Open Apple 1987 01 January 1987 Vol 2 No 12

Open-Apple Releasing the power to everyone. How to keep from getting pregnant Of all the literary works that I've collected, the most dog-eared is a small volume I've had more than 12 years and still refer to at least weekly. An early section of this work has a title very similar to the one on this article. That section begins like this: Say you are given a new racing Ferrari that you will be forced to drive for the next seventy-five years. No doubt you'd clean the motor with Q-Tips and hover over its gas and oil mixtures like a French chef pouring aspic. Inexactly the same way, you are imprisoned in the vehicle of your body (for seventy-five years!) with a chassis that can never be exchanged nor replaced. Poor fuel reduces your speed because it starves your body, mind, and spirit. Don’t blame your psychological problems all on early home life and your parents. You may be creating your own problems right now—or at least not actively correcting them—by constant consumption of Non- Foods, those over-processed, over-priced products that litter our grocery shelves... The book, known world-wide as “The Campus Survival Cookbook,” by Jacqueline Wood and Joelyn Scott Gilchrist, goes on to chop and slice home the wisdom of eating a decent breakfast. The book comes to mind because here at Open-Apple you can be sure we'd never begin an article with a title like this just to trick you into reading it Before we get down to the bare facts, however, | was wondering whether you Il-Plus and Ile users knew that you can easily add a connector to your computer that's much like the disk connector found on the back of the IIc and IIgs? An unlimited number of slots. This little connector has all kinds of potential uses. Apple's technical manuals call it the Smartport Bus connector. Theoretically, you could connect all kinds of devices to your Apple II through this connector— printers, modems, robots, and compact disk players, to name a few. What's neat is that all of these devices could be connected to your computer at the same time and use up just one slot. This is because Smartport Bus-compatible devices can be daisy-chained together—each is supposed to have a connector on the back that the next device can be plugged into. (To avoid having your devices jerked all over your computer room, however, | recommend giving the robot plenty of cable and plugging him in at the end of the chain.) Unfortunately, this is all theoretical—at the moment there aren't any printers, modems, or robots that will connect to a Smartport Bus. The only things you can successfully connect to the Smartport Bus today are disk drives, We'll look at the kinds of disk drives that work on the Smartport Bus in a moment—first though, let's examine two other ways of connecting daisy- chains of devices to Apple IIs. One of these is the AppleTalk network, which is available only on the new IIgs. Much like the Smartport Bus, AppleTalk allows you to connect your computer to a chain of devices such as printers, modems, and disk drives. AppleTalk has the advantage of allowing more than one computer to be a part of the chain. Consequently, AppleTalk also has the disadvantage of making you share your devices with other people. Unlike the Smartport Bus, whichis all theoretical except for disk drives, AppleTalk, at the moment, is all theoretical except for printers. The other method of daisy-chaining devices is called the Small Computer System Interface (SCSI, pronounced “scuzzy’). Apple announced its SCSI controller card, which works in any slotted Apple II, including the II-Plus, in late September (see Open-Apple, October 1986, page 2.67). Other vendors January 1987 Vol. 2, No. 12 ISSN 0885-4017 newstand price: $2.00 photocopy charge per page: $0.15 are announcing similar cards. The advantage of SCSI is that itis a computer- industry-wide standard. SCSI-compatible devices work with any computer that has a SCSI Bus connector. Consequently, SCSI devices should have bigger markets and lower prices than Smartport Bus devices, which will probably all be manufactured by Apple, Inc. The SCSI standard allows several different types of devices to be connected to the daisy chain, including direct-access devices, such as disk drives; sequential-access devices, such as tape drives; printers; “processor devices,’ such as other computers; write-once