488 lines
24 KiB
Plaintext
488 lines
24 KiB
Plaintext
|
Newsgroups: comp.sys.amiga.reviews
|
|||
|
Path: menudo.uh.edu!usenet
|
|||
|
From: koren@fc.hp.com (Steve Koren)
|
|||
|
Subject: REVIEW: Commodore Amiga 4000
|
|||
|
Message-ID: <1992Oct26.173622.22620@menudo.uh.edu>
|
|||
|
Followup-To: comp.sys.amiga.hardware
|
|||
|
Keywords: Amiga, computer, hot topic, commercial
|
|||
|
Sender: amiga-reviews@math.uh.edu (comp.sys.amiga.reviews moderator)
|
|||
|
Nntp-Posting-Host: karazm.math.uh.edu
|
|||
|
Reply-To: koren@fc.hp.com (Steve Koren)
|
|||
|
Organization: The Amiga Online Review Column - ed. Daniel Barrett
|
|||
|
Date: Mon, 26 Oct 1992 17:36:22 GMT
|
|||
|
|
|||
|
|
|||
|
PRODUCT NAME:
|
|||
|
|
|||
|
Amiga 4000
|
|||
|
|
|||
|
|
|||
|
BRIEF DESCRIPTION:
|
|||
|
|
|||
|
This is a review of the Amiga 4000, the latest machine in the Amiga line
|
|||
|
of personal computers from Commodore.
|
|||
|
|
|||
|
The machine as reviewed is:
|
|||
|
|
|||
|
Amiga 4000
|
|||
|
Commodore 1960 Multisync Monitor
|
|||
|
6 Mb RAM
|
|||
|
68040 CPU/25 MHz
|
|||
|
120 Mb HD
|
|||
|
1.76 Mb floppy drive
|
|||
|
"AGA" chipset
|
|||
|
|
|||
|
This particular machine was apparently one of the first 200 produced.
|
|||
|
|
|||
|
|
|||
|
LIST PRICE:
|
|||
|
|
|||
|
Check with your dealer. The original MSLP is US$3699, but the street
|
|||
|
price seems to be quite a bit cheaper. Prices certainly vary
|
|||
|
geographically as well.
|
|||
|
|
|||
|
|
|||
|
COMPANY INFORMATION:
|
|||
|
|
|||
|
Commodore Business Machines, Inc.
|
|||
|
1200 Wilson Drive
|
|||
|
West Chester, PA 19380 USA
|
|||
|
|
|||
|
(The machine is produced in England, and the keyboard and mouse are
|
|||
|
produced in Malaysia).
|
|||
|
|
|||
|
|
|||
|
OBTAINING THE MACHINE:
|
|||
|
|
|||
|
I had a very difficult time hunting down a place to buy a 4000. Four
|
|||
|
successive calls to the "Commodore Dealer Locator" got me phone numbers
|
|||
|
of supposed dealers, but in all cases the dealers either had gone out of
|
|||
|
business, or no longer sold Amigas when I called. This was a bit
|
|||
|
frustrating. After two weeks of searching, I eventually found a dealer
|
|||
|
about 75 miles away by talking to someone who had bought an Amiga there
|
|||
|
a while ago. The chore of finding the computer in the first place was
|
|||
|
one of the few bad things I have to say about this machine. I don't
|
|||
|
think most people would go through the trouble I did in order to buy the
|
|||
|
system. I believe it would be beneficial for Commodore to 1) vastly
|
|||
|
increase its dealer base in the US, and 2) keep its dealer database up
|
|||
|
to date, since calling 8 non-existent dealers does not give a very
|
|||
|
professional image of the company.
|
|||
|
|
|||
|
|
|||
|
HARDWARE:
|
|||
|
|
|||
|
The 4000 comes in a desktop style case, a bit smaller than an Amiga
|
|||
|
2000. The keyboard is essentially identical to the 2000's keyboard, but
|
|||
|
mouse is a more rounded "beetle" style mouse, instead of the more
|
|||
|
angular 2000 mouse. The 4000 has a key and lock which can be used to
|
|||
|
shut off all keyboard and mouse input to the machine (including the
|
|||
|
C-A-A reboot combination, but not including the power switch). The
|
|||
|
power switch is on the front, along with LEDs for power and the internal
|
|||
|
HD.
