ramsiri@blake.acs.washington.edu (Enartloc Nhoj) (02/04/90)
I thought i'd pass on my experience with Supra v3.37 and GCR.
Upon getting GCR about 2 weeks ago, i reformatted my Quantum 80S
with Supra 3.37. According to many sources, including Dave Small
and Supra themselves, Supra will allow you to create partitions
greater than 16 MB (i have TOS 1.0) as long as the larger partition(s)
are not used for GEMDOS. I proceeded with the following setup
under the Supra format utility:
C:\ 9
D:\ 9
E:\ 6
F:\ 15
G:\ 4.5
H:\ 10
I:\ 30 (30.19 = remainder)
I planned on using C-F for GEMDOS and G-I for Mac.
(I know these partition sizes/order may look strange, but the
were set like that for a reason.. obviously)
I watched Supra format the 164,000+ sectors totalling about 84MB.
0 bad sectors.. all looked fine.
I loaded up a few ST utilities.. including Spectre. Ran the
SPectre program to set defaults etc. When i came to the "format"
menu .... i noticed some strange things: Not only did Spectre
see the above 7 partitions, but there were partitions 8 and 9
of 12MB and 6MB resepectively. Partition 7 (I:\) still showed
30MB.. as a matter of fact, all the partitions were recognized
by Spectre and formatting the last 5 (including the "suprises")
worked flawlessly. I must have gone back into GEM desktop to
"show info" on my drives a zillion times.. the GEMDOS drives
added up correctly. Figuring i was not risking damaging anything...
i proceeded to install Spectre .. load in the Mac System file(s)
etc.. loaded up a few programs.. and so on.
The Mac OS saw all 5 partitions with sizes of 4.5,10,30,12 and 6
in that order. I was curious about the last partitions even
existing.. so i wrote to them .. ran programs off them...
i wrote to all five partitions.. in each case, the sizes
updated correctly.. as i put stuff in and took stuff out.
I somewhat suspected the 30 MB partition was not for real..
even though both Spectre and the Mac OS "saw" 30 MB in there.
For two weeks, i have used GCR with only a few crashes.. not nearly
as many as I get with GEM/TOS, ironically. Anyway, this morning,
i tried running Multi-finder.. well.. there were instant problems
recognizing drives and so on.. i didn't know if it was a function
of cache or system allocated memory or what..since i didn't know
much about multi-finder yet... in any event, i had to turn multi
finder off and work from finder. Since i don't have that much
mac stuff.. i created a bunch of folders in the 30MB partition
and copied programs and data into it....
Wouldn't you know it.. i couldn't put anymore than 12 MB inthe
partition.. even though the OS said there were 17MB or so left..
Well.. of course, by this time i realized that the partitons
were identical in size to the ones i had BEFORE i reformatted them
for Spectre/GCR.
Seems that the Supra format utility didn't do what it said it would
do... thinking back.. i definitely recall Supra prompting me
at the time of formatting that the 30MB partition could NOT be
used with GEMDOS... i clicked on "OK" or whatever..to say..
"go ahead". Either i "think" i clicked on there and didn't..
maybe i am at fault.. however, how does one explain Spectre's
"seeing" the 30MB... passing that info to the MAC OS and the
MAC OS updating the partiton etc...
Anyway.. this is the 6th time i have reformatted my drive..
not a fun thing to do... the result is that i continue
to "throw away" lots of ST software.... I figure today.. when
I reformat my drive, I'll make a small C:\ for a few startup
programs.. another larger partition for NOTATOR.. the
rest of my drive will be mac.
Any hints about the Supra format...? Do i have a bad version?
I remember going through hell when i first got the Supra
format utility that didn't like my Quantum... even though
it recognized the controller.... Is this another Supra
surprise? or another case of USER IDIOCY.. which i admit
i am rather successful at....