read-multiple (WORM) devices, such as optical disks; and read-only devices, such as compact disks (CD-ROM). A small disadvantage of SCSI in the Apple II world is that it can only be used in slotted Apples. The Apple IIc can't use it. In addition, SCSI, like the Smartport Bus and AppleTalk, is mostly theoretical at the moment. While SCSI was designed to handle a wide variety of devices, the only thing you can hook to an Apple II through a SCSI card right nowis a disk drive (and only a hard one at that). Disk drives and the Smartport Bus. Both SCSI and Apple’s Smartport are “intelligent” buses. This means that the devices connected to the bus are all supposed to have their own little computers built into them and are supposed to be able to respond to a few high-level commands. They don’t have to be told in detail such stuff as how many stepper motor steps are needed to move their disk drive arm from track 12 to track 16. Dumb devices, on the other hand, have to be told everything in exhaustive electronic detail. One major advantage the Smartport Bus has over SCSI, however, is that, in specific situations, itcan also support dumb devices. The dumb device you'll most frequently find connected to the Smartport Bus is a 5.25 inch floppy disk drive. On the Apple IIc, you can plug one dumb 5.25 disk drive into the end of the Smartport daisy chain. It will appear to be connected to slot 6 as drive 2 (even though other Smartport devices in the chain will appear to be in slot 5). On the Apple IIgs, you can plug two 5.25 drives in, again at the end of the chain. They will both appear to be connected to slot 6, Any Apple-compatible 5.25 drive is capable of working in a Smartport chain, incidentally, but not all of them have the right connector. Older drives have what's called a 20-pin header connector. Newer drives have the Smartport-compatible DB-19 connector. You can buy adapters to convert one to the other, or you can build an adapter using the instructions published here in December 1985 (page 90). “1 WENT LOCATED THE FROBLEM YET." 2.90 Open-Apple On the new Apple ligs, you can also plug “Apple 3.5” dumb drives (as opposed to “UniDisk 3.5” intelligent drives) into the Smartport Bus connector. Apple 3.5 dumb drives work with the Smartport Bus only on the Ilgs, however. The IIc Smartport Bus (available on any IIc that has the “3.5 ROM” or the “Apple Memory Card” upgrade) requires UniDisk 3.5 drives. And the Ile/II-Plus Smartport Bus (that is, the UniDisk 3.5 controller card), supports only UniDisk 3.5 drives. Itwas not designed to work with either 5.25 floppy drives or the non-intelligent Apple 3.5 drives. Somewhere there is someone who knows for sure how many drives can be connected to the IIgs Smartport Bus connector, but we haven't found that wizard yet. All references agree that two 5.25 drives can be installed at the end of the chain and that these will appear to be connected to slot 6. All references agree that one or two dumb 3.5 drives can be connected at the beginning of the llgs Smartport chain, and that these will appear to be connected to slot 5. According to Setting Up Your Apple ligs, “You can attach up to four drives to the port by connecting them to each other ina daisy-chain fashion....Only two of the drives can be 3.5 inch drives.” (And only two can be 5.25 drives, we might add.) According to the beta draft of the Apple IIgs Firmware Reference Manual (available from the Apple Programmers and Developers Association, see September 1986, page 2.57), “Current implementations of ProDOS only support two devices per port or slot. If more than two devices are connected to the device chain, devices beyond the second cannot be accessed with ProDOS 111 ProDOS 12 will support up to four devices on Smartport. ProDOS 12 will map the two devices beyond the second so that they will appear to be connected to slot 2....No remapping of units to slot 2 will be necessary with ProDOS 16 since ProDOS 16 will support more than two devices per port or slot. And, from the beta draft of the Apple IIgs ProDOS 16 Reference, “Up to four devices may be connected to Smartport. However, because ProDOS 16 supports only 2 devices per slot, the third and fourth devices are treated as if they were in slot 2.” More about this shortly. Distinguishing the bus from the bus station. As we've discussed in Previous issues, the word “bus” is used in electronics