Open Apple Vol 8 No 11 DEC 1992

A2-Central S A journal and exchange of Apple II discoveries Open-Apple i; December 1992 Vol.8, No.11 ISSN 0885-4017 newsstand price: $3.00 photocopy charge per page: $0.20 Coping with fast modems By Dennis Doms You may have noticed that several companies are now carrying high-speed modems at drastically lower prices than a few months ago. Although normally advertised for (and sometimes packaged with soft- ware for) Mac and PC compatible systems, these modems can also be used on Apple II systems. There are many communications standards in use and their names look like alphabet soup, so grab a spoon and let's get a taste of the prominent modem standards. The early personal computer modem standards were Bell 103 (300 “bits per second” or “bps,” also called the “baud rate”) and Bell 212A (1200 bps). These standards were used mostly in the United States and Canada. As modems became more common, standardization moved into the international realm of a committee known as the CCITT. CCITT v.22 is the international 1200 bps standard (it’s similar to Bell 212A) At the next level, things merged; CCITT v.22bis (2400 bps) is used pretty much around the world. At 9600 bps several standards were tried but CCITT v.32 seems to be emerging as the “winner” by popular consensus. Some high-speed modems also support CCITT v.32bis for 14,400 bps. These stan- dards, implemented on a chip set available to modem manufacturers, are the ones supported by the recent plethora of high-speed FAX modems. As with all these standards, the modem must be communicating with another modem supporting the same standard. (That is, a v.32 modem can’t communicate at 9600 bps with another modem using a non-v.32 standard.) This is especially confusing for the several 9600 bps protocols marketed, but v.32 is becoming the defacto “standard”. (If you think 14,400 is fast: a v.32fast standard for 28,800 bps is being proposed.) Many modems also support error detection/correction and data compression protocols. An error-correcting modem sends additional information with the data to allow another error-correcting modem (of the same standard) to determine if something has gone wrong in transmission and to attempt to compensate for it. A modem with data compression is smart enough to “compact” the data so it takes less time to transmit, and the receiving modem (of the same standard) is smart enough to recognize and decompress the data before passing it on to your computer. Several proprietary standards evolved into the Microcom Network- ing Protocols 1 through 5 (MNP 1-5). MNP protocols | through 4 encompass various types of error checking and MNP 5 (which includes the 1-4 standards) adds data compression. MNP 5 is the most common implementation today and offers data compression up to about 50 per cent of the data stream. That is under ideal circum- stances, however. On a precompressed file (such as most files archived for downloading) MNP 5 will actually increase the time need- ed to send the file. Many BBS’s and online services, including GEnie, use MNP 4, which provides error correction but not data compression, for this reason. The common standards you'll see on high-speed modems are v.42 for error correction and v.42bis for data compression. When you set up a high-speed modem connection you need to use two modems with compatible standards and you need to make sure that your terminal (for most Apple II users, your computer with com- munications software) is capable of communicating at the maximum speed of the modem. There are some facets to this last part — the maximum communication speed of your computer, that aren’t obvi- ous. Let's take a look at them. Because high speed modems can compress data, they can actually transfer more data through the phone line than the bps rate itself would suggest. With v.42bis a total compression of 4 to 1 is possible. This means a 14400 bps modem can conceivably trans- fer at a rate of 57600 (4 times 14400) bps, at least for short bursts when the data can be compressed by that ideal 4 to | ratio. This means your computer's serial port needs to be set to 57600 baud so that your modem can communicate with your computer as fast as it does with the other modem. Whether your serial port can meet this challenge depends on what computer and what communications soft- ware you're using, The Apple Ile Super Serial Card is pretty much limited to about 19200 bps on a “stock” Ile. Using an accelerator to speed the "THATS How To ATTAIN ENLIGHTENMENT? COMMAND CPTION CAPITAL E 2" 8.82 A2-Central computer beyond the “factory original” one megahertz may help relia- bility with some programs. The serial port software built into the Ilas will support communica- tions up to 19200 bps; this limit is primarily imposed by the fastest speed at which the programs can grab data from the serial port. The built-in routines are “general purpose” so that diverse programs can use them; a communications package can improve upon these rates by using its own specialized routines that bypass the built-in software. A programmer who does this on the Ilas has to be very careful to insure that the serial port is available; although the Ilcs is technically not a multitasking machine, it’s conceivable that a co-resident pro- gram such as a desk accessory might be using the serial port. The Mac serial port hardware is nearly identical to what’s in the Ilas, though the Macintosh system software is obviously not the same and very fast models of the Mac exist. We decided to test some Apple Il combinations along with both slow (an 8 MHz Mac Classic) and fast (a 20 MHz Ilsi and 25 MHz PowerBook 170) Macs to see how they fared. We used an MS-DOS PC for the host in our tests, The PC has an advantage in that it can use a version of a serial chip that “buffers” (remembers) characters when the computer isn’t quite ready to deal with them. By using a serial port with this type of interface chip (a 16550) we were able to get much better results than with non- buffered chips. Although the microprocessor in a PC can run at very high speeds, in most cases the internal data transfer circuitry, known as the bus, is limited to 8 MHz for compatibility. We performed some speed tests using the fastest Apple II drivers we could find (which happened to be in Proterm 5.0) to compare with some results using Mac and PC systems, which we'll report later. Since Apple II users are a diverse lot it’s hard to be specific about the best combination and ultimate achievable speed, but overall, using optimized drivers (primarily a function of the software you're attempt- ing to use) and faster hardware (usually achieved by accelerating the processor) will help. Since we plan to push the serial port for all it’s worth, we have to be sure that it can shout “Wait!” if it can’t keep up. As usual in the computer world, there’s a software and a hardware way to do this. Software handshaking involves sending control code char- acters in the data stream to indicate when to start or stop the data flow. The most common is “XON/XOFF” handshaking. When the device is nearly “full” it sends the “XON” character (usually defined as control-S) to the sender to tell it to stop