|
|||
|
|
|||
|
|
|||
|
UNPACKING AND SETTING UP:
|
|||
|
|
|||
|
This task went very quickly and painlessly. The system as shipped is
|
|||
|
essentially ready to plug in and go - the operating system is already
|
|||
|
installed on the hard drive, and the hard drive is configured for
|
|||
|
booting. There was just one small glitch on my machine - on some early
|
|||
|
4000s, the hard drive was formatted in the OFS ("Old FileSystem")
|
|||
|
format, which is substantially slower than the newer FFS ("Fast
|
|||
|
FileSystem"). From what I hear, Commodore has since corrected this
|
|||
|
problem. It was not much trouble for me to reformat the hard drive and
|
|||
|
reinstall the operating system. Although this isn't a recommended
|
|||
|
approach, I got through it with no trouble without reading the
|
|||
|
documentation, just by booting the install disk and clicking on things.
|
|||
|
The OS install utility is quite user friendly and intuitive, and you can
|
|||
|
pick what parts of the operating system you do and do not wish to
|
|||
|
install.
|
|||
|
|
|||
|
One thing I noticed immediately is that the 4000 is a quiet machine. My
|
|||
|
old 2000 is fairly loud, and the 4000 seems to be only about half as
|
|||
|
loud when running. The hard drive is essentially silent, and only the
|
|||
|
fan can be heard, but it is quieter than the 2000's fan.
|
|||
|
|
|||
|
|
|||
|
INITIAL SYSTEM CONFIGURATION:
|
|||
|
|
|||
|
The operating system originally boots in 640x200 mode, similar to a 2000
|
|||
|
or 500. However, the AGA ("Advanced Graphic Architecture") chipset in
|
|||
|
the 4000 supports many other higher resolution modes. There are monitor
|
|||
|
configuration files that control the resolution and scan rate of the
|
|||
|
various graphic modes supported by the 4000. The Workbench screen can
|
|||
|
be run on any of these and changed by a tool in the preferences drawer.
|
|||
|
After some amount of fiddling, I settled upon the "SUPER72 Super High
|
|||
|
Res Interlace" mode. On my system, this mode gives a solid display of
|
|||
|
896x628 pixels (which I'll round to 900x630 for simplicity, although it
|
|||
|
is a 4x2 pixels short of that in reality). The scan rate in this mode
|
|||
|
is 25 KHz, which is enough faster than the 15 KHz interlace modes in the
|
|||
|
2000 that it seems to eliminate flicker. However, this might depend a
|
|||
|
little on lighting conditions. When I booted the system in this mode at
|
|||
|
the dealer, I could detect a bit of interlace flicker, but when I tried
|
|||
|
this mode at home, the display appears quite solid. With my anti-glare
|
|||
|
screen on the monitor, I cannot detect any flicker in this mode at all,
|
|||
|
unless I look very closely for it. It certainly seems to be a genuinely
|
|||
|
usable mode, quite unlike 640x400 on non-flicker-fixed A2000's. In
|
|||
|
order to display it, I had to adjust the vertical size knob on the
|
|||
|
monitor.
|
|||
|
|
|||
|
|
|||
|
WORKBENCH 3.0
|
|||
|
|
|||
|
The Amiga 4000 includes a new release of the Amiga operating system.
|
|||
|
Release 3.0 includes support for the AGA chipset of the 4000. The AGA
|
|||
|
chipset can support up to 256 directly accessible colors in any
|
|||
|
resolution mode from a 24 bit palette, and up to 252,208 simultaneous
|
|||
|
colors in "HAM8" mode. (HAM8 mode is excellenct for graphics
|
|||
|
applications, but isn't suitable for word processing or textual
|
|||
|
applications).