circles to mean a group of wires or “lines” canying related signals. It's ok, however, for those of you who aren’t magnetically attracted to this meaning of "bus" to think of the Smartport Bus as a large yellow vehicle for moving data back and forth between your computer and the devices that are attached to it What is important, however, is that you distinguish between the Smartport Bus, which consists of cables, connectors, and electronic signals, and a Smartport-compatible protocol converter, which is the software interface that sits between your programs and the Smartport Bus. The reason it's essential that you make this distinction is that a Smartport-compatible Protocol converter can also be used to access devices that are connected to other Kinds of buses, or even to no bus at all. Think of a Smartport-compatible protocol converteras a bus station. It is clearly marked so that all programs can find it. It has a consistent ticket window so that all programs can use it. What's beyond the bus station’s departure gate, however, can be just about anything. Apple's SCSI card, for example, uses a protocol converter that looks just like Smartport to access devices on the SCSI Bus. Most programs don’t distinguish, nor should they, between a SCSI Bus device, a Smartport Bus device, or even an Apple Memory Card. All of these devices use a protocol converter that looks like Smartport, but only one of the three actually uses a Smartport Bus. A Smartport-compatible protocol converter can support up to 127 devices. Each device on a single Smartport chain is assigned a unit number. The protocol converter itself gets to be unit number zero, the others are 1 through 127. However, the Smartport Bus almost certainly doesn’t have the spark to access that many devices. Its electronic signals will decay long before they get to the last device in the daisy chain. Likewise, a SCSI bus, by definition, will support nowhere near 127 devices. Worse yet, as we saw a few paragraphs ago, ProDOS 111 doesn’t want to deal with more than two devices per Smartport; developing versions of ProDOS appear to deal with only four devices per Smartport (though that’s not totally clear). This could be a significant limit in terms of the llgs, because in addition to any disk drives plugged into the slot 5 Smartport Bus connector, the IIgs supports both an internal RAMdisk and an internal ROMdisk. Vol. 2, No. 12 These two devices also appear to be connected to the slot 5 Smartport daisy chain. (The IIgs automatically creates the RAMdisk, when instructed to through the control panel, out of extra RAM installed in the IIgs memory expansion slot. This memory expansion slot is also capable of holding ROM —although no cards are currently available that support this. This ROM can be configured several ways, including as a ROMdisk.) Thus, if the RAMdisk and ROMdisk were being used, the maximum number of devices (not counting one or two 5.25 dumb drives) that could be connected to a IIgs slot 5 Smartport Bus connector would be two. How to find the bus station. If you are writing a program that needs to be curious about what kinds of storage devices are available to it, you can easily doa “slot scan” to look for Smartport-compatible protocol converters. Slot scans have a long history. The tradition began with the Autostart Monitor in the Apple I!-Plus. When you tum an Apple II-Plus or later Apple II on, the Monitor scans the slots, beginning with slot 7 and working downward, looking for a disk drive. It boots the first one it finds. As we've explained in previous issues, each Apple II slot has a 256-byte memory space dedicated to it. What appears in this memory space is a set of machine language routines. These routines actually live ina ROM chip on the controller card plugged into that slot. (On the IIcand IIgs there areanumber of built-in cards that follow the same format. This address range, in hexadecimal, starts at $Cn00, where n is the slot number. In decimal this translates to byte 49152 + (SLOT*256). Ifyou let ADR = 49152 + (SLOT*256), then what the Autostart Monitor looks for is a 32 at ADR+1 ($20 at $Cn01), 0 at ADR+3 ($00 at $Cn03), 3 at ADR+5 ($03 at $Cn05), and 60 at ADR+7 ($3C at $Cn07). All 5.25 Disk I!