AudioVision monitor connections — In response to last week’s discussion about connecting a Quadra 840AV to the AudioVision monitor, we learned that Apple has no plans to modify the 840AV’s (or the Centris 660AV’s) motherboard to add a different port for connecting to the AudioVision monitor. No matter what kind of Mac you have, you need to make sure to purchase an adapter along with the AudioVision monitor. Apple corrected its price list (as of 16-Aug-93) to avoid future customer confusions.

This article, a deliberate takeoff of urban legend article, "Your Co-worker Could be a Space Alien," goes out with a grin to people who do telephone technical support.

Many callers are actually space aliens who sound human, but you can expose these visitors by looking for certain tip-offs, say experts.

Odd or mismatched software and hardware. — Aliens might run "MacWriter on their Apple Microsoft," or use "PerfectWord on a Claris 610."

Strange or unusual software habits. — An alien might not understand the hierarchical filing system, so she might put all her files and applications in the same folder. Aliens can also exhibit paranoia, so an alien may worry that if he throws out one document, all other documents will disappear. Perhaps this occurs in different solar systems.

Bizarre sense of humor. — Aliens often exhibit inappropriate senses of humor. For instance, an alien may burst out laughing in the middle of a serious discussion about the difference between clicking and dragging.

Puts you on hold frequently. — Aliens have trouble with human language and often must put you on hold to consult references in order to figure out what you are saying.

Keeps a written or recorded diary. — Aliens worry that they won’t retain every subtlety of the conversation and feel more comfortable recording it.

Misuses everyday items. — Aliens have trouble with human-based objects, so an alien would be more likely to use a mouse upside down. Aliens having Pivot monitors are constantly pivoting them. Scientists speculate that they are trying to align their monitors with a distant magnetic wave system, but tests have been inconclusive.

Constant questions about customs. — Aliens want to know why thing happen as they do, so they may interrupt every suggestion you make to ask why or to suggest another way of doing it. Don’t be put off by this behavior, aliens can’t help it.

Secretive about personal life-style and home. — Aliens don’t like to give anything away for fear of blowing their cover. If a caller has a great deal of difficulty telling you, say, the names of the fields used in a print merge, and then finally tells you the information is classified, you are probably assisting a space alien. Another tip-off is aliens that cannot reveal the names of their hard drives.

Frequently talks to himself. — The alien may be rehearsing or practicing what he will say next.

Displays a change of mood when near certain high-tech hardware. — An alien may become nervous or hyperventilate when near computer hardware. Aliens also become unobservant and have trouble reading labels. If a customer cannot identify the name of the computer, and the name is affixed to the front of the computer, you may have a space alien on your hands.

Experts point out that a caller would have to display most if not all of these traits before you can positively identify her as an alien.

For the majority of Macintosh users, those without copy-protected software, this article will be a minor curiosity. Lucky you. For those left, you will want to read this to save yourself the hours of aggravation that I went through installing The Latest Upgrade. Specifically, we’re dealing with professional music software.

Music software (specifically, MIDI sequencers, patch editors and librarians) has been copy-protected basically forever. I bought a copy of Mark of the Unicorn’s (MOTU) Performer in 1988, and it came on a copy-proof floppy disk which had to be "keyed" into the Mac for the program to run. The program won’t run without a key disk, and the key disks carry formatting errors to circumvent duplication. If the key floppy gets damaged for any reason, you’re lost, regardless of the number of dollars per second being burnt in a production studio, or the size of an audience. If you live in the U.S. you stand a chance of getting a replacement, but if you live outside the U.S. the support is probably such that a change of career is a better idea.

A couple of years later a new scheme came along, and was adopted both by MOTU and by Opcode who by this time had a pro-level sequencing package of their own. There were still key disks, but there was also a procedure for installing on a hard disk. An "install count" on the key floppy would decrement in exchange for a copy of the application on the hard disk. At this time, another mechanism would kick in to avoid duplication of the hard disk copy. The application’s location on the disk would be stored in a secret format in hidden key files also on the disk, preventing making a useful copy of the application from the hard disk since the copy wouldn’t have the keying information.

