delph@tesla.UUCP (02/08/84)
Date: Tue, 7 Feb 84 17:25:46 est
From: delph (Robert Del Favero)
Message-Id: <8402072225.AA12898@tesla.UUCP>
To: micro@net
Subject: Macintosh: Setting the record straight
We have been developing software for some time now, being one
of the "100's" of software developers that Apple describes in their releases.
We have developed software for most popular personal computers and have a great
deal of experience with these products from both the user's and the developer's
points of view.
There has been a great deal of discussion about the Macintosh in this
newsgroup recently. We are concerned that a great deal of misinformation has
been presented regarding this new machine. Additionally, we take exception to
some of the opinions ventured to date.
Most of the arguments we have heard articulated against the Mac indicate
to us that some people are missing the import of this machine. The Macintosh
is intended to be an appliance! That's it. Not a mainframe, not a super-mini,
not the answer to every hacker's dreams.
To quote the Apple developer's document, _Inside Macintosh_:
Macintosh is intended to be the first mass-market personal computer.
It is designed to appeal to an audience of non-programmers, including
people who have traditionally feared and distrusted computers. To
achieve this goal, Macintosh must be friendly. The system must,
once and for all, dispel any notion that computers are difficult to
use. Two key ingredients combine in making a system easy to use:
familiarity and consistency.
In our opinion, Macintosh realizes these design goals exceptionally well.
It has by far the most consistent, well-organized user interface we have
developed software for.
For those who want a development machine, the Mac is utterly inappropriate.
However, you would be amazed at what an insignificant percentage
of the personal computing market consists of those that don't really want
(or need) to hack on the machine. While most people will need to program
somewhat, the packages that Apple and third party developers will deliver
in the next nine months, will suffice. The plain fact of the matter is:
most people out there are truly "users". They want to purchase high-quality,
well-designed, familiar and consistent software. That's what (and who) the
Macintosh was built for.
Some interesting points have been raised in the discussions so far:
1. What cost slots? Indeed, most of the usual slot applications have been
built into the machine already. Furthermore, two RS-422A ports, each with
tremendously high speeds and bandwidth, provide plenty of flexibility for
expansion. The absence of slots will increase the cost of additional
devices, but most users won't have any such devices.
According to Steve Jobs, "add a $5 bill [for a microprocessor] to the
peripheral, and then talk serially, rather than have every single user
pay an extra few hundred dollars for the price of the slots that may
never get used." Since most peripherals cost in the $100's (at least)
it seems that Apple has quite properly addressed the issue, and saved
most customers--the ones who have no need for the extra devices--a bundle.
2. Lack of control key? Are you kidding? It is quite precisely the point of
the Mac interface to eliminate the control key. However, the funky
"cloverleaf" (called the "command") key performs a similar function for
expert users. In fact, there are several "short-cuts" permitted in the
user interface. Double-clicking replaces the menu selection for "Open"
and command key combinations are provided to short-cut frequently pulled
menu items, for example. Apparently, not even Apple can remain completely
true to the faith.
3. Not a "serious machine"? Hardly; the Mac has more raw horsepower than any
microcomputer on the market today. It will zip along with the best of them,
both in terms of "learning curve" time and processing time. It has the
finest graphics facilities on the market, and exploits them better than any
other machine exploits its own. Try to get honest-to-god overlapping
windows on the PC!
The applications available for Mac will be of higher quality than
those for any other machine, simply because the potential is there for the
software developers to make it happen. Apple made it far easier to
use the standard interface and integration facilities than to ignore them.
You can roll your own interface, but it would be far less efficient.
While Apple intends the Mac to be a "toaster", it is a heavy duty toaster.
If a Differential Equation solver is needed, it will run on this baby,
faster and easier than on any other stock PC. And it will be easier to
write in that ease of use. Don't be put off by the notion of the computer
as "friend". This is a real machine, it's just easier to use than a VAX.
[Yeah, trademarks all over the place...]
4. "Less predictable than the PC"? This is definitely a matter of taste, but
we can't help but wonder if you've ever used a PC OR a Mac.
5. Of Mice and Menus...
a). There is no textual command language in the Macintosh Finder at all,
contrary to !tesla!kinch (309@tesla). There are some textual commands
available, but in all cases they duplicate Mouse actions.