sending data as soon as it can. When the receiver is ready again, it signals the transmitter by sending it an “XON” character (usually a control-Q). XON/XOFF is very simple and requires only that the transmitter be ready to accept the signal and stop transfer until told it’s okay to resume. The trouble with XON/XOFF is that it requires the software on the transmitting side to take some time out of sending data to occa- sionally watch for the XOFF character. It also means that the transmit- ter has to be careful not to send more data between checks for the XOFF character than the receiver might reasonably be expected to still have room for (this is subject to tricky limitations and there are no precise “rules” for the reserve space to expect). These factors cause XON/XOFF to interfere with the rate of data flow as the commu- nication speed increases. The other solution is hardware handshaking, where the receiver uses a hard wired connection to tell the transmitter to stop or start transmission. In this case, when the “stop” signal is applied, the trans- mitter’s peripheral interface will tell the transmitting software it’s not ready to send the next character. Since the software normally checks to see if the interface is ready before stuffing the next character in it anyway, this doesn’t require the software to expend extra effort (and time) checking for readiness like XON/XOFF does. Also, since the sig- nal is transmitted at the speed of electricity through wire (instead of at the communications bps rate) the transmitter is able to stop without any risk of overrunning the receiver's buffer. The drawback for this Vol. 8, No. 11 timeliness is that, since the handshake signal can’t be generated on the same wires as the data, you need more wires in your cable to implement it. You can make a cable for a connection using software handshaking with as few as three wires (transmit signal, receive sig- nal, and signal ground); hardware handshaking requires at least two extra wires (more if you want to continuously insure your connection is intact) for bidirectional control. There are several schemes for hardware handshaking; with modem communications, several different handshakes may be used for different purposes. Overall, communications equipment is divided into two general classes: Data Terminal Equipment (DTE) and Data Communications Equipment (DCE). In general you can think of a device that produces output in human-readable form as DTE and a device that moves data to another location as DCE. You computer is DTE because it presents information on a screen. Likewise, printers are usually considered DTE since they put information on paper. A modem is DCE since it sends data to another modem but doesn’t dis- play or print data itself. Handshaking is often defined in terms of these types of equipment. Overall, the common signals are: Signal Acronym = Signals Ground % protective (shield or frane) ground ting indicator nH indicates phone line “ring” Signal ground % comon baseline for signal reference Transnit data bi) data signal output Receive data RD data signal input Data carrier detected DC device is connected (from DCB) Data temninal ready DIR device is ready to send data (to DCE) Data set ready DSR device is ready to send data (from DCE) Request to send Ras device wants to send data (to DCE) Clear to send cs device is ready to accept data (from DCE) Secondary CTS sens (alternate) device is ready to accept data (£rom DCB) These signals are defined here as applying to data communications equipment, but the other end of the signal will connect to data termi- nal equipment (that is, RTS is a signal defined as to DCE, but could also be thought of as a signal from DTE). Some are redundant, such as SCTS. Exactly how these signals are used for handshaking (in the case of modem communications) varies with the implementation of the DTE (computer) to DCE (modem) connection. Differences in the way man- ufacturers have decided to design their serial ports has caused more than a little fragmentation of the RS-232 “standard”; the limitation usually comes from trying to minimize the size of the serial connector. The use of these handshake signals is defined by the RS-232 specification of the Electronic Industries Association. These can be explained in an analogy to the telephone. DTR and DSR indicate whether the person speaking (DTR) and the phone (DSR) are each ready to initiate a conversation. RI is the bell to indicate someone wants you to pick up the phone. DCD indicates whether the phone at the other end of the line is on the line and available for conversation. (Even if neither side is actually saying anything.) RTS and CTS indicate whether or not the speaker wants to speak (RTS) or is ready to listen (CTS). TxD and RxD correspond to the phone’s mouthpiece (for transmit- ting data) and earpiece (for receiving data). A typical exchange may go like this: The phone rings. (RI asserts itself; the modem passes this on to the computer). You pick up the phone. (DTR is asserted to indicate the terminal is responding to the ring.) December 1992 The phone isn’t dead. (The modem is asserting DSR as “true,” indi- cating the modem is ready.) The phone gives you background noise instead of a dial tone. (DCD is indicated by the modem to verify it has established communication with a remote system.) You prepare to speak. (RTS is asserted.) There’s no voice coming through at the other end, so you know they're waiting for your response. (CTS is asserted.) You say “Hello”. (TxD.) The other party is ready to answer. (RTS.) You're ready for them to speak. (CTS.) You hear “Hi, this is Beth.” (RxD.) You want to respond (RTS) but you're waiting for them to finish (CTS). Then you speak (TxD). This loop could go on for hours. One way it could end is for the phone at the other end to die (DCD is lost). Another is that your phone could suddenly fail (DSR is lost). There are other more grisly possibilities, but let’s assume everything precedes normally... At the end of the conversation, either the remote party hangs up first (DCD is lost, telling you the other modem has hung up) or you hang up (DTR is set false, telling your modem to hang up). At this point, you're ready to start again. Back in the real world, it isn’t always this simple. Not all serial devices use all these signals, and so both computers and modems may use alternative methods of monitoring events. You can conduct a serial conversation with as few as three signals (TxD, RxD, and ground) by using software handshaking. However, this isn’t ideal for high speed communications; the telephone equivalent would be explicitly saying “Okay, you can speak now...please wait...go ahead” to let the other party know you're keeping up with their conversation. Hardware handshaking, even with interface peculiarities, is the more efficient solution, and becomes increasingly desirable (and even necessary to maintain speed) as the communication speed increases. Hardware handshaking is also necessary in situations where software handshaking codes might interfere with interpretation of the data stream, such as when raw data is being transferred. (There are soft- ware methods to coordinate these