The advantage now is that you don’t have to mess around with floppies. But the disadvantages to legitimate users are significant. Disks cannot be optimised without destroying the install key. And of course, backups are useless; damage the disk for any reason, and you lose the key forever. If you get an update of the program, you have to go through a deinstall/reinstall procedure. You rely on the scheme working properly, and it doesn’t always. A legally installed copy will, every now and then, declare itself illegal and refuse to run.

But at least it stops piracy and keeps the software companies in business, right? Wrong. Leaving aside the arguments about whether illegal copying of software eats into sales or not, you can circumvent these protection schemes. These methods are not trivial, but are reasonably well known. So all the schemes do is inconvenience legitimate users.

Anyway, a lot of music software seems to use the same scheme. Or, I should say, "used." There is now A New Scheme, featured in Performer 4.2, the latest version of Opcode Galaxy, and on its way in Vision and MAX. It works as follows: instead of installing a copy of an application with a specific key, you attach a key to a hard disk. The application can be moved around (or upgraded, an important benefit) without wrecking the key. That’s the Major Pro. Major Con: the protection key is buried deep within the disk. I suspect a pointer within the boot blocks, but that’s only a guess. The installation process is now one of authorising a disk, rather than installing an application.

The more complicated the scheme, the more that can go wrong. The more complex the scheme, the more it has to assume about the target configuration, and the less likely the assumption is to be correct. The Performer installation blew up several times over a period of several hours. I partition my disks with Silverlining, and yet the installation instructions and scripts assume a single-partition disk. Installation has to be to the boot "disk" – so I had to resize partitions to make space. The instructions stated that any copies of the application on the disk would run. Wrong: the application is keyed from its own partition. Not a serious drawback, but the wording of the instructions shows this eventuality hadn’t been considered, which is worrying. So, I had to move the authorisation tag from one partition to another, which failed several times, due to a bug in the copy protection machinery which causes it to not always recognize a legal key.

Hopefully your experiences with these schemes won’t be as problematic as mine. But if they are, the above notes may be useful.

Floppy-based copy protection stinks, yes? Magical hard disk installation schemes stink too, yes? So why not use a hardware dongle instead? This is the course taken by Steinberg for their Cubase sequencing software. It uses an ADB dongle. And it works well… except on PowerBooks, where sleep mode causes the dongle to malfunction. Of course, we all know that copy protection schemes serve no purpose except to inconvenience legitimate users, and the pressures of the music business (studio recording and live work) make this more significant than in some other fields. So, is there any chance of such schemes being dropped? The only indication I have is from Mark of the Unicorn, and I quote, "not a chance." Let me leave you with an excerpt from Apple’s document "Antipiracy Technologies," on the subject of key floppies and hard disk installation schemes:

"Since this kind of copy protection depends on specific characteristics of the hardware, the copy protection may occasionally malfunction, thus preventing your customer from legally using the program he or she bought. Also, you have no way of knowing whether the floppy-disk copy protection that you use will work with future hardware from Apple and other companies.

"Technical reasons aside, floppy-disk copy protection is bad because it breaks the most important law governing antipiracy technologies: Thou Shalt Not Annoy the Legitimate User. Limited-use installer programs and key disks are inconvenient to use, and they often cause legitimate users problems. When this copy protection malfunctions, you suddenly have a very unhappy customer, one who sees copy protection as a useless encumbrance that doesn’t deter the software pirate but that penalizes the legitimate user for being honest."

[If you have comments about this article, please hold off on them for now. Copy protection usually beats up a storm of debate, but Adam’s book is keeping us busy and we just don’t have time to read your thoughts this week. Thanks for understanding. 🙂 -Tonya]

Apple’s newest Macintoshes are fast machines with new video and sound capabilities. To review, the new Centris 660AV, formerly known as the Cyclone, uses a Centris 610 case; likewise, the new Quadra 840, once known as the Tempest, lives in a Quadra 800 case. As usual, Apple’s naming scheme comes from marketing folks who think it’s funny that people try to memorize this stuff (especially considering that the latest rumors have Apple ditching the Centris name entirely and calling all of the old Centris machines Quadras). At least AV seems to stand for audio-visual. Introductions aside, what’s the deal with these Macs and should you buy one?

