hsgj@batcomputer.UUCP (02/20/87)
I received some mail indicating that I should post my review of the
MicroBotics MAS-20 hard disk. So here it is.
Disclaimer: I have no connection, other then possessing a MAS-20,
with MicroBotics company. This is the only hard disk I have ever
personally possessed, so I can't compare it to other Amiga hard disks.
Maybe someone else would like to do that. The format of this review
is (loosely) based on those found in InfoWorld magazine.
A Review of the MicroBotics MAS-20 Hard Disk for the Amiga 1000.
*** Features: ***
The MicroBotics MAS-20 is a 20 Megabyte hard disk for the Amiga
Computer. The MAS-20 consists of a 14.5 x 7 x 3 inch metal box, in the
same color as the Amiga system unit, along with a power cable, a cable
to the Amiga, and a floppy disk. The MAS-20 also comes with a "manual"
which consists solely of one (1) one-sided photo-copied piece of paper,
which is the installation guide.
This drive is well constructed and has a solid look and feel.
All plugs, ports, switches, and lights are well mounted. No exposed
wires or other "unprofessional" marks are evident. The drive features
an "in use" light on its front, a pass-thru printer port (see below),
and an SCSI interface so that one may chain other SCSI devices off the
back of the MAS-20. While operating, the disk drive is not completely
silent, but is by no means loud. The volume is about that of a room
sized window fan going at medium speed.
The MAS-20 connects to the Amiga in an interesting way; through
the Amiga's parallel (printer) port. Even though it is connected in
such a dubious fashion, MicroBotics includes software to make the use
of the drive transparent to the user. Once the drive is mounted, it
can be accessed as DH0:, just like one of the built-in floppies. To
mount the drive, one must include a special command in the Startup
Sequence of your Workbench floppy.
*** Performance: Satisfactory Minus ***
One buys a hard drive for two reasons. Primarily, the drive is
for mass storage; 20 Megabytes can hold as much information as 23 standard
880K floppy disks. Such a large amount of storage confers savings of
time and space. Time, because 23 disks worth of software can be
accessed at once, and space, because you no longer need a disk rack to
store floppies. The second reason for purchasing a hard disk is for
speed. A hard disk can typically deliver data much faster then a floppy
drive.
The MAS-20 fulfills the primary storage function. Coupled with
AmigaDOS's tree-directory structure, a pratically unlimited number of
files can be stored in directories and subdirectories on the hard disk.
Drive Partitioning (where one physical drive can be "split" into two or
more logical drives), while a feature of the MAS-20, is not needed, for
the Amiga operating system can efficiently deal with the great amount
of storage space on the hard disk.
The MAS-20 (so far) has been very reliable. It has been in
operation for exactly one month, and in that time span has lost no data.
Under normal operation, the device driver has never become confused or
balked at any AmigaDOS request.
What the MAS-20 fails to offer is speed. I have performed several
ballpark tests on what I would consider to be typical applications. All
tests were run from the CLI, with no other processes active other then
the Initial CLI. No "AddBuffers" or any other disk-speedup software was
used.
Speed Tests:
Three general tests were run on DF0: (a floppy drive), RAM: (the
Ram Disk (v1.2 version)), and DH0: (the MAS-20 hard disk). These tests
were: a directory of C:, which contained 54 files; the time to load and
run Kermit, which is 113,712 bytes long; and the time to copy the entire
C: directory (which is 431 blocks) to the RAM: disk. In the case of testing
RAM:, the C: directory was copied into a subdirectory on the RAM: disk.
All Test results are "back of the envelope figures". Tests were
timed in the following manner:
1> RUN DATE +
command +
DATE
This sequence of commands will print the date, run the command, and then
print the date. The finest resolution of timing, then, is one second. All
tests were performed twice on each drive, and the results in all cases
were identical times for each test. For Kermit timings, the Close Box on
Kermit was clicked immediately after the Kermit window opened.
Test Results:
Following are the three tests, and the timings (in seconds) for
each drive to complete the test. Lower times are better.