-compatible floppy disk controller cards have these values in these bytes. In enhanced Apple Iles and later Apples, the Monitor doesn’t look at all four of these identification bytes—just the first three. This is because Smartports and some hard disk controller cards don’t have a 60 at ADR+7. It would take several paragraphs to explain why —and we did so just a couple of months ago—so if you'd like to know more see “How many drives have we here?” in the November 1986 issue, page 2.79. (And puta note on that article to refer to this one for more information about counting the number of drives connected to a Smartport protocol converter.) The important thing to note in this discussion is that since Smartport protocol converters don’t have a $3C in that fourth ID byte, they won't automatically boot on startup when installed in older Apples. What Smartport protocol converters do have in that fourth byte is a zero. Thus, the complete “signature” of a Smartport-compatible protocol converter is: 49153 + (SLOT#256) = 32 49155 + (SLOT#256) = @ 49157 + (SLOT#256) = 3 49159 + (SLOT#256) = 0 You can do a slot scan that identifies Smartports like this: 100 REM * Slot scanner * 110 FOR SLOT = 1 107 120 : ADR=49152 + (SLOT*256) 130 : IF PEEK(ADR#1) < > 32 THEN 190 140 : IF PEEK(ADR+3) <> @ THEN 190 150 : IF PEEK(AOR+S) <> 3 THEN 198 160 : IF PEEK(ADR+7) <> @ THEN 190 170 : SMART(@)=SMART(@)+1 : REM count Smartparts found 180 : PRINT “Smartport found in slot “;SLOT 190 NEXT : PRINT 200 IF SMART(@)=0 THEN PRINT “No Smartports found.” : END ProDOS and Smartport. Like the Monitor, ProDOS also doesa slot scan. This happens as ProDOSis starting up. During this scan ProDOS attempts to identify all of the storage devices available on the computer and decides how itwill deal with them. Like newer Monitors, it looks at just the first three of the four ID bytes. Ifit finds a card that matches, it then looks at ADR+255. If this byte holds a zero, ProDOS assumes the card is a 16-sector floppy disk controller (and prepares to deal with it using its built-in “floppy driver’ software). A 255 in this byte would indicate a 13-sector floppy controller, if such a thing still existed (in this case ProDOS would ignore the device). Any other value indicates that routines ProDOS should use to deal with the device are stored on the card itself (in this case ProDOS assumes the entry point to these routines is at ADR + whatever was in ADR+255). ProDOS was specifically designed to deal with disks and other storage devices. In general, it's best to leave things such as buying Smartport Bus ‘sCn07 = $00 January 1987 tickets and putting blocks of data on the bus to ProDOS. It's much easier to type in CATALOG and get list of the files on the disk in the drive connected to the Smartport Bus than it is obtain the same information by driving the bus yourself. Evenso, ProDOS doesn’t knowa Smartport from a Dumbport. What todays ProDOS sees when it runs into a Smartport protocol converter is a ProDOS- compatible block storage device. ProDOS expects such devices to respond to just four commands — status, read, write, and format. In order to execute one of these commands, ProDOS stores the command number and other needed data in bytes $42-$47 (66-71) and calls the card's ProDOS entry point. You and Smartport. By going directly to Smartport yourself, however, there are a few things you can accomplish that ProDOS can’t. For example, ProDOS has no command for ejecting a disk from a 3.5 drive. Smartport does. Smartport protocol converters have a number of commands. Most commands can be sent to any device in the daisy chain. Most of the commands are useless for most programmers. For example: Init. This command resets the Smartport and all devices connected to it This command can't be made to a specific device, but only to the Smartport as a whole. Leave this command alone. Format. This call tells a disk drive, or any “block storage” device, to erase itself and prepare all blocks on the medium for reading and writing. Operating-system specific information, such as free block or free sector maps or blank directories are not a part of this call. A device that requires additional information to format itself is supposed to respond to this call by indicating that a successful format was completed without actually doing anything. | recommend you leave this command alone. Use the format command in the ProDOS Machine Language Interface if you really want to destroy... mean format,..something, ReadBlock, WriteBlock. These commands