If you don’t care about audio-visual applications, you might buy an AV Mac for the 68040 chip, faster NuBus 90 architecture, and SCSI DMA (Direct Memory Access). If you care about sound, video, or telephony, then the AV Macs may be the place to be, since they sport a DSP chip (similar to the one used in the now-obsolete NeXT machines) to keep these capabilities humming along. A few of quirks worth noting: AV Macs don’t support A/UX, they only support 32-bit addressing, and the serial port has an extra pin. Although the port works with the usual 8-pin cables, the new ninth pin carries power to a pod (more on pods below), and lets you turn on your Mac by calling it on the phone.

In the Macintosh AV room — If it all works as described, you can use the new DAV (Digital Audio Visual) port to bring in video from VCRs and video cameras. You can play the video on the screen in the provided Video Monitor program, convert it to a QuickTime movie, convert video frames to PICT images, play around in the included VideoFusion Fusion Recorder, and output video to VCR tape or an attached TV screen.

The DSP chip works with Apple’s new and still-unfolding GeoPort architecture to provide the capabilities of a modem, fax board, speakerphone, answering machine, or even an Ethernet- or ISDN-based videophone. The videophone gives 15 frames per second in a small window, so we’re not talking about putting someone onscreen as in Star Trek. Each AV machine comes with an adapter box, known as a pod, that attaches the Mac to the phone system. This is so Apple can make different pods for phone systems in different countries, not to mention ISDN and digital PBX, which should all be available sooner or later.

The AV Macs also include PlainTalk – formerly know by the code name Casper – software that works with the DSP chip to turn speech into computer commands and text into speech, of sorts. I’ve heard a PlainTalk version of a paragraph from TidBITS, and although it sounded like a computer, I understood every word. Evidently, there are a few different voices to pick from. Currently the speech-recognition works for adult, English-speaking, North American voices, but more options should come soon.

The AVs come wired so you can speak to them, as in, "Computer, Control Panel. Computer, Memory." You can create complex voice macros using QuicKeys or AppleScript (a special version of QuicKeys ships with the computer). To help the Mac pay attention to you, you preface every command with a keyword (like Computer in the example above) and the AVs come with a special microphone. Note that the microphone and PlainTalk ship with Macs sold in the U.S. and Canada, but everyone else has to buy them separately. I wonder if this will change when PlainTalk can recognize more voice types.

Sound Manager 3.0 comes as part of the System Enabler, so you can play with sounds, distorting or enhancing them, and you don’t have to buy any third-party software to at least have some basic fun.

Conclusions — The AV Macs seem like a grand experiment. What features will prove popular? Which ones will be ignored? At this time, it’s unclear if developers will write software and continue to write and support software that takes advantage of these Macs, or if developers will focus on the larger PowerPC market.

These Macs could be sensational fun. Not that the capabilities can’t be created on other Macs with the right hardware and software, but that Apple has put a lot of possibilities together in one package, so you get a blazingly fast Mac along with software to play with. Not only can you create and view movies, but you talk and listen to the Mac. I wouldn’t buy a Video Spigot board and all the trappings, because I don’t know enough about video to justify the purchase, but if I had an AV Mac, I’d explore all of its capabilities. Whether the new abilities are up to reliable business uses, intelligent educational applications, or great works of art remains to be seen, but if they don’t measure up, Apple is bound to improve them next time around.

The MessagePad, Apple’s new Personal Digital Assistant, is a type of Newton. So, you may ask, what exactly is a Newton? Newton represents an architecture that combines hardware, software, and application technologies. The Newton architecture doesn’t try to create a miniature computer. Instead, by discarding bulky desktop computer trappings like keyboards and hard drives, the architecture provides mobile yet powerful systems that have rich communications functions. Importantly, Newton is not proprietary. Apple intends to license it, and has already done so to Sharp, Siemens, Motorola, Cirrus Logic, and Matsushita.