Thanks for the insights
-kevin
ramsiri@blake.acs.washington.edu
BTW: i have written to David Small about this a couple of times,
but he's been too busy to respond... so anyone out there
who can tell me what i am doing wrong will be much
appreciated.dsmall@well.UUCP (David Small) (02/06/90)
[The basenote lists some bizarre problems encountered with a Quantum 80-s drive, which apparently fooled Spectre into thinking it was 105 mb long.] Kevin, I'm sorry I missed you via Email; I can't get to the WELL as often as I'd like, as we're putting the final touches onto 2.5 (very time intensive). I think you'll find that what happened is Supra's software left glop in the trailing ends of the partition sector. GEM ignored it. However, the Spectre utilities saw the glop, and somehow interpreted it as reasonable partition data. I really have no idea how it got there; perhaps a previous user had formatted the drive as a 100 meg unit, and gotten far enough to stamp out a partition sector? I'm puzzled. Anyway, I'm glad in a way that the problem is resolved, and you've found the top possible sector # on your drive. Another way to find this is just to run PARTCOPY with a long sector count, from the drive to itself, until it reports a failure. Naturally, there's always the possibility the Spectre software screwed up; however, I recall *many* times seeing trash in the "bottom" of the Supra partition table. I usually go zero it out by hand when I find it. I'll bet that's what happened to you. -- thanks, Dave / Gadgets
hcj@lzsc.ATT.COM (HC Johnson) (02/06/90)
In article <5586@blake.acs.washington.edu>, ramsiri@blake.acs.washington.edu (Enartloc Nhoj) writes: > I thought i'd pass on my experience with Supra v3.37 and GCR. > > Upon getting GCR about 2 weeks ago, i reformatted my Quantum 80S > with Supra 3.37. According to many sources, including Dave Small > and Supra themselves, Supra will allow you to create partitions > greater than 16 MB (i have TOS 1.0) as long as the larger partition(s) > are not used for GEMDOS. I proceeded with the following setup > under the Supra format utility: > > I watched Supra format the 164,000+ sectors totalling about 84MB. > 0 bad sectors.. all looked fine. > > When i came to the "format" > menu .... i noticed some strange things: Not only did Spectre > see the above 7 partitions, but there were partitions 8 and 9 > of 12MB and 6MB respectively. Partition 7 (I:\) still showed Here are some "gotchas". First, as far a SUPRA and Spectre are concerned, there are 12 partitions on a disk. All of the information is kept in sector 0. It sounds as if you have corruption in some of the areas. Use suputl.prg to set to 0 the partitions that done exist. I have noticed that some garbage values prevent suputl from doing this. Then use supedit.prg to (CAREFULLY) set the partitions to 0. Second. there are two kinds of disk drives. SCSI drives take care of defective areas automatically. All other drives require that the formatter handle the problem. Most (all?) SUPRA controllers contain an Adaptec 4000 (MFM) or 4070 (RLL) controller. Adaptec controllers EXPECT to be told the defect list for the disk. Adaptec then makes these sectors disappear; they are not numbered or accessable. ===== What follows is My Opinion, not SUPRA's ===== Supra's formatter, hdx, does not support this mode in the Adaptec. Instead, SUPRA (and ICD and ATARI) format all sectors, then try to map the bad ones out logically. In TOS, this is done by gathering them together in an invisible file. This does not help Spectre or MINIX at all! Of course, If you do not have the bad block list for your drive you have a problem. Then you really need to buy the Adaptec manual ($5.00). =================================================== This is why I wrote my own formatter (minihdx.prg, and partitioner (mkfs.prg). These are available if you can send me EMAIL. Howard C. Johnson ATT Bell Labs =====NEW address==== att!lzsc!hcj hcj@lzsc.att.com
ramsiri@blake.acs.washington.edu (Enartloc Nhoj) (02/08/90)