While it is possible to develop applications in the baroque style, it is
much harder to do so than to follow Apple's lead.
b). The point-and-press dialogues can be quite rich. They don't "go back
to first grade mentality." It is possible to specify a wide range of
information using the Toolbox Dialogue Manager, when more than a simple
noun(s) and verb(s) interaction is needed (q.v. BYTE, p. 37).
c). As for the problems of abandoning home row on the keyboard, we too
were skeptical at first. We can only suggest that the current skeptics
try producing a document of some size using this technology. It takes
a short while to understand the tradeoffs involved, but you'll like the
mouse a lot once you get used to it.
d). To my knowledge, there seems to be no "objectionable penalty in
delayed response" involved in the Mac user interface. The only noticeable
delays seem to be due to disk accesses. [All the user interface stuff
is in ROM pipelined to the processor. It actually runs faster than code
in the RAM!]
6. The dot-matrix printer running in hi-resolution graphics mode is faster
than an Epson MX-80, and the print quality is substantially better. It
is not a laser printer, but one's on the way (for "cheap", and already
fine-tuned for the Mac). We find it quite suitable for our business and
personal correspondence.
7. Compatibility? It certainly isn't an IBM PC, and vive la difference!
[On a Mac, you could show the accents!] True, we can't easily plug
a spreadsheet from another machine into the Mac. This can be a
legitimate problem for some institutions, but not for very many
households (Jerry Pournelle's may be the only exception).
On the other hand, the computing concepts are readily "upward compatible"
to the Mac/Lisa, where they will run better and faster, and be easier to
use. LAN's will admit transfer of information in sites with many diverse
machines. [And baby, was this machine built to be networked!]
Frankly, if you only have one machine, compatibility only provides the
advantage of a software base. Let's be realistic about this. No software
developer willingly ignores a market. There will be enough Macs on the
market that they will be forced to produce Mac software to remain
competitive.
The thing about compatibility is that a software developer has to aim for
the least common denominator machine of a given class. It would be
foolish to market an MS-DOS product that wouldn't work on nearly every
MS-DOS machine out there. It may be that Macintosh's so-called lack
of expandability will benefit the Mac's owners, since all software will
be configured and tuned for his exact configuration, and not castrated to
run on the look-alikes. (There's no difference between a high-end Mac
and a low-end Mac.)
8. Repairs? Repairs will be swift--they'll give you a new one. We have
every confidence in the repair technician at our local Apple dealership.
Apple repair techs are trained and certified by the parent company. Third
party entrepreneurs can apply for the repair program, but the Apple
requirements are severe: Strict minimum stocking facilities; must repair
anybody's machine; strict pricing on everything except labor. Apple's
reputation for quality service is substantial. Maybe the horror stories
you're familiar with had to do with IBM's PC track record?
Note: He who does not buy extended warranty gets what he doesn't pay for!
(So true of any machine!)
Concluding (at last):
We's hate to do development on the Mac, and we're the masochists that wrote
Wizardry on an Apple ][ Pascal system with no hard disk. But the machine
is a winner for the *users* who make up most of the consumer market.
Responding to the techies who are disappointed, we can only quote the
late, great Frank N. Furter: "I didn't make him for you!!"
On the other hand, go buy a Lisa 2 (you'll want a hard disk and the extra
half-meg of memory to be really happy, but it cooks as is). You can hack
*her* to pieces if you want.
Andy Greenberg
Robert Del Favero Jr.
Andrew Greenberg Associates, Inc.