An Outside Glance — The MessagePad typifies the design goals of the Newton architecture. The system is diminutive enough to carry around like a paperback book. It weighs just under one pound, measures .75 inches thick, and has a 7.25 by 4.5 inch face. A low-power, reflective, LCD, 336- by 240-pixel, black-and-white screen displays information. The MessagePad has no keyboard; the primary input device is a passive plastic stylus. A resistive film over the screen senses and tracks the stylus motions. The MessagePad comes with an RS-422 serial port and a low-power, half-duplex, infrared transceiver built in. The serial port uses a Mac mini DIN-8 connector, and like the Mac, this port doubles as a modem port and LocalTalk network connection. The infrared transceiver uses Sharp’s infrared communications protocols to "beam" data at rates of up to 19,200 bps (typically 9,600 bps) to another Newton at distances up to three feet. For communications, there’s a seven-ounce fax modem that runs off the MessagePad’s power charger or two AA alkaline batteries.

Pad Power — Stand-alone power comes from four AAA alkaline cells or a rechargeable nickel-cadmium battery pack. Apple estimates battery life at two weeks for alkaline cells, and one week for the battery pack. A small lithium cell preserves memory while the batteries are changed, and an ingenious interlock mechanism prevents you from removing the battery pack and the lithium cell at the same time. You plug into the wall using a universal power adapter (100V – 240V, 50 – 80 Hz), which also recharges the battery pack. Different power adapters are available for the U.S., Japan, Australia, and Europe.

Inside the MessagePad — The heart of the MessagePad is a general-purpose 20 MHz ARM610 RISC microprocessor made by Advanced RISC Machines. The ARM610 combines on a single chip an ARM6 RISC CPU, a 4K cache, a write buffer, and an MMU (memory management unit) designed to Apple’s specifications. The write buffer enhances performance by storing up to eight 32-bit data values and writing them to main memory independently of the CPU. The MMU implements a two-level page-table structure and certain extensions that make it suitable for object-oriented systems. A read-lock-write operation performs indivisible memory updates, which are essential for testing and setting semaphores or handling shared memory in a multitasking operating system. The ARM610 can handle Big-Endian (Motorola) and Little-Endian (Intel) addressing modes, and has fast interrupt response features that make it ideal for real-time processing. It also consumes little power (less than .5 watts), and is fully static (when the processor’s clock stops, its internal state is preserved and it consumes little power). The small ARM instruction set (it has only ten basic instruction types) provides high-level language support, yet is straightforward enough to allow assembly language programming. Last but not least, the ARM610’s manufacturing costs less than $25 in quantities of 100,000.

A custom Apple ASIC (Application Specific Integrated Circuit) controls the MessagePad’s power and the various memory and I/O subsystems, including the serial port, the infrared transceiver, and the LCD screen. It handles sound generation, DMA (Direct Memory Access) for sound and serial data, the PCMCIA card interface, timing signals, and calendar functions. By the way, if you wondered, PCMCIA stands for Personal Computer Memory Card International Association.

The MessagePad uses 4 MB of ROM to store its operating system and several built-in applications. Included here is the handwriting recognition engine written by Paragraph International, a U.S.-Russian joint venture in Sunnyvale, California and Moscow.

The MessagePad has 640K of SRAM (Static RAM) for working memory and data storage, of which approximately 200K is free for user data. There’s no hard drive to save data, so users will probably want a flash memory card for backups. I had a chance to try the 2 MB PCMCIA RAM card from Epson. I inserted it into the card slot, and switched the MessagePad on. Although this card was originally designed for PC notebooks, the MessagePad recognized it, and asked to erase it. After several seconds, I had an additional 2 MB of memory to work with – a good show for Apple PIE, Epson, and hardware standards. The Newton Intelligence allows you to file individual schedules and notes to the card, or make a backup of all data.

Other cards let you add functionality to the MessagePad. Lots of cards should be coming from various vendors, and one of the main cards to look forward to using is the PCMCIA Messaging Card, which consists of a one-way wireless pager that can receive text messages.

The Newton Operating System — The MessagePad uses the Newton OS, or Newton Intelligence, in Apple parlance. The Newton Intelligence (NI) is an object-oriented, multitasking OS from the ground up. However, it borrows a number of useful components from the Macintosh. Portions of QuickDraw – the Mac imaging engine – handle the display. The NI uses the AppleTalk protocol stack to support name look-up, zones, data streams, and printing. There’s currently no support for AppleTalk Remote Access. However, the NI is designed to be extensible, and Apple plans to distribute software updates and extensions via online services. Apple built the NI on several software components: the Recognition Architecture, the Communications Architecture, the Information Architecture, and the Intelligent Assistance.

