Slashdot Mirror
A Review of the 128KB Macintosh
bfwebster writes "The physicist John Wheeler famously quipped that 'Time is nature's way to keep everything from happening at once.' The web flattens time by making more of the past accessible. Here, then, is a reprint of BYTE's official review of the original 128KB Macintosh from the August 1984 issue. The article highlights the radical break with other PCs that the Mac represented, while at the same time giving the first real warning of Steve Jobs's least-productive tendency: pre-emptive and often arbitrary constraint of end-user options (e.g., no memory expansion on the 128KB or announced 512KB Macs, even though the 68000 processor had a lovely, flat 16MB address space, as opposed to Intel's 808x segmented hell)."
as opposed to Intel's 808x segmented hell
How hard is it to write a submission about a product without taking a cheap potshot at the competition? Was this really necessary?
If you own a Ford, does your car drive better if you talk shit about imports whenever you're not driving your car?
It's eriee how similar this statement is to the statements which we get every time Apple launches a new product even today... "a .wma icon was included with the iTunes app in Mac OS X Tiger" or a while back it was possible to unlock the "secret colour screen" on your iPod 3rd gen. (it made the screen turn blue.)
Also similarly, the author says he actually wouldn't like colour, and he's glad Apple left this feature out. (Remind anyone of Steve Job's current stance on the video iPod?)
My 3D Texturing Skinning work (under construction)
Byte was such a great magazine. It tried to cover a wide range of computer and technology related subjects. I really miss it.
"Leo Fender was in a 'state of grace' when he designed the Stratocaster." -- Paul Reed Smith
Slow news day it may be, but the introduction of the Mac *was* a historic event. The Byte article is a nice reminder of that.
I had a Mac 128 w/2 drives. The thing that made the Mac immortal wasn't necessarily the user interface, though the user interface was indeed revolutionary.
.
The thing that made the Mac immortal was the fact that anyone could "publish" documents from their desktop without needing complex typesetting systems or knowledge of traditional "publishing" and commercial printing processes.
At the time, most people with home computers didn't even have printers, which were expensive, error-prone, often massive, and didn't produce pretty output. All non-industrial printers at the time were either dot matrix or daisy wheel (using letter blocks like a typewriter to pound letters through a ribbon) impact printers and had only one typeface at one size. On dot matrix printers the quality of these letters was horrible (think NINE dots of vertical resolution per letter for consumer-grade printers or FIFTEEN dots of vertical resolution for business class printers). Very expensive printers might have a second "high quality" typeface that you could select by pressing a button on the printer, but this typically wasn't much better.
Basically, the process of creating a printed document with a computer had, until the Mac, been one of simply typing ASCII into a very basic editor program (Linux users: think pico or similar; Windows users, think Notepad), then sending it to the printer directly as a stream of characters, which it would output using its single available ugly, low-res typeface and size. No formatting, no fonts, no graphics, certainly (even the dot matrix printers generally didn't have any graphics capability whatsoever--it just wasn't included; only the ability to accept a stream of ASCII and dump it out to the page was in the ROM). What little formatting could be performed (left/right justification, line spacing, etc.) was often set in a word processor as a document property globally, and wouldn't be displayed on the screen as you typed.
The Macintosh and relatively cheap ImageWriter printer changed all this radically; you could format text using multiple typefaces, set them to a range of sizes, boldface, italicize, even full justify (!), and not only would these things appear on the screen as you did them (beyond magical in an era in which most PCs also only had the ability to display ASCII on their screens, lacking graphics capability unless you had expensive hardware like a so-called Hercules card, IIRC, still mono), but they could be output to the printer and would appear on the page just as they did on the screen. And you could even mix text and graphics
This kind of capability was unheard of because it had never before been available to the consumer at any price, and certainly not in a system that required no specialized knowledge to use.
You knew the Mac was an important computer historically from the moment it was released, because within a month or two, in any city or neighborhood, every newsletter, advertisement, flyer, poster, city council report, whatever that hadn't been commercially printed had obviously been done on a Mac. Everyone knew what a Mac was and knew that it was the computer that could be used to publish readable, visually pleasing, professional documents straight from your office or bedroom, for just a few thousand dollars.
STOP . AMERICA . NOW
Although the 128k had many a kludge with respect to memory management, multitasking, etc. I'd argue that Apple had the right approach when it came to telling developers what to expect. Direct interactions with hardware were frowned on. Apple's early design guidelines were very explicit about NOT assuming anything about the hardware, file system, display, etc. Developers that took this advice to heart could create applications that were future compatible.
The result is that I still use some applications on a near daily basis that were introduced in 1987/1988. These apps could run on a Mac Plus (System 6, 8 MHz 68000, 2 MB RAM, 800k floppies) and now run on a dual-G5 (OS 10.3, 1.8 GHz G5, 1 GB RAM, 160 GB SATA HD).
Apple may not have designed pre-emptive multitasking into their early systems, but they did create a development ethos that meant that early applications were not incompatible with the major changes in both hardware and OS that occured later.
Two wrongs don't make a right, but three lefts do.