|
|||
|
|
|||
|
The Workbench 3.0 screen can be configured to any depth from 1 to 8
|
|||
|
planes. Depending on your resolution mode and tolerance to update
|
|||
|
rates, you may find that anywhere from 4 to 8 planes provides a suitably
|
|||
|
fast environment. In my 900x630 workbench (actually a 1024x768 virtual
|
|||
|
workbench displayed in a 900x630 physical display), I find the update
|
|||
|
rate adequate at 5 or 6 planes (32 or 64 colors). Seven and 8 plane
|
|||
|
displays can get slow at this high resolution, but they do better at
|
|||
|
lower resolutions such as 640x400. In fact, when I was playing with
|
|||
|
this system at the store, I compared the interactive performance of the
|
|||
|
4000 to a nearby 386/33 machine running windows 3.1. Both machines were
|
|||
|
running 8 plane displays at an identical resolution, and the Amiga was
|
|||
|
quite a bit faster than the 386 for window updates. Although I didn't
|
|||
|
time either one, here is my subjective impression of the speed of the
|
|||
|
4000 user interface compared to several other systems I have used a
|
|||
|
reasonable amount. The rating factor is "snappiness", whatever that
|
|||
|
means. Remember, this is subjective, and compares things like moving
|
|||
|
windows, scrolling scroll lists (which depends less on resolution), the
|
|||
|
speed with which windows pop up, etc. So graphics performance isn't
|
|||
|
directly correlated with this, and "tricks" of the OS, such as AmigaDos
|
|||
|
3.0's method of only scrolling needed bitplanes for CLI windows, can
|
|||
|
affect things:
|
|||
|
|
|||
|
System & UI approx system cost "snappiness" of UI
|
|||
|
-------------------------------------------------------------------
|
|||
|
Amiga 3000, 1 plane WB, 640x400 $1K-3K 2.0
|
|||
|
Amiga 3000, 2 plane WB, 640x400 $1K-3K 1.0
|
|||
|
Amiga 3000, 3 plane WB, 640x400 $1K-3K 0.7
|
|||
|
Amiga 4000, 4 plane WB, 640x400 $3K-4K 2.5
|
|||
|
Amiga 4000, 4 plane WB, 900x630 $3K-4K 1.5
|
|||
|
Amiga 4000, 5 plane WB, 900x630 $3K-4K 0.8
|
|||
|
Amiga 4000, 8 plane WB, 900x630 $3K-4K 0.6
|
|||
|
80386/33 clone, Windows 3.1, 800x600x8 $1K-2K 0.3
|
|||
|
HP 720 workstation, 8 plane, 1280x1024 $8K-12K 3.0
|
|||
|
|
|||
|
Workbench 3.0 supports the use of IFF images as backgrounds for both
|
|||
|
the workbench screen, and workbench windows. I currently have a
|
|||
|
640x400 image of a bicycle in the background of my workbench (which is
|
|||
|
a backdrop window), and have a smaller 320x200 image in the background
|
|||
|
of my workbench windows. Although this doesn't provide any real
|
|||
|
"functionality", it does look very nice and provides an easy way to
|
|||
|
visually distinguish windows. The effect is quite pleasing. All
|
|||
|
workbench windows share a common image, but since the edges of the
|
|||
|
windows will tend to use many different colors, it is easy to see
|
|||
|
where one window stops and another starts.
|
|||
|
|
|||
|
Workbench running on a 900x630 screen looks very nice. The extra
|
|||
|
resolution allows the use of higher resolution fonts on screen, which
|
|||
|
makes for a much more professional look. I am using a 20 point
|
|||
|
Compu-Graphic Times font for my window titles, which is a readable size
|
|||
|
on this display, a 15 point font as the system default, and a 15 point
|
|||
|
proportional font for icons. The 15 point font takes about as much
|
|||
|
screen real-estate as a 9 point font did on the 2000's 640x400 interlace
|
|||
|
screen, but provides much more resolution for nice looking characters.
|
|||
|
In addition, the display is much sharper and easier to read.
|
|||
|
|
|||
|
The display quality of the 4000's output is very high. Although my old
|
|||
|
2000's display was inferior, I feel the 4000's output, when sent to a
|
|||
|
suitably good monitor, is truly of workstation quality. I use a
|
|||
|
1280x1024 Sony display attached to an HP workstation every day at my
|
|||
|
job, and while it gives more resolution than the 4000 does, I don't
|
|||
|
think the 4000 lacks anything in sharpness or clarity in comparison. In
|
|||
|
fact, since most people with 4000s will probably use a 14" monitor,
|
|||
|
900x630 is about as much resolution as is practical at this size. In
|
|||
|
order to move to 1024x768, I believe that at _least_ a 16" monitor is
|
|||
|
needed. Since there is a minimum physical size of text that is
|
|||
|
comfortable to read, having more resolution on a small monitor meets
|
|||
|
diminishing returns after a point. Larger monitors than 14" are quite
|
|||
|
expensive.