The Recognition Architecture — The Recognition Architecture deals with stylus input, and a text recognition engine – or recognizer – handles printed text, cursive text, or a mix of the two. The recognizer is trainable, and adapts to your handwriting over time (about 150 words, if you use the built-in training program). You can also specify the type of text you write and the character shapes in the Preferences section to short-circuit the training process. The recognizer uses a 10,000-word dictionary to assist in word recognition. This makes for fast word identification, but it also means that the results of a misinterpreted word, or for a word missing from the dictionary, can be wide off the mark. The user interface allows new words to be added to the dictionary. The overall design makes the MessagePad easy to adapt to overseas markets. By changing dictionaries, you can make the Newton Recognition Architecture handle new languages.

A graphics recognizer detects specific objects and stroke symmetries as you sketch. A round scribble becomes a circle; four perpendicular lines become a rectangle. These recognizers detect several predefined gestures to edit or erase text. Both the text and graphics recognizers can be switched off so that the MessagePad only records digital ink. However, it’s better to have the recognizers in action, because text and object-based graphics require less storage. Also text can be searched for, or used by other applications. Due to storage constraints and usability problems, the MessagePad doesn’t support deferred recognition – the ability to store screens of ink and then extract text and graphics later. However, nothing in the Architecture precludes this feature from being added in the future.

The Communications Architecture — The Communications Architecture supports a variety of I/O devices. With a serial cable, the MessagePad prints to Apple serial printers. An optional cable, available in the Print Pack, stores printer drivers and allows the MessagePad to print to a wide variety of parallel printers. The same serial port allows you to print to PostScript printers through a LocalTalk connection. The infrared transceiver lets you exchange data with another MessagePad or a Sharp Wizard 9600. This architecture is extensible, so new devices can join the repertoire.

The Information Architecture — The Information Architecture levers off the NI’s object-oriented design. As data enters the MessagePad, the Information Architecture tags (or names) it and compresses it as required. Data is saved in an object called a frame, a structure composed of data stored in tagged locations called slots. Slots can contain data values, code, and even other frames. Frames use a unified data model which eliminates data translation overhead and the redundant copies such translations often make. This common data format also promotes data sharing among Newton applications. As objects, high-level meanings can be attached to frames: address, for instance, can represent "One Phoenix Mill Lane."

A collection of related data frames is known as a soup, and soups reside in physical locations called stores. A soup of contact names might exist on a store that’s located on a RAM card. Soups maintain indexes to the frames within them. What this means is that all information entered into the MessagePad automatically becomes part of an object database that can be searched, modified, and displayed in different forms by various applications. All information is processed as Unicode, which means the MessagePad can readily handle foreign languages.

The Intelligent Assistance — The Intelligent Assistance uses the soup of data objects to make plausible connections between information and your actions. For instance, to see the Intelligent Assistance in action, draw a sketch in the NotePad and ask the MessagePad to "fax to Lisa." The Intelligent Assistance will assume you want to send the sketch. Next, it will locate everyone named Lisa in the object database. If there’s more than one, the MessagePad prompts you to make a choice. The Assistance then retrieves the fax number associated with the desired Lisa and generates a fax image. If a fax modem is connected to the MessagePad, the Assistance transmits the fax. Like the rest of the NI, the Intelligent Assistance rules can be extended to handle new applications or devices.

The synergy of the Intelligent Assistance with the object database makes the MessagePad a winner. The ability to locate information within the system and act on it in rational ways is a major improvement. Once a contact and her relevant information is captured in a MessagePad loaded with communications options, you can easily call her, fax her, schedule appointments with her, or send her email. The Newton OS eliminates the many redundant operations (launching an application, locating a file, opening the file, and searching for data in the file) that you do on a desktop computer to accomplish the same thing, and it’s a model desktop computers should adopt. The Newton implementation is by no means perfect, but it’s a huge step in the right direction.