TEST RAM: MAS-20: DF0:
-------------------------- ---- ------- ----
DIR C: (54 files in dir) 02 05 09
-------------------------- ---- ---- ----
RUN KERMIT (113712 bytes) 03 11 14
-------------------------- ---- ---- ----
COPY C:#? TO RAM: 09 23 51
-------------------------- ---- ---- ----
As you can see, the RAM: disk is incredibly fast. However, the hard
drive is not that much faster then the floppy, except when one is
reading many small files (eg the copy). I would estimate that in a
mixed bag of user actions, the hard drive would provide a 50% boost
in speed.
Note that since the MAS-20 can read a lot of small files very
fast, it performs very well in showing the WorkBench icon files. I do
not have benchmarks on this, having erased the great majority of my
.info files. However, I would imagine that for WorkBench users, the
MAS-20 speed up would be about 100%.
CPU Speed tests:
Aside from measuring disk speed, one must also take into account
the power that the drive steals from the CPU. Standard floppy drives
use something called DMA, which allows them to transfer data from the
disk to the computer, without bothering the central processor. This
leaves the CPU free to run other programs in the multi-tasking system.
The MAS-20, however, relies on a software driver program to shuttle
data from the printer port into the computer memory. This can potentially
eat away a lot of CPU performance.
To test the effect of the driver on the CPU, I wrote a program
which was extremely CPU-dependent -- it will take the square root of PI and
print out this value, and it does this over a thousand times. The program
uses the AmigaDOS DateStamp() routine to accurately time itself. I then
created a 102,060 byte long text file. The test, then, was to run the
CPU dependent program (which takes about a minute and a half) and then
while it was running, copy the huge text file from disk and into RAM:
three times. To make a fair test, I created a NewCLI window wherein the
calculations program ran. In the standard window I wrote the following:
1> run copy file ram: +
delete ram:file +
copy file ram: +
delete ram:file +
copy file ram: +
delete ram:
However, I did not press RETURN until the calculations program had reached
iteration 100 out of 1000. The data I found is reported below:
Test: Time (minutes:seconds:ticks)
---------------------------------- ----------------------------
Calculations running by themselves 1:38:09
---------------------------------- ----------------------------
Calculations plus copy from df0: 1:47:26
---------------------------------- ----------------------------
Calculations plus copy from MAS-20 1:58:42
---------------------------------- ----------------------------
Note: There are 50 "ticks" in a second.
As you can see, the MAS-20 consumed twice as much CPU power as did the
normal floppy drive. For people who run a lot of CLI's or programs at
one time, and if they use the disk a lot in each program, this CPU
lossage can be a big burden.
Performance Conclusion:
The MAS-20, then, earns very good or excellent marks in storage.
However, the poor speed performance drags its overall performance score
to a satisfactory, and the effect on CPU power pulls the score down to
a Satisfactory Minus.
*** Documentation: Poor to Unacceptable ***
The entire printed documentation supplied with the MAS-20 was
one sheet of paper. One column of this lone, poorly photocopied, page
was of a hardware installation diagram. The remaining column advertised
some of MicroBotic's products. The hardware installation figure was
adequate, being that the hard disk is so easy to connect to the Amiga.
MicroBotics also supplies a floppy disk. On this disk are the
installation "Execute" command files, some mount files that the system
will need, a readme file, and a demo copy of TxED from MicroSmiths (NOT
MicroBotics). The disk, then, is the only real "documentation". However,
it was very poorly done. To begin with, the disk has .info files only
for the executables. But the executables can only be run from the CLI,
as the CLI is the only command invoker that has the EXECUTE command!
On the other hand, the documentation files do not have any .info files
at all, and thus are completely invisible to the WorkBench user, who
might want to look at them with NotePad. Clearly, the WorkBench user
is getting no help at all from MicroBotics, and in most cases it will
be the WorkBench user who needs the most help. Do not purchase the
MAS-20 if you are going to require hand holding.
The on-disk documentation is very clear for getting set up. To
do this, all one must do is make a copy of the *original* WorkBench V1.2
disk, and then run MicroBotic's install Execute script, which will put
their DHMount command, plus a custom parallel.device, onto your disk.
I made the mistake of attempting to run their install script on my own
copy of workbench, where I had deleted many of the files in the C:
directory, such as IF and ELSE. This caused the EXECUTE script to
fail. It was not until, in frustration, I re-read the readme file
several times that I realized that I needed all of those standard files.