|
|||
|
|
|||
|
At any rate, the 4000's display in 900x630 mode is, in a word,
|
|||
|
beautiful. The only potential problem is that some people need a 10% or
|
|||
|
so faster refresh in order to no be bothered by the interlace, but I
|
|||
|
don't think this will be a problem for most people under most lighting
|
|||
|
conditions.
|
|||
|
|
|||
|
AmigaDos 3.0 and the AGA chipset support HAM8 in any resolution mode.
|
|||
|
This means that it is possible to have a 900x630 display in up to 252000
|
|||
|
simultaneous colors from a 24 bit pallete, for near 24-bit quality
|
|||
|
graphics. Further, it is possible to animate HAM8 graphics in any
|
|||
|
resolution mode (although the practical limit is probably 640x400 due to
|
|||
|
bandwidth restrictions in this generation of graphics chips). As far as
|
|||
|
I know, there are no animation players yet which support HAM8 animation,
|
|||
|
but that should change fairly soon. The few HAM8 still-frame images I
|
|||
|
have seen look wonderful.
|
|||
|
|
|||
|
|
|||
|
HARDWARE EXPANSION:
|
|||
|
|
|||
|
The 4000 has 4 card slots, one 5.25" drive bay, and two 3.5" drive bays.
|
|||
|
The 3.5" bays can accept either two 1" tall devices each (for a total of
|
|||
|
four 3.5" devices), or one larger device. Commodore ships the machine
|
|||
|
with 1.5" tall devices, so you will need to replace one or both of them if
|
|||
|
you wish to install 2 devices in each bay. Further expansion will need an
|
|||
|
external case and power supply (which costs US$40-$85). The 4000 can use
|
|||
|
Amiga 3000 Zorro-III cards, which gives it a good supply of expansion
|
|||
|
devices already on the market, such as 64 Mb RAM cards. It can also use
|
|||
|
Amiga 2000 Zorro-II cards, although these will not take advantage of the
|
|||
|
4000's superior bus speed. The 4000's CPU is on a daughterboard and can be
|
|||
|
upgraded when faster versions come out. There have been rumors of future
|
|||
|
CPU boards with an on-board DSP. It is not yet known whether the 4000 is
|
|||
|
upgradable to the next generation of AGA chips. I hope the answer is
|
|||
|
"yes".
|
|||
|
|
|||
|
|
|||
|
SOFTWARE COMPATIBILITY:
|
|||
|
|
|||
|
Most properly written applications seem to work fine under AmigaDos 3.0.
|
|||
|
Immediately after powering up my 4000, I transferred my bare minimum
|
|||
|
"working set" of software, which consists of the following:
|
|||
|
|
|||
|
SKsh 2.1 beta
|
|||
|
GNU Emacs 18.58 (port by David Gay)
|
|||
|
ISpell 3.1ljr (port by Loren Rittle)
|
|||
|
Half a dozen commodities
|
|||
|
|
|||
|
All of the above installed and worked without any trouble at all, and in
|
|||
|
fact, I am currently using GNU emacs and ISpell to type this review.
|
|||
|
Emacs is talking to ISpell through ARexx without any trouble at all, in
|
|||
|
exactly the same manner as on my 2000.
|
|||
|
|
|||
|
I have only tried a few software packages so far other than the above.
|
|||
|
These I have found to work:
|
|||
|
|
|||
|
DPaint IV
|
|||
|
Excellence!
|
|||
|
A-10 Tank Killer 1.5
|
|||
|
Most 2.0 freeware Commodities or utilities
|
|||
|
|
|||
|
I have tried a few PD "screen hacks" and such which failed on the 4000:
|
|||
|
|
|||
|
Oing! (a screen hack with bouncing ball sprites)
|
|||
|
World (puts a 3d rotating globe in a workspace)
|
|||
|
|
|||
|
On the other hand, a few other screen hacks ("Wavebench") seem to work.
|
|||
|
It seems a fair assessment that compatibility for properly written
|
|||
|
software is very high, but games or utilities which break the rules
|
|||
|
stand a good chance of not running under AmigaDos 3.0.
|
|||
|
|
|||
|
A number of popular AmigaDos software suppliers have already announced
|
|||
|
AGA versions of their products, and a few are already on the market.