transfers, but they require diluting the data with further control codes.) To use hardware handshaking you need the proper connections and coordinated software support. We took several examples out of the ProTERM manual since it actually recommended the cabling it pre- ferred. The Apple fle Super Serial Card and most PC serial ports rec- ognize nearly all of the defined handshaking signals. The Apple Super Serial Card uses a DB-25 (D-shaped connector with 25 pins), but only 10 wires are actually carried from the card to the connector. The missing signal is the “ring indicator’. The PC recognizes 9 signals on the common “AT-style” DB-9 con- nector (D-shaped connector with 9 pins); SCTS and frame ground are omitted. Some PCs serial interfaces may use the older XT-style DB-25 connector; on ones we've seen, you are often limited to the same nine signals. With plenty of wires, hardware handshaking for the Ile is no prob- lem, a standard RS-232 cable with all 25 pins wired straight through will work fine. (Actually, you only need the wires defined on the Super Serial Card interface, but full cables aren’t hard to find and generally cost little or no more.) The Mac and the gs have a problem due to a dual nature of their serial port. To support AppleTalk, the Mac and Ilas use a high- speed serial interface design that uses a serial standard called RS- 422. There is enough correlation with RS-232 and RS-422 to allow connecting the two, but the RS-422 design calls for two transmit and two receive signals. The transmit and receive lines are configured with “negative” and “positive” pairs and the signal is expressed as a differ- ence in the signal levels across the pair. This means noise is less like- ly to interfere with the proper detection of the signal data and the sig- nal is more reliable. It also means that with RS-422 it is necessary to dedicate four wires (instead of two) to the data flow. A2-Central 8.83 Apple designed the Ilas and Mac serial ports using an 8-pin connec- tor. The supported signals are DTR (handshake out), “HSKI" (hand- shake in), TxD-, TxD+, ground (reference and supply), RxD-, RxD+, and “GPI” (general purpose input), Apple seems to like non-standard nomenclature, so we went to the Ilas schematic (in the Apple Has Hardware Reference, Second Edition) to find out that HSKI is connect- ed to CTS and GPI is connected to DCD. You have to be careful with modem cables on the Ils and Mac. If you plan to use a high-speed modem (9600 bps or higher) you'll want to invest in a hardware handshake cable (some companies sell these designated as “Hayes V-series modem cables”). ProTERM’s manual does caution that you can’t use hardware handshaking, transmit inter- rupts, and true carrier detect all at the same time on the Ilas inter- face. ProTERM does not enable the modem’s carrier detect feature (“&C1" in most modern modem‘s command set) for this reason. The Apple IIc only has five signal lines on the modem port: DTR, DSR, TxD, RxD, and ground (power and signal common). The Ilc port uses a DIN standard connector with only five pins; the IIc-Plus uses the same five signals mapped to the llas-style mini-8 DIN con- nector. Five signals might seem to give the IIc a minimum comple- ment of signals for bidirectional handshaking. But, to complicate matters, Apple’s labels for those pins aren’t strictly kosher. Each “DSR” pin is actually connected to the “DCD” pin of its respective 6551. The real “DSR” pin from the Port 1 6551 ACIA (the serial interface component used in the llc) is connected to an EXTINT* signal on the Ilc’s disk port. The “DSR” pin from the Port 2 ACIA is connected to the KSTRB signal to generate keyboard inter- rupts. Each of these signals is treated as an input from an “external” source (disk port or keyboard) in order to generate an event interrupt. (if you need to know more about this, consult the Apple Ilc Technical Reference manual, especially Appendix E on interrupts.) According to ProTERM's manual, the limited number of signals allows hardware flow control in one direction only. ProTEKM simulates bidirectional hardware handshaking in its driver. Once the handshaking is right, the issue is throughput; how fast the information flows through the system. Using hardware handshaking instead of XON/XOFF ensures the maximum amount of data can be pumped through the modem/computer connection. The remaining question is, “Can the computer keep up with the modem?” The theoretical maximum throughput of the v.32bis/v.42bis modem is 57,600 bps (14,400 bps times a possible data compres- sion factor of 4), Therefore the modem will communicate with its computer terminal at bps rates up to and including that speed. The actual amount of data passing through the modem may be somewhat less. Any increase over the basic bps rate (9600 or 14,400) from the data compression aspect will be limited by how “compressible” the data is. Files compressed with Shrinkit (or a simi- lar high-efficiency compression utility), for example, are already about as compact as a reasonable “lossless” (without any loss of informa- tion) compression algorithm can make them and further effort in com- pression is just wasting time. In this case, the modem will effectively be transferring data at its basic speed. Another limitation is the error rate of communication across the phone line. These modems monitor the signal quality of the phone line and can “fall back” to a lower bps rate if the line quality causes so many errors that data could actually be sent faster at a lower speed. When the line quality seems to improve, the modems can step up to a faster speed. (You might not think that 2400 bps can be faster than 9600 bps, but it may be if a block of information can be sent clearly the first time at 2400 bps as opposed to having to be sent sev- eral times at a higher speed before it arrives intact at the other end.) In other words, using the maximum theoretical speed between the modem and the computer is not an indication that all information is 8.84 A2-Central transferred at that speed, or even that much of it is. Using the maxi- mum bps rate just insures that the data will always flow with as little obstruction as possible. In the real world, the computer may have limits, too. We tried to determine the maximum bps effective rate at which several popu- lar Apple II configurations (as well as a few other computer systems) could transfer data reliably. It turns out that, depending on your con- figuration, you may be better off setting your computer/modem com- munication speed as low as 19200 bps to get the best performance. In fact, certain configurations may require that you set the rate even lower (to 9600 bps). The systems tested were: an Apple Ics with an 8MHz ZipGSX with 16K cache, a Mac SE, a Mac Ilsi, and a Mac PowerBook 170. The soft- ware used was ZTerm on the Macintosh, and Talk is Cheap and Pro- TERM on the Apple Il systems. Files were downloaded over a null modem cable from a PC running ProComm Plus (DOS) in host mode. The file transferred was an uncompressed application file of about 214016 bytes in size. For Talk is Cheap we used the YMODEM transfer protocol. YMODEM transfers blocks of data of 1024 bytes along with a