmann) writes: ..... lots of detail about 5.25 & 3.5 disks deleted ...... > >I would personally love to see postings from people who have experience with >this setup. We've heard from a few people who've just started down that road, >and are very positive. How about people who've done it longer than a month >or two? > > >-- >Peter Holzmann, Octopus Enterprises |(if you're a techie Christian & are >19611 La Mar Ct., Cupertino, CA 95014 |interested in helping w/ the Great >UUCP: {hpda,pyramid}!octopus!pete |Commission, email dsa-contact@octopus) >DSA office ans mach=408/996-7746;Work (SLP) voice=408/985-7400,FAX=408/985-0859 I have had a PS/2 50 since July 1987. Within an hour of unpacking it, I discovered the inability of the IBM drive to know which disk was which. I can format DD disks to 1.44M, or HD disks to 720k, or vice versa. I bought ONE box of HD disks (at $58!!!!) when I bought the machine. Turns out I needed at least one. Apparently the drive in the PS/2 can sense which disk is in the drive (somehow), but in all but one case chooses to ignore it. When you make a backup copy of the IBM Reference Disk, it won't work with a 720k disk. Has to be a HD disk. Anyway, I used DD disks as high density disks (Nashua brand are my favorite) EXCLUSIVELY for about 1.5 years. We're talking 60 or more disks, some used for compressed hard disk backup (PCTools put 1.6M on a 1.44M disk!), some used for regular day-to-day use. I have not lost ONE BYTE. NOT ONE. I have had bad sectors appear on old software distribution disks that got relegated to the SPARE bin by virtue of software updates, etc. In particular, the disks that Microsoft use to distribute C and ASM on have a tendency to come up with bad sectors on format. Whenever a DD disk that was pushed to HD use came up with bad sectors on format, I would reformat at DD, and not try to use it HD. The Nashua's I used NEVER ONCE came up with bad sectors on format, and HAVE NOT EVER FAILED IN ANY WAY YET. About 6 months ago, I bit the bullet and bought 2 boxes of HD disks to use in my backups, but for everyday use I STILL do 1.44M formats on 720k disks. Not a byte lost in two years. It is clear to me that the quality of disks from manufacturer to manufacturer varies (as the MS and other distribution disks indicate), but if the disk manufacturer is known, and its NOT Acme Storm Door and Disk Company, then why not use them? Clearly, my purchase of some HD disks for backup recently indicates that my level of trust is not absolute. Your milage may vary..... Discaimer: I am not connected with Nashua, Central Point Software, Microsoft, IBM, or any other conflict-of-interest trap explicitly or implicitly mentioned in this posting. Nya-nya.
trd10523@uxa.cso.uiuc.edu (08/24/89)
I have a 1.44M 3.5" drive that never checks for the HD hole, so I've been using DD disks formatted at 1.44M for about 4 months now. Here's what I've found: 1. The better quality the disk, the better luck you'll have doing this. I tried using bulk 3.5" disks (about 70 cents a piece) and only 6 of 10 would format 1.44M without bad sectors. I'm now using 3M or Sony disks which yield 90% or better success (Sony being the best). The Sony disks cost me 1.30 a piece, but that's still a far cry from 2.80 for a HD. 2. Compatibility with other drives is limited. A friend of mine has a PS/2 which also does not check for the HD hole. He's using Sony disks too, but when I tried to diskcopy some of his disks, about 10% got read errors from tracks 76 to 79. I have had no problems yet reading my own disks made on my own drive. I still use DD disks at 720K for hard disk backups, though, just to be safe.
las) (08/25/89)
In article <1989Aug23.151302.2622@Octopus.COM> pete@octopus.UUCP (Pete Holzmann) writes: }In article <5221@mtuxo.att.com> gjp1@mtuxo.UUCP (XMRK4-G.PATEMAN) writes: }}(George E. Lehmann) writes: }}}DON'T USE 3.5" DS/DD DISKETTES FOR HD (1.44MB) USAGE!!!!! [Meta-discussion on debate regarding use of 720K, 3.5" disks as 1.44M disks.] }[Somebody ought to get a box of each of several }brands of 720K disks, do an archival backup on each box at 1.44M, then report }on the readability of the disks over the next year (or more!)...] What? Test a net proposition? Are you crazy? :-) regards, Larry P.S. If I had a 1.44M I would do this. If nobody else tries it, I will when I get my new PC but that probably won't be until the beginning of the year. -- Signed: Larry A. Shurr (cbema!las@att.ATT.COM or att!cbema!las) Clever signature, Wonderful wit, Outdo the others, Be a big hit! - Burma Shave (With apologies to the real thing. The above represents my views only.) (Please note my mailing address. Mail sent directly to cbnews doesn't make it.)
davidr@hplsla.HP.COM (David M. Reed) (08/25/89)
I provide support for a lot of DOS users, and this is one area I frequently can not help them. If they format a low density disc (360K) as a high density disc (1.2M), they will typically get about 600K of bad sectors and the rest is okay. HOWEVER, the problem that I get called about is when they can no longer get their data back off the disc. It seems that many of the sectors DOS had considered good when it formatted the disc are really somewhat questionable, and experience deterioration over time. (Most of the users who have had this problem used the wrong disc type by accident when they formatted their disc, and generally I have found that Norton Utilities will not recover most of the data that ends up in these marginal sectors.) I am now getting experience of the same kind of problem from those who format a disc certified for 720K as 1.4M, creeping deterioration of many sectors of the disc. So my advise is: If your data is valuable, and your time is valuable, and the manufacturer will not certify the disc for the higher density (when they stand to make more money for almost no more investment or resources), then why take chances? I wont.
uchuck@ecsvax.UUCP (Charles Bennett) (08/26/89)
I can speak from experience. We have tested, not by choice, both Sony
and Verbatim brand 720Kb, 3.5" diskettes. And have found data
degradation to the point of complete loss in approximately 2 months.