|
|||
|
|
|||
|
Since 2.0, software which opens custom screens should allow the user to
|
|||
|
choose the monitor file for the screen resolution to be used. Such
|
|||
|
software, even if written under 2.0, will run with the higher resolution
|
|||
|
modes under 3.0 with no changes. However, some software which is not
|
|||
|
that smart will still use old modes. If several screens each have
|
|||
|
different scan rates, the multisync monitor will have to re-sync when
|
|||
|
the user changes screens on the Amiga, which adds a slight delay of
|
|||
|
about 0.5 sec. If a screen of one resolution is dragged partially down
|
|||
|
over a screen of another, the front-most screen sets the scan rate of
|
|||
|
the output.
|
|||
|
|
|||
|
|
|||
|
OS 3.0 FEATURES & BUGS:
|
|||
|
|
|||
|
AmigaDos 3.0 supports object classes. I haven't had a chance to play
|
|||
|
with these too much yet, but I can describe the basic concept. If, for
|
|||
|
example, a desk top publishing program wants to load in an image for
|
|||
|
inclusion in a document, it previously had to understand the format of
|
|||
|
the image. I.e., it had to call the IFF shared library to read in IFF
|
|||
|
images, or include code to read jpeg or GIF images if it wanted to read
|
|||
|
those formats. If a new format came along, the program had to be
|
|||
|
re-shipped. With AmigaDos 3.0 file classes, all that changes. A
|
|||
|
program can read an image class, and AmigaDos will call the appropriate
|
|||
|
handler to extract information from the image itself without the program
|
|||
|
having to even know what type of image it is. Thus, if a new image type
|
|||
|
called "zpeg" comes along, all that needs to be done is install a new
|
|||
|
object class for zpeg images, and all the old software will be able to
|
|||
|
suddenly understand the new image type. The same applies for sounds,
|
|||
|
text, animations, and any other type of object. This is a powerful new
|
|||
|
feature in AmigaDos 3.0 that has not been developed much yet, but has
|
|||
|
great potential.
|
|||
|
|
|||
|
CLI windows in AmigaDos 3.0 seem to be smarter than they were in 2.0.
|
|||
|
The console device, apparently, only scrolls the bitplanes that
|
|||
|
absolutely need to be scrolled. This means that if you are just
|
|||
|
displaying text in the standard color, the console device may only have
|
|||
|
to scroll one or two bitplanes instead of the 5 or 6 there may be in
|
|||
|
your screen. This makes using CLI windows fast even on deep workbench
|
|||
|
screens. (Disclaimer: I don't know for sure that this is what is
|
|||
|
happening but I suspect it quite strongly.)
|
|||
|
|
|||
|
There are some bugs in AmigaDos 3.0 yet. The "multiview" object viewer
|
|||
|
crashes easily, and occasionally the palette preference tool does odd and
|
|||
|
unexpected things to your palette. But overall, I have not yet found any
|
|||
|
critical bugs which would prevent me from using the system. Most of the
|
|||
|
ones I have found are just minor inconveniences which I'm sure will be
|
|||
|
fixed for future versions of 3.0.
|
|||
|
|
|||
|
|
|||
|
BENCHMARKS:
|
|||
|
|
|||
|
The following benchmarks compare the Amiga 4000 to:
|
|||
|
|
|||
|
- An Amiga 500, 68000/ 8 MHz, no fast RAM
|
|||
|
- An Amiga 2000, 68000/ 8 MHz, fast RAM
|
|||
|
- An Amiga 2500, 68020/20 MHz, fast RAM
|
|||
|
- An Amiga 3000, 68030/25 MHz, fast RAM
|
|||
|
|
|||
|
The tests were all performed with AIBB_4.65, ("Amiga Intuition Based
|
|||
|
Benchmarks", by LaMonte Koop). In all cases, the 3000/25 is used as a
|
|||
|
comparison base:
|
|||
|
|
|||
|
Machine: 500/00 2000/00 2500/020 3000/030 4000/040
|
|||
|
----------------------------------------------------------------------
|
|||
|
Integer tests:
|
|||
|
WritePixel 0.25 0.26 0.68 1.00 2.81
|
|||
|
Sieve 0.11 0.11 0.56 1.00 1.11
|
|||
|
Dhrystone 0.18 0.18 0.48 1.00 3.43
|
|||
|
Sort 0.13 0.14 0.45 1.00 2.68
|
|||
|
Matrix 0.10 0.11 0.52 1.00 1.52
|
|||
|
IMath 0.05 0.05 0.51 1.00 2.29
|
|||
|
Memtest 0.16 0.17 0.61 1.00 1.20
|
|||
|
TGTest 0.50 0.52 0.82 1.00 1.44
|
|||
|
InstTest 0.17 0.17 0.44 1.00 1.78
|
|||
|
Float/Double:
|
|||
|
Savage 0.01 0.01 0.51 1.00 1.19
|
|||
|
FMath 0.05 0.05 0.40 1.00 4.72
|
|||
|
FMatrix 0.14 0.14 0.46 1.00 1.04
|
|||
|
Beachball 0.01 0.03 0.39 1.00 6.50 (!!)