verification value calculated from the data in the block and then waits for an acknowl- edgement from the recipient. If the block is not received intact it is sent again. The more times a block isn’t correctly received the more times it must be resent, so the number of errors and transfer rate is an indication of how successful the connection is. For ProTERM and ZTerm we used ZMODEM transfers, a more robust and complicated protocol. The easy way to think of ZMODEM is that it sends data without waiting until the recipient identifies an error. At that point the recipient tells the sender which block of data was bad and it is resent. Since ZMODEM doesn't wait for acknowledgement of each individual packet it can proceed much more quickly until an error actually occurs. With the Apple II system we timed tests with and without accelera- tion to see how much aid more CPU speed would be. Also, on the Ilas we tried some transfers with AppleTalk enabled to see if AppleTalk’s use of CPU time would hinder high-speed transfers (look for the “AT” setting in the comments). Transfer times were reported in seconds and in characters per second (cps values in parentheses were calcu- lated in situations where the receiving program didn’t report the figure itself). Here's what we came up with: System (bps rate) Tie ops errors = comment Ynoden Apple 85¢ (9600) failed Ihe; Talk is Cheap Apple $5C (9600) 26 (90) 0 fe; talk is Cheap ‘Apple SSC (19.2K) failed Apple Is (9600) 3M (682) 0 2.80mi; talk is Cheap Apple Tes (19.2K) 313-———(684) 0 Apple Tes (9600) 26 906 Mhz; Talk is Cheap Apple Hes (19.2%) 191 1133 ) taoden Apple $8C (9600) m1 888 0 lia; Provera 3 Apple $8C (19.28) 199 1075 2 Apple SSC (19.2) 195 1097 0 Siz; Provera 3 Apple S6C (38.48) failed Apple Ties (19.2%) 1S 0 at off; ip off Apple Tics (38.48) 158183 9 AY off; Zip off Apple Tes (19.2%) 101783 0 af off; Zip on Apple Tes (38.48) 177 120915 AP off ip on Apple Ties (384K) 12S 9 AD off; $2 fast Apple Tie (19.2%) 134518 on ip off Apple Ties (38.4) 872 25 136-—-AT on; tip off Apple Tes (19.2K) 61788 0 AT on; Zip on; $2 fast Apple Te (38.4) 530 43 AY on; Bip on; $2 fast Vol. 8, No. 11 Mac $B (9600) a 05 () Mac SE (19.2K) failed Mac Iisi (19.2%) 2B 195 0 Mac isi (38.4%) (6 8 Mac PBITO (19.2K) 17180 1 Mac PBI70 (38.4) failed The upper end on the Super Serial Card seems to be around 19200; even though ProTERM completed the transfer the time was not significantly faster than at 9600 bps. Talk is Cheap didn't com- plete (it usually got the file header and died with continuous errors) at 19200, but worked fine at 9600. The Ilas serial port firmware seems solid at 19200. ProTERM's low- level driver seems to be pushing things at 38400 but does complete the transfer. Setting slot 2 to “fast” (marked by “S2 fast” in the com- ments) seemed to help, so we left it there for the tests with AppleTalk on. AppleTalk does seem to interfere some; 38400 becomes slower than 19200 due to the number of errors ProTERM has to compensate for. According to Guide to the Macintosh Family Hardware, Second Edi- tion (page363) the “classic” Macs (it wasn’t clear whether this meant only the 8MHz compact versions such as the Plus, SE, and Classic) support a maximum bps rate of 57600 through the Mac ToolBox (higher rates are possible by using externally clocked synchronous transfers, but consumer modems are asynchronous beasts). However, it looks to us like the Classic Macs could only muster a sustained throughput somewhere between 9600 and 19200 bps. The Ilsi made it to 38400 with transfers cleaner than the Ilas, but the PowerBook 170 didn't, dying about 6 to 10 percent of the way into the transfer. (We've successfully used these systems as UNIX terminals at 38400 bps; the limits show up during raw transfers.) Most common current serial ports on PC's will work at least to 38400 bps (the limit for our test). Older serial ports based on a non- buffering chip may not make it that high. The operating system may also influence the throughput; operating systems like Windows and OS/2 that are graphics-based may lower the maximum attainable speed on some systems. What we're trying to convey is that the new high-speed modems do work on Apple Il models with varying degrees of success. And in cases where the Apple I! does have limits, moving to another type of computer may not solve the problem. If you plan to use a software package or interface other than the ones we have mentioned, check with the respective manufacturer to see if the pack- age will support sustained high-speed transfers and what kind of cabling you will need. Also recognize that most people are interested in high-speed trans- fers for the downloading of files. Such files are usually provided as compressed archives and as such will not be further compressed by v.42bis. For these files a successful transfer rate of 19200 bps will exceed the maximum attainable throughput on a 14400 baud modem and even 9600 bps may not represent a severe dropoff.—DJD Miscellanea Apple has finally joined the modern world by going mail- order. Apple has a new 40-page catalog that includes Mac and Apple Il software, peripherals, and accessories as well as PowerBook com- puters. With traditional Apple dealers seemingly an endangered species in some areas and Apple II dealers being even rarer, the new Apple dis- tribution channel finally allows the rest of us to actually get hold of true Apple software and peripherals without having to search high and low. The computer and electronic “superstores” that Apple products seem to be moving toward just don't seem to always have a complete array of Apple products available, especially Apple Il products. December 1992 Several accessories that are Macintosh peripherals also work on the Apple Il and are listed in the catalog: Inageliriter 11 Apple Desktop Bus House Personal Lasertiriter WIR Appleco 150 ‘Apple MIDI Interface Apple SuperDrive Apple Extended Keyboard Apple Keyboard 11 Stylettriter (IN with systen 6.0 only) Add to this several Apple II specific peripherals: Apple Joystick II Apple II SuperDrive Controller Card AppleHouse Ile Apple Disk 5.25 Apple II High-Speed SCSI Card Super Serial Card Tie Extended 80-colum Text Card Apple II Video Overlay Card Apple Ie Accessory Kit with 5.25 Controller Card ‘Apple Ie Memory Expansion Card (slot 1-7 card) Apple II Workstation Card (AppleShare network interface for Ile) And a few software items: HyperCard Ios Apple Ue Systen Software 6.0 Apple II Aristotle Menu Software (AppleShare network menu) ‘Access II 1.3.1 (archaic communications software, not recomended) The mere presence of some of these items in the catalog may wake up some customers to the fact that the Apple II is still around. There are some things we'd like to have seen included (like Apple II CPUs) but being able to point someone to a location to order com- mon Apple ll interface boards and peripherals is still a great relief. Apple has established a toll-free order number at 800-795-1000 (FAX 302-678-9200) which is in operation around the clock; the voice number doubles as the Customer Service number. (There’s no inter- national phone number at this time, although they're working on that.) Orders called in by