This was my mistake, but I think it could have been averted with
printed documentation. After I got a vanilla copy of the workbench, the
setup script ran perfectly.
MicroBotics should have put, either in writing or in the readme
file, what their execute script was doing. All it really does is put
their DHMount command in your s:Startup-Sequence file, and give you a
new parallel.device, and then runs FORMAT over the hard disk. I could
have done this myself, without relying on the execute script, if they
had clearly listed what commands one must type from the keyboard.
What really irks me about the lack of documentation is that
three new commands, plus the SCSI port, are never documented at all!
MicroBotics includes PARK, BDFNEIGHBOR, and DISKVERIFY commands. Park
appears to be obvious to me (it parks the disk heads, good to do before
transporting the drive). However, what about the others? Does
BDFNeighbor mean anything to you? Also, what does DiskVerify do? I am
afraid to even test it out, for fear that it might format the disk, and
then check it for bad sectors, or something like that. No where are any
of these explained.
Finally, SCSI users are out in the cold. There is an SCSI port
on the back of the hard disk, but no where is there any documentation on
how to hook up another device to it. I know nothing about SCSI, but if
it is anything like RS-232 then there are probably several "standards".
At the least, one should know what "standard" the MAS-20 is using.
In summary, the documentation (or lack thereof) supplied was
atrocious. Sure, you don't need anything to actually operate the drive.
But in a product costing almost $1000 you at least expect a users manual.
Commands are undocumented, SCSI is ignored, and WorkBench users are out
in the cold. I am being overly generous in giving the MAS-20 a poor
documentation rating.
*** Ease of Use: Very Good ***
The MAS-20 is easier to use then a normal floppy disk. All of
the AmigaDOS commands, without exception, operate on the hard drive
just as they would on a floppy disk. And with the MAS-20 you never have
to insert or eject floppy disks, so in this respect, the MAS-20 is easier
to use then a floppy drive. There are only a few constraints which
keep the MAS-20 from earning an excellent rating in this category. The
first is that, according to the manufacturer, the drive must always be
turned on before the Amiga is turned on. Several times I forgot this
fact and, after powering up the Amiga, I had to shut it down and then
power up the MAS-20, and then reboot the Amiga. A second problem is
that, again according to MicroBotics, if you want to use the printer
you must turn it on before turning on the Hard Drive, and then leave
the printer on for the whole computing session. This restriction arises
because, as you recall, the printer is attached to the hard disk, and
not directly to the Amiga as is usual. Often, I will turn on my printer
just for a quick print job, and then turn it off again. The MAS-20
hinders my style by not allowing this. Actually, I have tried powering
up the printer in mid-session, printing, and powering-down. No ill
results occur unless you accidentally make another print request with
the printer still turned off -- then the MAS-20 driver gets confused.
Aside from having to turn on the hard disk first, and to never turn the
printer off once it has been used, the MAS-20 places no restrictions on
the user. It earns a Very Good in this category.
*** Setup: Very Good ***
The hardware setup of the MAS-20 thankfully was quite easy. All
you have to do is plug in a (supplied) printer cable from the MAS-20 to
the Amiga printer port, plug the standard power cable (supplied) from the
hard disk to the wall socket, and (if you have a printer) plug the printer's
cable into the back of the MAS-20. This is as simple and painless an
operation as anyone could desire. The software setup, as mentioned in
the documentation section, was a bit of a problem. If you did exactly
as they mentioned -- copied the workbench disk, and ran their script,
everything worked flawlessly. Being a little different caused problems,
that would have been solved in a snap by some decent documentation.
The MAS-20 can only be used if one includes their DHMount command
in the Startup-Sequence. This means that one has a "boot disk" that must
be used to start up the hard drive. In practice, this is fine, because
the only time one uses the floppy drives are for the initial boot, or
for doing backups. Therefore, one need make only one "boot disk".
*** Servicability: ??? ***
The MAS-20 was rugged enough to survive a New York to Chicago
journey in the back seat of my car, without suffering any ill effects.