|
|||
|
SWhetstone 0.02 0.03 0.38 1.00 0.56
|
|||
|
DWhetstone 0.02 0.02 0.37 1.00 3.40
|
|||
|
FTrace 0.01 0.01 0.42 1.00 3.10
|
|||
|
CplxTest 0.04 0.04 0.47 1.00 3.25
|
|||
|
|
|||
|
Generally, it can be seen that the 4000 averages about 2 to 2.5 times
|
|||
|
faster than the 3000 for integer operations, and about 3 to 3.5 times
|
|||
|
faster than the 3000 for floating point operations. For the Beachball
|
|||
|
test (which ray traces a beachball on the screen), the 4000 is a
|
|||
|
staggering 650 times as fast as an unexpanded Amiga 500, and about 215
|
|||
|
times faster than an Amiga 2000 with fast ram. Most of the CPU bound
|
|||
|
tests of the 4000 come out about 5 to 10% slower than my PP&S 68040 card
|
|||
|
on my 2000, which runs at 28 MHz instead of 25. However, the 4000 feels
|
|||
|
snappier in actual operation due to the AGA chipset. CPU board upgrades
|
|||
|
to 33 and 40 MHz 68040s promise even more speed from the machine. The
|
|||
|
CPU performance of the 68040, coupled with the AGA chipset's enhanced
|
|||
|
color modes, make this essentially the perfect 3D rendering platform for
|
|||
|
those who can't afford a Silicon Graphics workstation.
|
|||
|
|
|||
|
The following is a performance test of the 4000's internal disk drive
|
|||
|
using DiskSpeed 4.1:
|
|||
|
|
|||
|
CPU: 68040 OS Version: 39.106 Normal Video DMA
|
|||
|
Device: sys: Buffers: 128
|
|||
|
Comments: Amiga 4000 internal disk
|
|||
|
|
|||
|
CPU Speed Rating: 3097
|
|||
|
|
|||
|
Testing with a 512 byte, MEMF_FAST, LONG-aligned buffer.
|
|||
|
Create file: 26184 bytes/sec | CPU Available: 84%
|
|||
|
Write to file: 26462 bytes/sec | CPU Available: 85%
|
|||
|
Read from file: 158488 bytes/sec | CPU Available: 50%
|
|||
|
|
|||
|
Testing with a 4096 byte, MEMF_FAST, LONG-aligned buffer.
|
|||
|
Create file: 157322 bytes/sec | CPU Available: 78%
|
|||
|
Write to file: 163083 bytes/sec | CPU Available: 79%
|
|||
|
Read from file: 213989 bytes/sec | CPU Available: 74%
|
|||
|
|
|||
|
Testing with a 32768 byte, MEMF_FAST, LONG-aligned buffer.
|
|||
|
Create file: 329347 bytes/sec | CPU Available: 72%
|
|||
|
Write to file: 375143 bytes/sec | CPU Available: 71%
|
|||
|
Read from file: 559055 bytes/sec | CPU Available: 54%
|
|||
|
|
|||
|
Testing with a 262144 byte, MEMF_FAST, LONG-aligned buffer.