midnight eastern time arrive the next busi- ness day, according to the catalog (Saturday and Sunday orders roll over to arrive the following Tuesday). For more information, call and order the catalog. And, if you've picked up a used Apple... we're starting to get calls from new Apple Il users who've picked up their systems used. Often the previous owner couldn’t locate the original user manuals. Apple doesn’t print and package these manuals for resale, so they aren’t available through book dealers and similar sources. However, near the back of Apple’s new catalog, there’s a sidebar advertising replacement manuals for several computers. Among them are manuals for the Apple Ile and Ilas. If you’re missing original Apple manuals for your Apple Il, call Apple’s order number to see if they have the ones you need. Also, if you have any scrap of your original manual (or disk) left, you may be eligible for Apple's Media Replacement Program. Beagle Bros has scattered to the four (well, two) winds. In March Beagle Bros transferred most of its Apple II products to Quality Computers for distribution, upgrade and support. Beagle was hard at work on BeagleWorks, its integrated Macintosh software package (with word processor, database, paint, draw, communications, spread- sheet, and business graphics modules). Despite glowing preliminary reports from many Macintosh maga- zines, BeagleWorks arrived late and with some bugs and performance problems. This month we heard that Beagle President Mark Simonsen has sold the package to WordPerfect Corporation. WordPerfect is a major player in the word processing arena with a reputation built on outstanding product support. Mark Simonsen is now the Product Man- ager for the new Works package; Mark Munz also follows to Word- Perfect as the head programmer for the package. Several former Beagle luminaries had already turned to other pro- jects, including the founding of two companies still active in the Apple Il market: WestCode Software (Alan Bird and Rob Renstrom) and JEM Software (Randy Brandt). Beagle is a beacon for the Apple II that will be missed. A2-Central 8.85 Double-wide loads Although most of our readers have moved to the Ilas, many still use Ile or IIc computers or run older programs on their Ilas systems. A question that recurs with astonishing frequency is how to enable the double high resolution mode available on the 128K Ile and its progeny, and more specifically how to display pictures created using double high resolution drawing programs. The loading and displaying of double high resolution graphics (let’s start using “DHR” as an abbreviation) is interesting in itself because it requires several preliminary steps to avoid problems. Most significant is that half of the DHR screen is taken from the auxiliary 64K of the 128K Apple II it is working with. That's the same area that the ProDOS /RAM volume occupies. If you have files stored on /RAM and move something into the DHR screen, the odds are very good that your files will be destroyed. So one requirement is that we have to make sure we don't start copying things to the DHR screen before we have the memory reserved, The simplest way we've found to check /RAM for files is to BLOAD its directory (ProDOS lets you do that) and check the file count in bytes 37 and 38. If it's not “O” then /RAM is not empty; the user has to move the files somewhere safe first. If /RAM is empty, we need to reserve the auxiliary memory portion of the DHR screen. Apple has made this absurdly easy; just BSAVE a file the size of a high-resolution screen to the /RAM volume. /RAM is arranged so that not only is the first file stored in the DHR screen area, it is stored in the correct order for display. Using that knowledge, displaying DHR is easy. Save a dummy file to /RAM to reserve the memory. To load a picture, BLOAD the auxiliary memory half first, then BSAVE it over the dummy file. Then BLOAD the main memory half to the primary high-resolution screen area. Actually, it’s easiest to load the auxiliary half to the main memory high-res display buffer at $2000 (8192 decimal) before saving it to /RAM, then just load the main half over the auxiliary half to complete the picture. The fun and games is often in trying to guess how the DHR picture is arranged. Some programs save them as two files (one for each half of the picture). Some save the whole picture in one file, but either half may be first. In the following program you can set the type to “"1, “2”, or “3” to indicate one of the three popular DHR program formats (TimeOut Paint, Beagle Graphics, or Dazzle Draw); one of the three subroutines at the end of the program will load and display the two halves of the picture. All that’s left is to flip the softswitches to display the picture. You can let BASIC handle most of this by using “PR#3” to initialize the screen to double-wide text (80-column) mode and HGR to display the high-resolution graphics screen. Then POKE 49246,0 to enable the DHR display and, optionally, POKE 49234,0 to display the graphics full screen (without four lines of the text screen mixed in at the bot- tom). Of course, when done, put everything back to normal: POKE 49247,0 to disable DHR (and maybe reset the “mixed screen” setting with POKE 49235,0), use “TEXT” to turn off the graphics display, and issue “PRINT CHR$(21)” (print a Control-U character) to turn off the 80-column firmware. Here goes: 1000 REM == load/display double high-res == 1010 DIM B(7): REM use for parsing byte to bits 1020 TEXT : HOME 1030 IF PEEK ( - 1101) <> 6 THEN PRINT “Requires 128K Apple I1e/c/gs.": STOP 1040 GOSUB 1460: IF (MEM < 128) THEN PRINT “Requires 128K!": STOP 1050 PRI “Checking for eapty /RAM. ..* 1060 PRIT CHRS (4);BLOAD /AM,A$2000, 1512, 018" 8.86 A2-Central 1070 IF ( PEEK (8192 + 37) + 256 * PEEK (8192 + 38)) THEN PRINT “/RAM not eapty!": STOP 1090 REM — the following enables double high-res — 1100 PRINT CHR$ (4);"PRI3": REM activate 80-colum firmare 1110 PRINT : REX send sonething through COU? to initialize finmare 1120 REM (above also initializes 80STORE) 1130 HOR : REM activate high-res ‘1140 PRINT CHRS (4);"BSAVE /RAM/ER.X,A$2000,L$2000": REM reserve DER page in aux men ‘1150 PORE 49246,0: REM activate double high-res 1160 REM (You may want to select “full page" here, too.) 1170 REM For more information, see the appropriate Rardvare Reference ‘1190 REM — You have to figure out which format — 1200 REM Use T= 1 for Beagle Graphics Vol. 8, No. 11 1260 T = 2: REM Our example is a TimeOut Paint file 1270 F§ = “TESTPIC” 1280 REM Line 1340 calls subroutine appropriate to the file format in 7. 1290 REM In all cases, we: 1300 REM a) BLOAD the aux mem section to main mem high-res page 1 1310 REM b) BSAVE the main memory page 1 to aux mem file /RAM/ER.X 1320 REM (This works due to the way ProDOS treats the /RAM device!) 1330 REM c) BLOAD the main mem section of the picture. 1340 ON (1) GOSUB 1600, 1680, 1760 1350 PORE ~ 16368,0: WAIT - 16384,128: POKE - 16368,0: REM wait for keypress 1370 REM — following disables double high-res — 1380 PORE 49247,0: REM deactivate double high-res 1390 PRINT CHRS (4);"DELETE /RAM/HR.X": REM renove our file 1400 PRINT : PRINT CHR$ (21): REM turn off firmare 1210 REM «T= 2 for TimeOut Paint 1220 RMT = 3 for Dazzle Draw 1230 REM One of these will probably work. Otherwise... 