I never had a need to call up
MicroBotics, so I do not know what their phone support (if any) is
like, or whether they will take the drive back for servicing, or how
much such servicing would cost. Hopefully I will never have to know
this! Any prospective buyers who are interested in knowing MicroBotics
policy can call the company. They list their telephone number on the
serial number plate, which is attached adjacent to the power switch of
the MAS-20 drive. So visit your dealer, inspect the rear of the drive,
and get the phone number that way. One point in MicroBotic's favor is
that they shipped promptly and by UPS within a week of our placing
the order.
*** Miscellaneous ***
The MAS-20 lacks backup software. MicroBotics never claimed
that they had any backup software, so this is not a drawback to the
product. However, the product would be greatly enhanced if they had
included some backup software, that would save the contents of the
disk to floppies.
As an unrelated note, it has been mentioned that MicroSmiths
(NOT MicroBotics) provided a demo copy of TxEd on the disk that came
with the MAS-20. When I started up TxEd, the image forming the logo
of MicroSmiths company was completely garbled, no doubt because I have
a fast-memory board. The "readme" file for TxEd also claimed, without
any evidence, that half the Amiga community that used word processors
or editors was using TxEd. In my opinion, any company that blatently
lies and can't even display their own copyright logo, probably cannot
provide a worthwhile product. I enjoyed throwing TxEd into the trashcan.
*** Value: Satisfactory Minus ***
The MicroBotic MAS-20 hard drive is a reliable piece of hardware
which features an inexpensive connection to the Amiga, an expandable
SCSI port, and 20 megabytes of much-needed storage space. The drive
suffers from speed and an atrocious lack of documentation. I think
that these features and drawbacks offset themselves, leaving the
user's end impression of the MAS-20 as a so-so machine. We purchased
the MAS-20 for under $1000 -- however, prices may have changed since
then, so, if you are interested, it is best to investigate the drive's
cost at a local dealer. The MicroBotic MAS-20 rates a satisfactory
minus in overall value.
-- Dan Green
--
ARPA: hsgj%vax2.ccs.cornell.edu@cu-arpa.cs.cornell.edu
UUCP: ihnp4!cornell!batcomputer!hsgj BITNET: hsgj@cornellacmcmanis@sun.UUCP (02/20/87)
In article <221@batcomputer.tn.cornell.edu>, hsgj@batcomputer.tn.cornell.edu (Dan Green) writes: > As an unrelated note, it has been mentioned that MicroSmiths > (NOT MicroBotics) provided a demo copy of TxEd on the disk that came > with the MAS-20. When I started up TxEd, the image forming the logo > of MicroSmiths company was completely garbled, no doubt because I have > a fast-memory board. The "readme" file for TxEd also claimed, without > any evidence, that half the Amiga community that used word processors > or editors was using TxEd. In my opinion, any company that blatently > lies and can't even display their own copyright logo, probably cannot > provide a worthwhile product. I enjoyed throwing TxEd into the trashcan. > -- Dan Green Dan and others, the Demo copy of TxEd on the MicroBotics disk is old and out of date. At the time the TxEd documentation was written half the Amiga community (then about 20,000) *was* using it, that has of course changed. MicroSmiths and specifically Charlie Heath is committed to writing quality software at reasonable prices. Their two products TxEd and FastFonts both provide good capabilities for the machine. Now I know Charlie and like him, which makes me biased I suppose however I would hope one would reserve their opinion until seeing the product rather than the demo copy. -- --Chuck McManis uucp: {anywhere}!sun!cmcmanis BIX: cmcmanis ARPAnet: cmcmanis@sun.com These opinions are my own and no one elses, but you knew that didn't you.