read or write a block of data on the storage device. They are essentially duplicates of the ProDOS read block and write block commands. ProDOS-based applications are better off using the ProDOS MLI read block and write block commands than these. Open, Read, Write, and Close. These are special Smartport commands for character-oriented (as opposed to block-oriented) devices, such as printers and modems, Since there aren't any devices like this currently available, there isn’t much you can do with these commands. One thing somebody could do, but hasn't, is read and write 3.5 inch Macintosh disks. These commands must be used for this because Macintosh disks have 524-byte blocks, rather than the ProDOS standard 512-byte blocks that ReadBlock and WriteBlock use. There is no technical reason why programs such as Apple's System Utilities or Copy II Plus couldn't copy files to and from Macintosh disks as easily as they do to and from DOS 3.3 disks. The software just hasn't been written yet. Interested parties can obtain more information from Apple's “UniDisk 3.5 Technote #4,” pages 4 to 7. There are two Smartport commands that are kind of fun. They are: Status. This command can be sent to the Smartport itself or to any device connected to it It comes in four flavors that retum various types of information about the Smartport or about the device connected to it. Control. This command is used to send special control commands to devices. The 3.5 disk eject command is a special version of this call. Atypical Smartport protocol converter call looks like this: JSR DISPATCH DA HMO DA PARMLIST BCS ERROR the DISPATCH address is $Cn00 + what's in SCrFF + 3 one-byte command number tuo-byte parameter List pointer Notice that the command number and parameter list pointer are embedded right in the assembly language code. This is similar to how a ProDOS MLI call is done. The Smartport will jump over these bytes when returning control to the main program. Thus, the next statement executed will be the Branch on Carry Set. Ifan error has occurred, the microprocessors cary bit will be set and the accumulator will hold the error number. If, on the other hand, there was no error, the carry bit will be clear and the accumulator will hold a zero. The Smartport entry point is always three bytes beyond the ProDOS entry point. We can have our slot scan routine from a few paragraphs ago find this this address, which Apple's documentation calls DISPATCH, by adding the following line: 173 : SMART(SLOT)=AOR + PEEK(ADR+255) + 3 : REM get DISPATCH address Open-Apple 2.91 There are two types of Smartport calls, standard and extended. All Smartport protocol converters support standard calls. At the moment, the only one that supports extended calls is the one built into the IIgs. After a quiet paragraph or two for those who care, we'll ignore extended calls. (Extended calls have four-byte memory and block pointers. Standard calls have two-byte memory and three-byte block pointers. Extended calls are useful primarily on the IIgs (and future machines) because they allow data to be saved from or loaded into any of the many banks of linear memory. (Interestingly, a four-byte block pointer, combined with standard ProDOS 512-byte blocks, allows a maximum volume size of over 2 trillion bytes. This is the equivalent of about 512K for every Apple II ever manufactured. It should be enough to last for quite some time. (You can tell if a Smartport supports extended commands by looking at byte $CnFB (ADR+251). If the high (seventh) bit of this byte is 1, then extended commands are supported. In addition, bit 1, when set to 1, indicates the Smartport is connected to a SCSI bus. Bit 0, when set to 1, indicates the Smartport controls a RAMcard. Thus: 175 B=PEEK(ADR+251) 176 IF B=>128 THEN B=B-128 : PRINT “Extended “; 177 IF 8=2 OR B=3 THEN PRINT “SCSI “; 178 IF +1 OR B=3 THEN PRINT “RANcard “; From Applesoft, we can make a Smartport call like this: 1100 REM * CALL SMARTPORT * 1110 POKE 768,32 : REM JSR DISPATCH 1120 POKE 770,SMART(SLOT)/256 1130 POKE 769,SMART (SLOT) -PEEK(770) #256 1140 POKE 771,CH0 1150 POKE 772,11 1160 POKE 773,3 1210 POKE 774,141 : REM STA 778 1220 POKE 775,10: REM error code address is $308 (778) 1230 POKE 776.3 1240 POKE 777,96 1250 CALL 768 1260 ERR=PEEK(778) 1270 RETURN After making a call to a Smartport protocol converter, we should always check to see whether an error occurred. This Applesoft Smartport caller stores the returned error code in byte 778. If this code is a zero, no error occurred. Ifit’s not zero, the following lines tell us what happened: : REM parmlist address is $308 (773) + REM RTS 1900 REM * Smartport error messages * 1901 IF ERR= 1 THEN PRINT “S01. No such conmand.” 