1420 REM — outa here — 1430 TERT : HOME 1440 END 1240 REM ...you'll have to figure it out how to load it yoursel#. 1250 REM put filename into F¢ 1460 REM — get memory size from ProDOS (HACEID) — 1470 MACHID = PEEK (49048): REM (BP98 Ask (or tell) ’ Uncle Playing slots (again) Normally, when you go to the Control Panel and create a RAM disk, an icon called RAMS appears in the Finder. I have had a RAM disk icon in Finder, along with the hard drive icons and the 5.25 icon, for years. The problem started when I added the Apple 3.5 Disk Controller Card to my system with the SuperDrive. Now all of a sudden, when I switch slot 5 from “Smartport” to “Your Card" the RAM disk icon called RAMS disappears. I may have a RAM disk, but without the icon how do I know? Ihave had to split my drives up to be able to run all programs and have the option of keep- ing my RAMS disk in the Finder. The SuperDrive is still connected to the controller card but the original drive is back to the Smartport hookup. Any suggestions on how I can daisy-chain my 5.5 drive and still have access to the RAMS disk option and have Copy Il Plus recognize the drives? I’m tired of switching to slot 5 and rebooting every time I need to change direc- tions in my work. Rusty Schneeflock St. Joseph, Mo. As comedian Steven Wright has said, “You can't have everything...where would you put it?” The Ilgs Disk Port requires some built-in pro- gramming (“firmware”) assigned to slot 5 to support any attached 3.5 disks; this firmware also supports the RAMS volume. The built-in hardware and firmware was only designed to support the 800K (UniDisk 3.5 and Apple 3.5) drives, the /RAM5 disk, and (using additional firmware assigned to slot 6) Apple 5.25 drives. Drives that sufficiently resemble these drives also will work; there isn’t a third-party UniDisk 3.5 clone (unless you count Chinook’s Smart- Port-attachable hard disk), but companies like Applied Engineering and AMR make alterna- tives to the Apple 3.5 and 5.25 drives (though we know of one minor glitch with non-Apple 3.5s; the las will wait indefinitely when trying to read the drive when no disk is present). You need additional hardware and firmware to support the SuperDrive, which is why you need the new Apple Disk 5.5 Interface card. But if you install this card in slot 5 in the las, it has to be activated in place of the internal slot 5 (“SmartPort” support). When you disable the internal support, you disable its RAMS sup- port. That’s why RAMS is unavailable. As you've seen, you can’t have slot 5 set to both options at once. You have to decide which option is more important to you and stick with it, or rearrange your slots to accom- modate the 3.5 interface in a different slot, or switch back and forth. This is life. If you have an available slot you can install an Apple I! Memory Expansion Card (“Slinky” card) or work-alike such as the Applied Engi- neering RamFactor to get a RAM disk back without moving the 3.5 controller. Alternative- ly, you could move the controller to a slot other than 5; properly written programs shouldn’t care, and ProDOS 8 will “phantom” a third slot 5 drive to another slot anyway. Assuming you have just these devices, slot 2 can be set to “Your Card” and the 3.5 con- troller can be installed there. Or put in a Slinky card for a RAM disk of up to one megabyte. (If you like the latter idea, we've got 2-5 of these cards around here that we'd sell.) Due to a problem we located with the Apple SuperDrive Controller and the AMR 5.5 disks (you can't do a low-level format of an 800K disk in the AMK 3.5, though it will read and write them fine), I spent some time working on an update to Tom Weishaar's original article on Smartport device identification. Look for some more information in the future.—DJD multi MIDI; advanced BASIC Congratulations to Dennis Doms on his arti- cles on MIDISynth and MIDI generally. At last I know what Multi mode is and that is the way my synthesizers work. | hope there are to be more such articles and I look forward to reading them. W. J. Pattison Adelaide, S.A. Once you have the basic components of MIDI on your Ilias you may want to scout around for other music software. We've found a couple of programs of professional quality and several less complete (but also less expen- sive) programs for experimentation. At the professional level, choices are strictly limited. One thing you may want is more capable sequencing software. The best we've seen is Master Tracks Pro from Pass- Port designs; there are actually IIas and Ife ver- sions. If you find the $399 retail price frighten- ing, there is a lower priced Master Tracks Jr. for about $99. Unfortunately, both of these Programs have been discontinued by Passport Designs. The other professional tool is software to produce actual musical notation (scores). The best we've seen for this is MusicWriter from PyWare (800-222-7536), which is actually a series of three products that can produce scores of varying complexity. MusicWriter doesn’t require MIDI, it can also be used to edit scores directly. MIDI just provides an alter- native method of communicating with the pro- gram. (As with several Ilas applications, the use of an accelerator will greatly reduce the frustration factor of using this graphics.inten- sive program.) December 1992 1480 BYTE = MACEID: GOSUB 1520: REM parse byte into bits 1490 MM = 0: TF (2 + B(S) + B(4)) = 3 THEM MEM = 128 1500 RETURH 1520 REM — parse byte into bit array — 1530 FOR T= 0.107 1540 NBYTS = IND (BYTE / 2) 1550 IP (BYTE < > (NBYTE * 2)) THEW B(I) = 1 1560 BYTE = HBITE 1570 wenn 1 1580 RETUR 1600 REX — Format: Beagle Graphics (two files) — 1610 REX Main nem half is in F§ 1620 REM Aux men half is in FS.AUX 1630 PRINT CERS (4); "BLOAD “;F$;".AUX,A$2000": REM load aux portion to main Denory 1640 PRINT CHRS (4); "BSAVE /RAM/ER.X,BS2000,1$2000": REM save it to aux 1650 PRINT CHRS (4); "BLOAD *;PS;*,AS2000": REM load main menory portion 1660 RETURN A2-Central 8.87 1680 REM — Format: TimeOut Paint (single file) — 1690 REM Main mem half is second half of FS 1700 REM Aux mem half is first half of FS 1710 PRINT CHRG (4) ;"BLORD main menory “4FS4" ,AS2000,1$2000,B$0": REM load aux portion to 1720 PRIND CHRS (4);"BSAVE /RAM/HR.X,A$2000,1$2000": REM save it to aux 1730 PRINT CHRS (4);"BLOAD *;F$;",AS2000,1$2000,BS2000": REM load main menory portion 1740 RETURY 1760 REM — Fornat: Dazzle Draw (single file) — 1770 REM Main nem half is first half of FS 1780 REM Aur nen balf is second half of FS 1790 PRINT CERS (4);"BLOAD *;F$;",3$2000,142000,BS2000": REM load aux portion to main portion 1820 RETURN 1800 PRUNE CHRS {4);"BSAVE /RAM/ER.X,3S2000,1$2000": REX save it to aux 1810 PRINT CERS (4) ;*BLOAD *;P$;",A$2000,142000,B$0": REM load main memory These products also exist in versions for the He using Passport Design’s (also discontinued) He MIDI interface . Many of the less expensive “hobbiest” Products are available through The Big Red Computer Club (402-379-4680). These are programs that may satisfy many users though Jacking professional features. They also in- clude a group of