news@blackbird.afit.af.mil (News System Account) (08/09/89)
I was wondering, is the diff between the slower (i.e. Atari Megafile 30) hard drives and the quicker (28 ms) drives noticeable in day-to-day usage. I understand I can get a Megafile 30 for $499 and am debating about holding out for a faster (& more expensive) drive. On another note, how about some reports on these 'Frog' drives :-} with the removable media. Fast? Reliable? Standard? Compatible (with Atari- 'soon'-to-be-released-drives)? ------------------------------------------------------------------------------- Bill Hodges | Me? People who speak for the Air Force get bhodges@blackbird.afit.af.mil | paid a lot more than I do! I just work here. -------------------------------------------------------------------------------
dbsuther@PacBell.COM (Daniel B. Suthers) (08/09/89)
In article <1275@blackbird.afit.af.mil> bhodges@blackbird.afit.af.mil writes: >I was wondering, is the diff between the slower (i.e. Atari Megafile 30) >hard drives and the quicker (28 ms) drives noticeable in day-to-day >usage. I understand I can get a Megafile 30 for $499 and am debating >about holding out for a faster (& more expensive) drive. > >On another note, how about some reports on these 'Frog' drives :-} with >the removable media. Fast? Reliable? Standard? Compatible (with Atari- >'soon'-to-be-released-drives)? The faster drives are noticeably faster. If you are useing file based based utilities such as arc, a compiler, MTC (a multi tasking shell), or Minix you will notice a VAST improvement ( about 4 times faster ). Many graphics programs also write to the disk often. The FROG drive appears to be based on the Syquest drives. I have used their 10 meg disks, and find them to be usable, but pretty fragile. You can lose a disk by dropping it 2 feet onto a desk. The ones I used were also VERY susceptible to power hits. A 1 second outage was guaranteed to cause data corruption. A suprising feature was that the drives also are sensitive about their orientation. They don't work well on their sides. Keep in mind that it is a hard disk, and the heads can crash just like those with fixed media. If you make the mistake of putting a crashed disk back into a drive you might end up with a bad head which will trash any other $100 disks you insert afterwards. Don't get me wrong though. As soon as I can get a hands on demo I plan to consider the purchase my-self. It would be great for setting up a development disk, a MAC disk, and a dos disk without fear of accidently corrupting each other. I am aware of the risks, and can be careful enough to reduce them to a reasonable level. Who knows, they may have added auto parking, power tolerance, enhanced quality control and all the other things a personal computer user needs :^). Dan Suthers uucp: ... ihnp4!pacbell!pbeos!dbsuther
logic@wet.UUCP (Henry Kwan) (08/11/89)
In article <29194@pbhya.PacBell.COM> dbsuther@PacBell.COM (Daniel B. Suthers) writes: > >The faster drives are noticeably faster. If you are useing file based >based utilities such as arc, a compiler, MTC (a multi tasking shell), or >Minix you will notice a VAST improvement ( about 4 times faster ). Many >graphics programs also write to the disk often. > I would have to say that a drive's transfer rate make more of a difference than the average access time. Unless you are doing intensive database work which requires a lot of seeks or you have a badly fragged hard disk, the transfer rate will make more of a difference. Compare a ST506/412's 5Mbit/sec to something like an Imprimis 94181-702M's 15-24Mbit/sec (ain't sure on this one but it's somewhere in that neighborhood). The Imprimis will just blow your socks off. > >The FROG drive appears to be based on the Syquest drives. I have used their >10 meg disks, and find them to be usable, but pretty fragile. You can lose >a disk by dropping it 2 feet onto a desk. The ones I used were also VERY >susceptible to power hits. A 1 second outage was guaranteed to cause data >corruption. A suprising feature was that the drives also are sensitive >about their orientation. They don't work well on their sides. > I have not used the SyQuest 10MB cartridge drives but I have used the 44MB cartridge drives and find them to be quite reliable. Sure, you wouldn't want to risk a cartridge by dropping it two feet but then again, would you do the same to your fixed hard disk? The SyQuest 44MB drives are auto-parking so power failures have little effect. They also work well on their sides. Look at PLI's 44MB drive. They are suppose to work on their sides! > > [stuff deleted] > >Dan Suthers >uucp: ... ihnp4!pacbell!pbeos!dbsuther I am putting together a 44MB drive for a friend this weekend. Will report on how it runs on the ST. Should be fairly decent in any event. -- Henry Kwan - FWB, Inc. | "Experience varies directly claris!wet!logic@ames.arc.nasa.gov | with equipment ruined." cca.ucsf.edu!wet!logic@cgl.ucsf.edu | {claris,ucsfcca,hoptoad,lamc}!wet!logic | -- Tech Support
alderaan@tubopal.UUCP (Thomas Cervera) (08/12/89)
In article <399@wet.UUCP> logic@wet.UUCP (Henry Kwan) writes: >In article <29194@pbhya.PacBell.COM> dbsuther@PacBell.COM (Daniel B. Suthers) writes: > >[...] Compare a ST506/412's >5Mbit/sec to something like an Imprimis 94181-702M's 15-24Mbit/sec (ain't >sure on this one but it's somewhere in that neighborhood). The Imprimis >will just blow your socks off. But doesn't the DMA bus limit such a speed if you solder it to an ST ? -- Thomas Cervera | UUCP: alderaan@tubopal.UUCP SysMan RKOpdp (RSTS/E) | ...!unido!tub!opal!alderaan (Europe) D-1000 Berlin 30 | ...!pyramid!tub!opal!alderaan (World) Motzstrasze 14 | BITNET: alderaan%tubopal@DB0TUI11.BITNET (saves $$$)
logic@wet.UUCP (Henry Kwan) (08/13/89)
In article <676@opal.tubopal.UUCP> alderaan@tubopal.UUCP (Thomas Cervera) writes: >In article <399@wet.UUCP> logic@wet.UUCP (Henry Kwan) writes: > >>[...] Compare a ST506/412's >>5Mbit/sec to something like an Imprimis 94181-702M's 15-24Mbit/sec (ain't >>sure on this one but it's somewhere in that neighborhood). The Imprimis >>will just blow your socks off. > > But doesn't the DMA bus limit such a speed if you solder it to an ST ? > True. Then again, I'm of the school of thinking that overkill is better than underkill. The Imprimis drive will get something like 1,050K/sec on ICD's RATEHD while the ST506/412 drive will get something like 300K/sec. Real-life benchmarks are probably less glaring in difference but should still be quite noticable. I'm curious about the BMS-200 host adapter though. They claim that it supports a 1.6MB/sec transfer rate but I always thought that the ST's DMA bus limit was 1.33MB/sec. Or are they measuring it at the SCSI level? -- Henry Kwan - FWB, Inc. | "Experience varies directly claris!wet!logic@ames.arc.nasa.gov | with equipment ruined." cca.ucsf.edu!wet!logic@cgl.ucsf.edu | {claris,ucsfcca,hoptoad,lamc}!wet!logic | -- Tech Support
bms@bdt.UUCP (Vance Chin) (08/16/89)
In article <411@wet.UUCP> logic@wet.UUCP (Henry Kwan) writes: > >I'm curious about the BMS-200 host adapter though. They claim that it >supports a 1.6MB/sec transfer rate but I always thought that the ST's DMA >bus limit was 1.33MB/sec. Or are they measuring it at the SCSI level? > The Atari DMA bus will handshake at 2 Megabytes/second, this assumes zero turnaround time between ack's and drq's and the DMA bus. The problem is that the DMA chip tends to modify the data at this rate ( only a slight problem :-) ). The next step down is 1.6MB/sec due to the clock rates. Vance Chin