The PC's success was driven largely by business usage. In 1984, you would have bought a PC because it had one of the best keyboards that has ever been made, and it had an outstanding monochrome text display (with a crisp font and specialized long-persistence phosphor).
Basically, the PC was a standalone version of IBM's high-quality mainframe terminals. It was designed for people who needed comfortably to run business apps all day long. This is not something that you would want to do on the primitive color monitors of the day, and the Macintosh was a brand new architecture with a radically different UI and zero business software available at its introduction. The PC also had IBM's support and brand name; as they say, you'd never be fired for choosing an IBM.
When the clones came along and offered PCs to the public at low prices, people bought a computer just like the ones at work that they were already familiar with. The rest is history.
Like it or not, part of technology is making it accessible to common person. Geeks may look down on Apple's products for not being ugly black things straight out of a 1985 stereo room, but that's because geeks don't know the first thing about marketing technology.
I think you're missing something.
Apple was selling them at $2500 and couldn't keep up with demand. What makes you think they'd be able to keep up with the *increased* demand if they lowered the price? Not only would they have more unhappy customers who couldn't buy their computer, but they'd be making less money on each computer they sold.
Comment of the year
First, hardware:
;-)
1) Memory access hatch (like the battery one it had) and the ability to upgrade memory...
2) Non-proprietary battery. Have you ever tried getting a replacement one? It's not easy. Wouldn't a 9 volt or something have sufficed?
3) Attach the mouse to the keyboard.
4) Sell a "ROM upgrade" service... Allow older machines to become "newer" machines for a reasonable fee.
You're OK so far.
Second, software:
1) FREE dev kits. Those Apple kits were really expensive if I recall correctly.
2) I think the filesystem (if designed by me, now) would probably be optimized for tools like spotlight.
3) The kernel would likely be an exo-kernel.
4) I would support TCP-IP
Holy Crap Batman! (Oh wait, that's me. Well, yes. As I was saying...)
1) There was really no concept of APIs at the time of the first Mac. Anything special that was loaded into memory was a form of BIOS call or memory jump. For the Mac, the primary calls were for Disk I/O and QuickDraw. That was about it. As a result, no dev system really existed for the Mac. You coded in assembler or you didn't code at all. Building something like a C environment would have been a tremendous expense for Apple, and was completely out of the question. Note that the devkits they did plan never materialized.
2) First of all, are you really willing to eat up a good chunk of your 400K floppies with meta-data and indexes? Secondly, what would you search for? Most people didn't bother with directories on floppies because the floppy *was* the directory. i.e. You had a floppy for Project A, a floppy for SpreadSheet B, and a floppy for Graphics C. For quick indexing and retrieval, you labelled them and placed them in a thumb-through disk holder.
3) KERNEL?! You do understand that there really wasn't a concept of a kernel back in those early consumer computers, right? The DOS (Disk Operating System) consisted of standardized software calls to control the floppy drive. After that, they got the hell out of the way and let the user do whatever he wanted. The Mac was slightly more sophisticated in that it also controlled a graphics device, but not by much. Even the mere mention of the word "kernel" at the time would have had you laughed out of your job and told to go program for Big Iron where they could afford to waste 300K of RAM on complex hardware control and multitasking.
4) TC... I... what? Have you been talking to those weirdos over at DARPA again? They've got some crazy idea about networking all computers together using the same protocol. Personally, I think they're nuts. I hear that the network stack uses up dozens of K of memory, and that the packets are each 512 bytes a piece. 512 bytes! You could fit an entire memo in that! Not to mention that expensive hardware they need to chain the computers together. Haven't these guys heard of a serial port or an acoustic coupler? I tell you, unless hardware gets a LOT cheaper in the next ten years, this ARPANET thing is going nowhere. Well, maybe academics will use it.
Javascript + Nintendo DSi = DSiCade
Yeah, the Ph.D. is a great degree and allows you to get all sorts of prestigous, high paying jobs, but for less time and money you could get a Bachelor of Liberal Arts degree *and* an Associates Degree in a totally unrelated field.
Quantum materiae materietur marmota monax si marmota monax materiam possit materiari?
http://developer.apple.com/ http://www.gnustep.org/ http://www.opengroupware.org/ for starters...but anyway, what you may have meant was Objective-C didn't catch on outside the NeXT community, which didn't really matter as NeXT were kindof most interested in its use inside the NeXT community.
And as you say, the original version - meaning there was a rev in which that was fixed (BTW you can fit a floppy drive on an original cube - I should know, I've got four). Actually the ultimate in cubey goodness was to install the OS on a MO disk and install a small hard drive for the swap directory - that way you still got to take your environment around with you in your pocket, but could use some nice fast swapspace. Or just buy dozens of the slabs and leave them all over the place.
Again, a problem with an early rev - the slabs were cheaper than that. The NeXT was superior in almost every respect to the equivalent from Sun, and the OS and development environments are still superior to many other offerings available today. Given a choice between developing a bespoke app for NeXTOpenCocoaYellowGNUBoxStep or Qt/.NET/Java/GTK+/wxWidgets, I know which side my bread's buttered.