programs that use MIDI as an alternative sound input and output channel. Electronic Art’s Music Construction Set allows you to enter your music by placing notes on a musical staff and then playing them. Constructed songs can be played back through the Ilas speakers (using two instru- ment voices, one “treble” and one “bass*, for the notes on the respective staves) and MIDI via the modem port. You can also play the voices using the MIDI keyboard but the notes aren't added to the staff. Music Construction Set has no means of selecting MIDI configura- tion and there’s no mention of MIDI in the manual. In addition, the program runs as a Pro- DOS 8 application and doesn’t implement standard Ilas interface features (such as a “quit” option from the menus). Activision's Music Studio also allows you to enter notes on a staff and play them through both the Ilas speaker and MIDI port (modem Port only), but the MIDI implementation is much more complete (and documented). Through the “MIDI parameters...” item of the “Options” menu you can set up up to 16 instruments and assign their basic MIDI chan- nel, the preset (voice program) number to use, and a range of notes that will be accepted. The MIDI parameters are saved as part of the file ‘so you can archive them with your music. This gives you more control over the instruments; you can actually play several instruments through the use of program changes to alter voice assignments (in the MIDI device playing the data) or by assigning output to play back on more than one unit (by assigning the basic channels for different voices). Music Studio also has several non-standard quirks that cause headaches. The user inter- face is barely recognizable as a Ils desktop implementation (it really isn’t one), even failing to call the Desk Manager so that New Desk Accessories (and the New Control Panel) show in the Apple menu. It uses its own MIDI driver (APPLEMIDI) that assumes use of the modem port for output instead of Apple’s (APPLE.MIDI) driver that can be assigned to either the printer or modem Port. (Things work much better if you create the “APPLEMIDI” driver by duplicating the 6.0 “APPLE.MIDI” driver rather than copying over the driver supplied on the Music Studio disk. {also try patching the name in Music.Sys16)). I had to disable the ZipGSX before launch- ing Music.Sys16 to get the MIDI functions to initialize and recognize the synthesizer (once the program has started, you can go into the Classic Control Panel and re-enable the ZipGSX using its CDA). Also, the pathname for a “WAVES” folder that Music Studio uses is hard-coded so that it must exist in the root directory. (The installa- tion procedures supplied with the Big Red Computer Club version are careful to point this out.) The program assignment values for Music.Studio correspond to one plus the actu- al program number (program 1-128 in Music ‘Studio translates to 0-127 for MIDI). In general, there are many things wrong here. But if some of the glaring problems could be “patched out” (or patched in) this program might be the best economical MIDI application out there. Several of the remaining programs are “performance oriented.” From Electronic Arts there are Instant Synthesizer and Instant Music. Instant Synthesizer is a simulation of a sampling synthesizer which can play back tracks of music and assign voices from digi- tized sound files. One of the more hilarious examples is a fugue using animal voices (dogs barking, ducks quacking, lambs bleating), When loading the supplied songs, it physically goes out and checks the 3.5 drive for every file, even when installed on a hard disk (because the full pathnames are imbedded in the “package’ file that loads the instruments). Instant Synthesizer supposedly supports MIDI devices, but I couldn't get it to work with the current system software. Instant Music simulates a background combo in several styles (rock, country, classi- cal, jazz, or make your own) playing a cyclic accompaniment that you can “jam” with in a “music minus one” style. The major weakness of most of these pro- grams is their relatively limited use of the Ensoniq’s sound capabilities in terms of the sampled waveforms provided. Most are capa- ble of decent sound, but many of the standard instruments just don‘t sound that good. More recent programs (also available through Big Red) from some of our overseas friends fare better. Oliver Goguel (of the Free Tools Associa- tion) programmed NoiseTracker, a sequencer that runs from ProDOS 8 but requires a Ilas since it uses the Ensonig. It doesn’t include MIDI capability, but it does provide the ability to import and play Amiga “MOD” files. This file format is apparently popular; MOD players for both the Mac and PC compatibles (using a standard sound card like the SoundBlaster) have been spotted. Noisetracker's sound reproduction abilities are impressive; the intro music is stunning. Hiibert Alber’s SoundSmith is another sequencer program with stunning sound and it does support MIDI output with channel assign- ment in Omni, Poly, and Multi modes. You can enter data manually; | didn't find a way to use MIDI for input. Thad two problems: the program seemed to output MIDI signals a few octaves off from the synthesizer (the synth plays C#2 when Sound- ‘Smith lists C#5), and occasionally a MIDI voice would be left in a “Note on” state (as if the controller didn't quite shut down all the voic- es). The latter may be a feature; it occurs when you drag from one note to another on the 8.88 A2-Central screen keyboard and clicking the sustained note stops it. But the ‘Stop Sound’ command won't, and often doesn’t stop all the notes when used to halt the ‘player’. There are at least a couple of companies involved in musical applications of computer systems including the Apple Il. One is Sound it, Electronic Music Products, P.O, Box 5055, Peabody, Ma. 01961, 508-551- 6192. This company provides product sales and advice on computer music products, MIDI interfacing, and so on. Another is Creative Consultation Services (719-687-2210) which supports PyWare and other products. —DJD Not in Kansas anymore A2-Central On Disk for November 1992 has a new Ilgs Shrinkit program. The new GSHK works just fine. Iused it to unshrink the GSHK.DOCS file which appears to include new information because it is a couple of kilobytes larger than the former GSHK.DOCS file. Unshrinking it was no problem with the new GSHK but when I went to print it, using Apple- Works at first, | got an error indication and the file appears as only 1K to AppleWorks. Itried using Sneeze to print it and got nowhere with it; it kept coming back to the lp : © Copyright 1992 by Resource Central, Inc. Most rights reserved. Al programs published in A2-Central are public omain and maybe copied and distributed without charge. Apple usergroups and sgnfcan obers may obtain permission io reprint artes fom me fo te byspectcweten request Publisher: Editor: Tom Weishaar Ellen Rosenberg with help from: Denise Cameron Dennis Doms Sally Dwyer Dean Esmay Mary Johnson Jeff Neuer Becky Smith Jean Welshaar ‘A2-Central fed Open-Apple troup January, 1960—has boon pub- [shed monthly since January 1965, Word-ide pres (n U.S. dolar; aia