and
Sir-Tech Software
Available for comment at: !decvax!cornell!tesla!delphguy@rlgvax.UUCP (Guy Harris) (02/09/84)
> Most of the arguments we have heard articulated against the Mac indicate > to us that some people are missing the import of this machine. The Macintosh > is intended to be an appliance! That's it. Not a mainframe, not a super-mini, > not the answer to every hacker's dreams. I agree 100%; and it turns out that a lot of the resistance to these "new-style" systems seems to be based on familiarity with the "old" command-language interfaces and unfamiliarity with the new desktop interfaces. I find I can get things done very quickly on our Lisa (modulo the speed of the underlying machine) once you get adjusted to it, and the commonality of the user interface ideas (pointing, dragging, selecting, etc.) makes it easier to figure out how to do something new without even looking at the manual (often it takes a lot of time just to figure out where to look in the manual). > 2. Lack of control key? Are you kidding? It is quite precisely the point of > the Mac interface to eliminate the control key. However, the funky > "cloverleaf" (called the "command") key performs a similar function for > expert users. In fact, there are several "short-cuts" permitted in the > user interface. Double-clicking replaces the menu selection for "Open" > and command key combinations are provided to short-cut frequently pulled > menu items, for example. Apparently, not even Apple can remain completely > true to the faith. Nor should they remain "true to the faith." Such shortcuts *are* necessary to make the machine usable by people who have learned how to do a number of common operations. After a while, it *does* become a pain to go to the menu to open things; the double click to open a file becomes almost second nature and you can open and close objects quite quickly, almost automatically. > While Apple intends the Mac to be a "toaster", it is a heavy duty toaster. > If a Differential Equation solver is needed, it will run on this baby, > faster and easier than on any other stock PC. I don't know if I'd go *that* far; without a 68881, I doubt a Mac could solve a differential equation faster than an 8088-based PC with an 8087 (although without the 8087, the 8088 would fall behind the 68K). b). The point-and-press dialogues can be quite rich. They don't "go back to first grade mentality." Amen. (Besides, if you open/close/move/delete objects by "point-and-select", you make fewer typographical errors than doing so with commands, simply because you don't type as much. There is still the possibility for error, but I *know* I have to correct typos in commands more often than I have to correct mis-selections.) c). As for the problems of abandoning home row on the keyboard, we too were skeptical at first. We can only suggest that the current skeptics try producing a document of some size using this technology. It takes a short while to understand the tradeoffs involved, but you'll like the mouse a lot once you get used to it. Besides, a lot of cursor pads require you to take your fingers off the home row anyway; I find that even working with our function-key-heavy screen editor, moving off the home row doesn't slow me down much. The thing about compatibility is that a software developer has to aim for the least common denominator machine of a given class. It would be foolish to market an MS-DOS product that wouldn't work on nearly every MS-DOS machine out there. It may be that Macintosh's so-called lack of expandability will benefit the Mac's owners, since all software will be configured and tuned for his exact configuration, and not castrated to run on the look-alikes. (There's no difference between a high-end Mac and a low-end Mac.) YES. However, one could develop a system with most, if not all, the quality of the Mac on a "more conventional operating system" (MS-DOS, UNIX, etc.), although some things may be slower or more awkward. NBI uses 4.2BSD as the OS for their System One Integrated Work Station, for example. However, one has to beware that just because a package runs under UNIX doesn't mean it will run on every UNIX box in creation; there's a *lot* of software that runs on a Sun workstation that won't run on an IBM PC/XT, for instance. Just because you are using UNIX as an OS for a machine or a base for an application doesn't mean you *have* to target it to the "generic UNIX" market, although it would be foolish to write a spreadsheet solely for the Sun, given that the market for such a beast is rather small. But if you built a system of the class of the Lisa which runs UNIX, it would mean you could pick up third-party software for it even if it isn't the best sort of software you can get for the hardware. Presumably, though, you'd supply a lot of the software yourself or have it written specifically for that class of machine; the trick is you wouldn't have to have it *all* written specifically for your machine (although it's interesting to note that Interleaf, Inc.'s all-singing-all- dancing multi-font what-you-see-is-what-you-get word processing software, which currently runs on the Sun, is being ported to the Cadmus as well, so even "new generation" software can be made portable among "new generation" hardware which runs UNIX). Guy Harris {seismo,ihnp4,allegra}!rlgvax!guy
POURNE%mit-mc@sri-unix.UUCP (02/15/84)
From: Jerry E. Pournelle <POURNE@mit-mc> Your enthusiasm for the Mac is perhaps understandable, but to say that it has more raw horsepower than any other micro is perhaps an exaggeration. Did you mean bang for the buck? Becaus certanily I can put together a micro with more power. We understand there will shortly be a Modula-2 compiler for Mac, which is particularly interesting to us here.