In article <1267@lzsc.ATT.COM> hcj@lzsc.ATT.COM (HC Johnson) writes: >In article <5586@blake.acs.washington.edu>, ramsiri@blake.acs.washington.edu (Enartloc Nhoj) writes: >> I thought i'd pass on my experience with Supra v3.37 and GCR. >> >> Upon getting GCR about 2 weeks ago, i reformatted my Quantum 80S > >Here are some "gotchas". > >First, as far a SUPRA and Spectre are concerned, there are 12 partitions >on a disk. All of the information is kept in sector 0. It sounds as >if you have corruption in some of the areas. >Use suputl.prg to set to 0 the partitions that done exist. >I have noticed that some garbage values prevent suputl from doing this. >Then use supedit.prg to (CAREFULLY) set the partitions to 0. > >Second. there are two kinds of disk drives. SCSI drives take care of >defective areas automatically. All other drives require that the formatter >handle the problem. Most (all?) SUPRA controllers contain an Adaptec >4000 (MFM) or 4070 (RLL) controller. >Adaptec controllers EXPECT to be told the defect list for the disk. >Adaptec then makes these sectors disappear; they are not numbered or accessable. >Howard C. Johnson >ATT Bell Labs >=====NEW address==== Thanks for responding. The denoument: Briefly, When i got GCR, i went into Supra to low-level format my Quantum 80S (84MB drive). Starting from C:\, i worked my way down like so: 9MB, 9MB, 6MB, 15MB, 4.5MB, 10MB, and when i did CALCULATE, the window showed 30.19 .. so i set my SEVENTH partition to 30.19.. did CALCULATE and got EXCESS is 0.00. Partition I:\ was 30.19, Supra gave me a *WARNING* that this partition could not be used for GEMDOS... i clicked on CONTINUE.. i formatted the drive.. all 164,000+ sectors with 0 BAD SECTORS (the Quantum has a built-in SCSI).. I then went into Spectre and proceeded to format the last 3 partitions (G:\ H:\ and I:\) to MAC HFS format.. it was in SPectre that i noticed two more partitions of 12MB and 6MB respectively. The Spectre format utility shows the type and size of each partition.. all the sizes looked accurate .. including the 30MB partition number 7. For experiment, i decided to format partitions G-I PLUS the "new" partitions at the end for HFS mac mode... i wanted to see if i could write to the new partitions. Spectre formatted them in seconds.. i then booted up into mac mode and opened up all 5 windows.. checked sizes and so on... they all looked correct... i then started loading a bunch of PD stuff and so on onto the outer last two partitions.. reading and writing to them was perfectly flawless.. i then wrote to the 30MB and all was well there too.... it wasn't until i tried to really load up the 30MB that i found that it couldn't take more than 12MB... i finally realized that these partitions were the sizes i had BEFORE getting GCR.... I went back into Supra and discovered what had really happened. I don't think this is a case of user idocy.. others may in fact think so...: I examined the partition sizes in the format window... everything checked out fine.. but then i noticed beyond the 30MB partition.. i had to scroll the windo there.. were the two partitions J:\12MB and K:\6MB ..! now.. this was obviously the problem... My question is: why when i originally set the partition sizes did those two still exist even after the 30.19 meg set the excess to 0.00? Seem like Supra should have realized the numbers added up to over 100MB for an 80MB controller.. Never was there a warning.. also.. becasue the windo only shows 4 partitions at a time.. those last ones were not visible as i set my "last" one to 30.19. I must not have seen the grayed scroll bar... nonetheless, this doesn't justify the formatting routine to continue with "apparent" success...! Not only that, Supra did NOT format the 30MB partition, instead, it defaulted to the original 12MB... and then formatted the ones that shouldn't have been there.... what was strange, of course, is that the formatting utility window showed 30MB still... even though we all know there was only 12MB in there..... NOt only that.. i went so far as to use these partitions under the MAC OS.. anyway... when i look back at it... the formatting should never have proceeded... also.. i think that when someone resets the partition sizes and begins at C:\ works the way down until EXCESS = 0.00 .. doesn't seem like there should be any partitions lurking under some scroll bar.... -kevin ramsiri@blake.acs.washington.edu BTW: it would have been a more interesting problem if what Howard thought were actually true.. at least my aggravation would have been worth something...