lalvery nuded at no aol charge): $34 for 1 year, $60 for 2 yaar; $84 for 3 years. Al back sues ae curently aval fox $2 each; bound, indexed ec tons of cur fst sx volumes are $14.95 each. Volumes end wih the January issue an index forthe prior volume is incded wih the February issue. ‘The fll txt ofeach issue of AZ-Centralis available on 3.5 disks, along with selection of the best new pubic domain and shareware fles and pro- ‘grams, for $90 a year (newsletter and disk combined). Single disk are $10. Please send al corespondence to: A2-Central P.O. Box 11250 Overland Park, Kansas 66207 U.S.A. ‘A2-Central is id in an unprotected format for your convenience, You ‘are encouraged to make back-up archival copies or 6asyovead enlarged ‘copies for your own use without charg. You may also copy A2-Ceattal for dstbution to thers. The distrbution fee is 20 cents per page per copy dis- ‘tributed. ‘WARRANTY AND LIMITATION OF LIABILITY, We warrant that most of the information in A2-Ceatral is useful and correct, athough drivel and mistakes are included from ime to ime, usually unntentonaly Unsalsied subscribers may canoe their subscription at any time and receive a full ‘etund of ther last subscription payment. The unfilled portion of any paid wil be refunded even to satisied subscribers upon request OUR LIABILITY FOR ERRORS AND OMISSIONS IS LIMITED TO THIS PUBLICATION'S PURCHASE PRICE. In no case shall our company or our contributors be Kable for any incidental or consequential damages, nor for ANY damages in excess ofthe fas pad by a subscribe. ISSN 0885-4017 (GEnle mall: A2-CENTRAL Voice: 913-460-6502 Printed inthe U.S.A. Fax: 913-469-8507 main menu of Sneeze. So | tried to read the file using Copy II Plus (v8.2) and “View Files in Text” mode. Igot an “I/O ERROR BLOCK $76AE”. I guess the new GSHK works fine but the documentation file has been messed up in some way. Could | please get a copy of the GSHK.DOCS. file that I can use? Thank you! George Milonas Tacoma, Wash. Okay, everyone who bought a Ilas because you wanted GS/OS features, take one step for- ward. Now, everyone who’s still using ProDOS 8 applications to manipulate GS/OS files, take one step back... It could be a little more apparent, but the GSHK documentation is actually in a Teach file, not a text file. Teach is a Ilcs format that supports text with formatting codes (text styles, font types, and so on), and it may not be unusual to see Ilas programs supplied with a documentation file in this format since the Teach application was supplied as part of Sys- tem 6.0. The reason for adding Teach to 6.0 (as for adding TeachText to the Mac system software) is to give Apple developers a “univer- sal” file format to distribute program documen- tation in. Not everyone has AppleWorks, some people even have trouble finding a utility to display text files if they’re not using a word Processor that can convert them. Developers probably won‘t supply Teach since it requires 6.0 to run; if you have 6.0, you have Teach. If you dont have 6.0, Teach files can be read by most Ilas word processors including the one in AppleWorks GS and many of the Ilas text editors distributed as user-supported programs. i Although GS/OS sup- Ports the ProDOS file system it also supports some enhanced features and at least one new file storage type (the “extended” files with resource forks) that ProDOS 8 does not have any intrinsic ability to handle. Copy Hf Plus apparently tries to follow the file physically (block by block); since the extended file for- mat does not use the same indexing as a “nor- mal” ProDOS file. Copy H Plus misinterprets the index and looks for blocks in impossible Places. That's where the “I/O error’ came from. Reading a file can’t cause serious problems as long as the program traps errors. Copying o1 ipulatin dis disastrous if the program makes bad assump- tions and doesn't catch errors. When Ilas Sys- tem 5.0 arrived with extended files many peo- ple tried copying system files to their hard disks with their old utilities and ended up with corrupted systems that wouldn't boot. If you are using disks with GS/OS files it is essential that you use only GS/OS utilities to manipulate those files.—DJD Words over wires Thanks for your treatment of Apple Writer Il in your October newsletter. I use it exclusively Vol. 8, No. 11 in my law firm and have for many years. It net- works, and the features you describe are great. However, Iam very interested in your com- ments on page 4 that there is a built-in telecom- munications module that will work with a modem. Ihave an Applied Engineering Datalink 2400. How do | activate it without leav- ing Apple Writer IP If I print to the slot where the modem is, nothing happens. Do Ineed to make special settings off the Control-O menu (choice J) first? If so, what, and then what do Ihave to do to engage the modem. What other module features exist? 1 have the original documentation for both the DOS 3.3 and ProDOS versions of Apple Writer Ifand can find nothing on telecommunications. Brian C. Sanders Fort Walton Beach, Fla. The telecommunications capability of Apple Writer I is limited to that of a dumb terminal. There are a lot of problems with its compatibili- ty, though; it doesn’t like our Iias at all (even when using a Super Serial Card in slot 2) and it may be selective about the serial interface it wants to use (for example, it doesn't work with the Hayes Micromodem I, if any of you out there still have one), To set it up, type control-O for the “Options” menu, select “J” for “Set Printer/Modem Inter- face”, and pick slot 1 or 2. Then you'll be asked to enter a format string (baud rate, data bits, parity, and stop bits). We tried “2400,8,N,1° (to use the default speed of the interface, use “0” as the baud rate). To connect, type “control-Q” to go to the “Quit” menu and select “I” (“Connect Keyboard to Printer/Modem’). From here, you type com- mands to the serial port firmware (and modem) as if you were in BASIC or in a very dumb terminal program. For example, to dial you might type “AT DT<number>” for an AT command compatible modem. There are four “Escape” commands; you issue any of these by typing the Escape charac- ter followed by the first letter of the command. “Record” copies received text to the word pro- cessor’s buffer. “E)cho” echoes what you type (toggles the duplex between full and half). “Piilter’ filters out control characters other than necessary ones like carriage retum and line feed (highly advisable since receiving a firmware control code could crash your sys- tem). And “Q)uit” takes you back to the word processor screen. “Receive” gives you the ability to capture text. You can send text to the modem by “printing” it to the modem slot, but this can be tricky since you won't have any way to control aspects of how the data is sent (the data is just “bulk dumped’ to the serial port). There are no provisions for file transfers or other common features of real telecommunications programs, If it works for you, the telecommunications support may be useful in situations where you need to talk to another computer at a very basic level and just don’t want to leave Apple Writer. Otherwise, you're better off ignoring it—DJD