Erattic reading behavior occurs in less than 5 weeks. Machines used
where IBM PS/2 Mdl's 50 and 60.
YOU PAY FOR WHAT YOU GET - if you want to store 1.44Mb use the HD
diskettes; if not, 720Kb's are ok. BUT *ALWAYS* format them correctly.
--
Chuck Bennett e-mail (any): uchuck@unc.bitnet
University of NC uchuck@ecsvax.uncecs.edu
Medical CAI uchuck@med.unc.edu
Phone: 919-966-1134 uchuck@uncvx1.acs.unc.edujohne@hpvcfs1.HP.COM (John Eaton) (08/28/89)
<<<< < Actually it is the higher current causing self-erasure on the inner tracks < because of the lower coercivity of the DD media. Because the HD has higher < coercivity we can resolve a "tighter pattern", but require the higher current < to write. It's not the retentivity that makes the real difference but that we < have to pound on it much harder to make it accept the signal in the first < place. ---------- I suspect that 3.5 inch drives use the media sense hole to determine what density the media can support. Although IBM chooses to ignore this data, the drive probably doesn't and could use this information to set write current. This would add another dimension to the media fracas because it would make a difference in whether you formated on a PS/2 or used a hole punch. On a PS/2 you would write high density with low write current while using a hole punch gives you high density using high write current. John Eaton !hpvcfs1!johne
chuck@ncrcce.StPaul.NCR.COM (Chuck Rissmeyer) (08/29/89)
When I got my Toshiba ND356 1.44 3.5" drive, it didn't come with the proper instructions. I installed it the best I could. I found that Mr. Egghead's DSDD disks would format at 720k just fine. I also purchased a box of Kodak HD 3.5 for $16.00. I couldn't pass up that offer; the box said HD. Well I opened the Kodaks, but the disks themselves said DSDD. Since I got them at a good price anyway, I thought that I should just be careful the next time. But wait, theses DSDD disks wouldn't format at 720K (My AT BIOS at the time did not support 3.5" drives by the way). I brought my Kodaks to a friend whose drive did not look for the hole, and he formatted them at 1.44 without problem. I updated my BIOS to one that supported 3.5" floppies, and changed my drive jumpers to look for the 1.44 meg hole. Why wouldn't my old BIOS and driver software format the Kodak DSDDs at 720k but would format the Egghead DSDDs at 720K? The jackets were the same (no strategically placed holes). The answer has to be media density. I think the Kodak disk were HD disk in DD jackets, and my drive sensed the media. Since this episode, I have not tried to reproduce the original scenerio, but I found that even the Kodak DSDD disks can format at 1.44Meg without problem. Also, my new bios and drive configuration allow me to format the Kodaks at 720k or 1.44 meg (provided I punch the hole in the jacket for 1.44). Needless to say, I buy nothing but Kodaks, and none of their disks have formatted with bad sectors at 1.44 meg. All standard disclaimers apply, and I am in no way affiliated with Kodak or Mr. Egghead. chuck@stpaul.NCR.COM
andyross@ddsw1.MCS.COM (Andrew Rossmann) (09/01/89)
At work we have a CompuAdd 386/20 with a Seagate 4144R drive and an Adaptec controller (2332B?? 1:1??) This disk has a chronic problem of deterioration. It seems as if every Monday, several sectors go bad. A run through with Nortons Disk Doctor (3-4 times) cleans things up for a week or so. This computer was put together and formatted elsewhere, and sometime I'm going to break down and re-format it from scratch. Right now I just make FREQUENT backups. BTW, is the 4144R VERY senstive to movement? The computer is currently on a table that is not the most sturdy. In fact, severe head seeking can cause a noticable vibration in the table. Andrew Rossmann andyross@ddsw1.MCS.COM
karl@ddsw1.MCS.COM (Karl Denninger) (09/02/89)
>----- >Response 8 of 8 (5123) by andyross at ddsw1.MCS.COM on Thu 31 Aug 89 20:23 >[Andrew Rossmann] >(15 lines) > > At work we have a CompuAdd 386/20 with a Seagate 4144R drive and an >Adaptec controller (2332B?? 1:1??) This disk has a chronic problem of >deterioration. It seems as if every Monday, several sectors go bad. A run >through with Nortons Disk Doctor (3-4 times) cleans things up for a week or >so. Try a different controller. The ACB boards are junk IMHO. I would get a WD1006-SR2 first, and if that doesn't help, get the drive changed out (if under warranty). > BTW, is the 4144R VERY senstive to movement? The computer is currently on >a table that is not the most sturdy. In fact, severe head seeking can cause >a noticable vibration in the table. Not in my experience. We have run these drives on the test bench for several hours on a seek test without a "rigid" mounting, and had no problems. -- Karl Denninger (karl@ddsw1.MCS.COM, <well-connected>!ddsw1!karl) Public Access Data Line: [+1 312 566-8911], Voice: [+1 312 566-8910] Macro Computer Solutions, Inc. "Quality Solutions at a Fair Price"