|
|||
|
Create file: 429040 bytes/sec | CPU Available: 69%
|
|||
|
Write to file: 550858 bytes/sec | CPU Available: 64%
|
|||
|
Read from file: 909545 bytes/sec | CPU Available: 33%
|
|||
|
|
|||
|
Average CPU Available: 68% | CPU Availability index: 2106
|
|||
|
|
|||
|
Those needing more disk speed than this can get it when fast Zorro-III
|
|||
|
SCSI cards become available. (Actually, Zorro-II cards can be used now
|
|||
|
at some cost in speed). The speed of the internal IDE drive is
|
|||
|
acceptable, although the CPU utilization gets a bit high during high
|
|||
|
speed transfers. However, I doubt many people will notice this, and
|
|||
|
SCSI is always available for those who need the high end.
|
|||
|
|
|||
|
|
|||
|
WARRANTY:
|
|||
|
|
|||
|
The machine and monitor come with a "Commodore Gold Service" warranty. If
|
|||
|
the machine breaks in a period of one year, they will pick it up for free,
|
|||
|
fix it, and send it back also for free, by overnight express mail.
|
|||
|
On-site service is available for a small, annual fee ($49 or $79).
|
|||
|
|
|||
|
|
|||
|
PROBLEMS:
|
|||
|
|
|||
|
Aside from my initial difficulty in finding an Amiga dealer within a 3
|
|||
|
hour drive of my house, I have had few problems with the 4000. Although
|
|||
|
I have yet to try most of my old software, most of what I have tried has
|
|||
|
worked. The only exception is that some games and a few "screen hacks"
|
|||
|
have failed, but I expected that, and it isn't the fault of the 3.0
|
|||
|
operating system, but rather the fault of the games themselves.
|
|||
|
|
|||
|
The 4000 could really use more than 2 Mb of chip ram. 4 Mb would be
|
|||
|
appropriate.
|
|||
|
|
|||
|
There is really just one significant problem I have run into. The 4000
|
|||
|
has a feature called "mode promotion", which does two things. First, it
|
|||
|
attempts to force application screens that would have opened in 15 KHz
|
|||
|
interlace mode to open at a higher scan rate to avoid flicker. Second,
|
|||
|
it attempts to force application screens with a resolution of 200
|
|||
|
vertically to "scan-double" their output and eliminate visible scan
|
|||
|
lines. This effect is very pleasing - all those old 640x200 screens
|
|||
|
suddenly are a lot more pleasant to look at. However, on my 4000, scan
|
|||
|
promotion seems to force screens far to the right of the monitor, such
|
|||
|
that there is no way to see the whole screen. Neither fidding with my
|
|||
|
monitor or the overscan preferences was able to help. I'm not sure what
|
|||
|
the problem is, but for now I've kept scan conversion off so that the
|
|||
|
application screens are reasonably centered and visible.
|
|||
|
|
|||
|
|
|||
|
CONCLUSIONS:
|
|||
|
|
|||
|
The A4000 presents a significant expansion in the capabilities of Amiga
|
|||
|
computers. The original Amiga's graphics, while fantastic by the
|
|||
|
standards of 1985 when they were introduced, have recently begun to show
|
|||
|
their age. The AGA chipset gives the Amiga a true 24 bit palette and
|
|||
|
the ability to use hundreds of thousands of colors in any resolution
|
|||
|
mode. The potential improvements of future Amigas now lie primarily in
|
|||
|
graphics speed, and to a lesser extent, increased resolution.
|
|||
|
|
|||
|
I bought my first Amiga 1000 in 1985, with 256 Kb of memory and later
|
|||
|
upgraded to an accelerated 2000. When I became interested in the Amiga
|
|||
|
4000, I was at first unsure whether the abilities it provided were
|
|||
|
significant enough to warrant upgrading from my current system. Now
|
|||
|
that I have worked with the 4000, I am confident the answer is "yes".
|
|||
|
The 4000 takes a large step towards making the Amiga into a workstation
|
|||
|
class computer system.
|
|||
|
|
|||
|
|
|||
|
COMMENTS:
|
|||
|
|
|||
|
I can be reached electronically at:
|
|||
|
koren@fc.hp.com
|
|||
|
|
|||
|
or via phone:
|
|||
|
303-226-4985 (USA)
|
|||
|
|
|||
|
|
|||
|
---
|
|||
|
|
|||
|
Daniel Barrett, Moderator, comp.sys.amiga.reviews
|
|||
|
Send reviews to: amiga-reviews-submissions@math.uh.edu
|
|||
|
Request information: amiga-reviews-requests@math.uh.edu
|
|||
|
General discussion: amiga-reviews@math.uh.edu
|
|||
|
|