As for my problem with the guy, the entire premise of his column is that he is an ignorant prat. Once he came on board, Byte started degenerating into yet another PC rag.
-jcr
The only title of honor that a tyrant can grant is "Enemy of the State."
"the Macintosh design team crammed an unbelievable amount of power into the 64K bytes of ROM in the form of tightly written, highly optimized machine code. In doing so, the team provided standard user interfaces, so that most application programs on the Mac will be used in similar forms." ... does it ?
While certainly not just applying to the Mac's of yore. What happened to those days where the true art of bare metal programming was the pinnacle of geekdom? Just think how much faster and efficient todays software would be if we applied this philosophy to programming.
Of course things are more complex and hardware considerably more variable in these days of Open Source , cross - platform development etc. Wouldnt it be nice if we at least tried a little harder to avoid the bloat - just because machines get more powerful it doesnt mean you should let your code slip
Electronic Music Made Using Linux http://soundcloud.com/polyp
The introduction and review was in the February 1984 issue, with the Mac on the cover. This is the article for the geek; it includes block diagrams of the architecture and pictures of the motherboard.
The Feb 1984 issue also included an interview with the designers.
I was hoping TFA would be the February article, because it actually is very interesting. In it, they make a big deal about the justifcation for certain design decisions, most notably the lack of expansion slots. Instead, they included "virtual slots", in the form of "high speed" serial ports (RS-422).
Remember that they were trying to solve the problems of the Apple ][, one of was how the expansion boards fitted into the memory map. By eliminating expansion slots, they hoped that it would improve stability, by ensuring that the developers would have a fixed machine environment to work with. They thought that by including all the ports a user would ever need, there would be no need for expansion slots.
Then a couple of years later, Apple decided that expansion slots were good (with the Macintosh II).
It is kind of funny that with the iMac, Apple came all the way around back to the same port-expansion ideas that were discussed in the Febuary 1984 article and interview.
If anyone can find the Feb 1984 article and interview online, it is a good read.
to run real accounting software, which didn't exist on the Mac.
to run lotus spreadsheets with more than 128K of memory which you could not do on a mac.
to run insurance comapny risk-analysis software.
to run stock market tracking and modeling programs
to run video store rental software.
to attach them to a file server and run shared databases.
and so on and so on and so on. To do the things that let you PAY for buying the computer. The things we spent thousands and thousands of dollars for CP/M Z80 computers for, and now, we could do the same thing, faster and with more ram, and cheaper hard disks.
We were networking Z-80 systems with Zenith and Hazeltine terminals in '79 and '80. Once we got Corvus Omninet cards or Arcnet (!) cards and could network PC's at 1 Mhz and then at 2.5 Mhz, all bets were off. The money flowed like water for a while....
Meanwhile, there were a few long-haird weirdos in the back playing with macs and mice, and making pretty pictures. Which was fun and all, but it didn't pay the bills. Of course, Tim Jensen kept playing with the Radio Shack color computer, and having a darned lot of fun, and he and a couple of other guys were sleeping in the back of their shop working on the early video toaster.... but meanwhile, we were making actual money networking PC's with early versions of Novell when we gave up MP/M and TurboDos and went to all '86 processors.
And Tim _did_ hit it big, but in the meantime, those of us wearing suits and ties and selling pc's to lawyers to replace Wang dedicated word processors and to run conflict-of-interest databases (Many of those available for Macs yet?) or law-office-case-management software (another big mac vertical, right?) or large free-text indexing systems, with at the time (1984 remember?) huge 40 and 80 and _90_ meg hard disk drives managed to make decent money.
The macs, and the Amiga had a problem. All that bit-mapped screen stuff was fun and all, but no court in the country would take dot-matrix printouts seriously. No Daisy wheel support in the mac. C. Itoh and Xerox and NEC were were the $$$ where. Now, _later_ after the lasers showed up, even then, remember that the people PAYING to have the contracts wanted COURIER not some weird Times Roman font they'ed never seen before. And mac lasers were expensive compared to early HP and Canon and Oki lasers.....
You bought a PC in 1984 to do things that EARNED MONEY. You bought a mac to play with pictures.
Even as late as 1997, we still were installing monochrome monitors and text-screens. Why? 'Cause if _all_ you do all day is word-processing, multi-tasking DOESN'T make you any money. Even now, the fancy graphics and fonts and colors do little to enchance the operations of accounts receivable software. The biggest advantage of windows for accounting software is that the big screens allow you to see more of the accounting information at once. It _is_ nice to have AR and AP and GL all open at once.... but uh, the mac had little to do with _that_.
I was in the same dilemma as the parent poster: I was mightily interested in the Amiga, but the lackluster sales drones in the local computer store and the confusing (for newbies) articles in the Amiga magazines suggested that you needed a mainboard. Terms were thrown around that suggested that it was a closed club with a secret handshake.
Add to this a bunch of clueless advertisers who try to fit their entire catalog into the smallest possible ad block in the magazine, and the confusion was complete. I threw up my hands in despair and bought a Mac Plus instead.
The Amiga was a great machine, but it was hampered by bad marketing (and, I suspect, a little FUD from the IBM-compatible crowd).