fnf@mcdsun.UUCP (02/22/87)
In article <13742@sun.uucp> cmcmanis@sun.uucp (Chuck McManis) writes: >Dan and others, the Demo copy of TxEd on the MicroBotics disk is old >and out of date. > >Now I know Charlie and like him, which makes me biased I suppose however >I would hope one would reserve their opinion until seeing the >product rather than the demo copy. I agree with Chuck on this. MicroSmiths is definitely one of the more commited and concientious companies in the Amiga market. Demos can be a two-edged sword. Some people will see the demo and say "this looks really neat, might have a few problems, but let's get the released version". Others will say "gee, this thing is full of bugs, what a piece of trash, forget them, they're obviously incompetent...". -Fred -- =========================================================================== Fred Fish Motorola Computer Division, 3013 S 52nd St, Tempe, Az 85282 USA {seismo!noao!mcdsun,hplabs!well}!fnf (602) 438-5976 ===========================================================================
hsgj@batcomputer.UUCP (02/26/87)
In article <13742@sun.uucp> cmcmanis@sun.uucp (Chuck McManis) writes: >In article <221@batcomputer.tn.cornell.edu>, hsgj@batcomputer.tn.cornell.edu (Dan Green) writes: >> [I write some negative comments about TxEd demo ] >> -- Dan Green > >Dan and others, the Demo copy of TxEd on the MicroBotics disk is old >and out of date. [...] MicroSmiths and specifically Charlie Heath is >committed to writing quality software at reasonable prices. Their >two products TxEd and FastFonts both provide good capabilities for >the machine. >-- > Chuck McManis I received about twenty (20) mail files pertaining to the MAS-20 review. This would indicate that net readers have a high interest in hard disks, (though they may not be representative of the whole Amiga community). Anyways, I do not want to be a source of negative information, so I will clarify some statements I made earlier: 1) I had some complaints with the TxEd ->>> DEMO <<<- for some personal reasons. Each user should make THEIR OWN DECISION on any software they purchase -- especially since I was concentrating on writing about the MAS-20 and the TxEd remark just "slipped out". I stand by my opinions, but apologize to MicroSmiths for any bad pub I have given them. Remember, I saw just a demo, and an old one at that... Draw your own conclusions. 2) I recently found out that the MAS-20 I have is a dealer/developer copy, which has the "feature" of specifically providing NO documentation. Therefore, if you walk into a computer sales place as an "average joe", the MAS-20 you buy may (I have no concrete info on this) have a manual. If we take away the minuses from the lack of manual, that I cited, the MAS-20's "score (which, mind you, is quite subjective)" will be raised. Of course, buying a MAS-20 as a non-dealer/dev will up its price by a bit. 3) I had the use of a Sperry PC/IT with a 40 meg drive for one summer. Therefore, when I reviewed the MAS-20, the only hard drive I could compare it to in my mind was the Sperry's drive. Now, not to over-plug the competition, but the PC/IT had the fastest drive on the market at the time, so when I did get the Amiga hard disk, my mindset was to expect something on the order of this speed (eg something damn fast). As my stats showed, the MAS-20 did not show such a great speedup (although it is still twice as good as a floppy). However, as some folks at Commodore/Amiga have pointed out, the fault can most likely be attributed to the AmigaDOS, which is placing a very low bound on disk throughput. 4) Due to the AmigaDOS barrier, I suppose all hard disks will evidence performance akin to the MAS-20. Is this good? Make your own conclusion. I stand by my statement that the MAS-20 is reliable and has a lot of useful storage, but just aint fast. If you are looking for speed, spend the big bucks and get a ram expansion... 5) Perhaps Commdore will incorporate the mentioned upgrade to AmigaDOS in the A2000, so that users familiar with the IBM PC world will have disk speed on the Amiga comparible to what they are used to. ps: Thanks for all the mail -- it makes rn more interesting :-) -- Dan Green -- ARPA: hsgj%vax2.ccs.cornell.edu@cu-arpa.cs.cornell.edu UUCP: ihnp4!cornell!batcomputer!hsgj BITNET: hsgj@cornella
bj@well.UUCP (02/28/87)
.tn.cornell.edu> Sender: Reply-To: bj@well.UUCP (Jim Becker) Organization: Whole Earth 'Lectronic Link, Sausalito, CA Keywords: Hard Disk Review The harddisks that have been out there for the most part are not DMA, with intelligent controllers. I don't want to get into heavy DMA vs. non-DMA wars over this point. My point of this message is to state that use of the controller technology used in the A2000 and the Byte by Byte expansion architecture makes a system that is very fast and wonderful to use. I believe that this will be one of the factors that will greatly enhance the appeal of the Amiga. I sure hope so, that is.... Yes, I am somewhat biased in my opinions, being associated with Byte by Byte, but I think that this performance issue is a key one that you should evaluate if you are in the harddisk market. Enough said, please no flames -- my corns already hurt. -Jim Becker Terrapin Software