1904 IF € THEN PRINT “$04. Bad parameter list.” 1906 IF € THEN PRINT “S06. Snartport Bus error.” 1917 IF € THEN PRINT “S11. This device can’t respond to this command.” 1931 IF THEN PRINT “SIF. Interrupts not supported by this Snartport.” 1933 IF € THEN PRINT “$21. This device can’t respond to this subconmand.” 1934 THEN PRINT THEN PRINT THEN PRINT THEN PRINT "$22. "$27. “$28. “$28. Bad Control List.” 1/0 error.” No such unit number.” Disk is urite protected.” THEN PRINT “$20. No such block number on device.” IF ERR=47 THEN PRINT “S2F. No disk in drive/device is offline.” IF ERR>47 AND ERR<64 THEN PRINT “$30-3F. Device-specific fatal error.” IF ERRDG3 AND ERRCBO THEN PRINT “$40-4F. The future has arrived.” IF ERRD79 THEN PRINT “$50-SF. Device-specific non-fatal error.” 1580 1999 RETURN Before using the Applesoft Smartport caller we have to set SLOT equal to the slot of the Smartport we want to call, CMD to the command number we want (0 for Status, 4 for Control), and we have to fill out a parameter list. The parameter lists for the Status and Control commands are exactly the same. They start with a one-byte parameter count (which should be 3 for both commands), then have a one-byte unit number (zero for the Smartport itself, 1 to 127 for devices), then have a two-byte pointer to a Status List or a Control List, and finally have a one-byte subcommand. From Applesoft, we can set this up with: 1000 REM * Set up paraneter List * 1010 POKE 773,3 1020 POKE 780,UNIT 1030 POKE 781,16 : REN Status/Control List address 1s $310 (784) 1040 POKE 782,3 105¢ POKE 783,SUBCHO 2.92 Open-Apple What buses stop here? To find out how many devices are connected toa ‘Smartport, you do a status call with a subcommand of zero. The number of devices will be returned in the first byte of the Status List. By adding the following commands to all the stuff that’s gone so far, you have a program that scans a computer's slots looking for Smartports and then tells you how many devices are connected to each one found: 300 UNIT=@ REM call goes to Smartport itself 310 CHO=@ —: REM Status call 320 SUBCND=0 : REM ue’re asking for the nunber of units 330 FOR SLOT=1 10 7 2 IF SMART(SLOT)=@ THEN 390 + GOSUB 1000 : IF ERR THEN PRINT “ERROR-~ +: NOEV(SLOT)=PEEK(784) 380 : PRINT “The Snartport in slot “;SLOT;” has “:NOEV(SLOT):” devices.” 390 NEXT : PRINT + GOSUB 190@ : GOTO 390 99 END (An additional bit of information you can pick up from this call comes back in the second byte of the Status List, byte 785 in this case. If bit six of this byte is clear, it means that a device on the Smartport chain is sending an interrupt. This can only happen on the Ic and will only happen after someone invents a Smartport Bus device that sends interrupts.) Ifyou're like most programmers, you'd like a bit more information about the devices connected to the bus than simply how many there are. A Status call to a device (rather than to the Smartport itself) with a subcommand of 0 will return four bytes. The bits of the first byte, which is called the Device Status byte, give various kinds of information that I'll show you how to decipher in a moment. After that come three bytes that indicate the number of blocks on the device. I can see absolutely no reason to ever use subcommand 0, however, because a Status call to a device with a subcommand of 3 returns a Device Information Block or DIB. This 25-byte block starts off with the same four bytes that subcommand 0 sends. In addition, after that you get a word, such as “RAMDISK” or “DISK 3.5,’ that describes the device (the length of the word is in the fifth byte, the device name itself in the next 16 bytes, padded with spaces at the end if necessary). After that there are device type and subtype bytes and a two-byte version number. From Applesoft it is always a mess to figure out which bits in a byte are ones and which are zeros. We'll have to do it here, however, so I have a magic subroutine for you. You tuck the byte you want to analyze into D, call the subroutine, and it sets the values B(7), B(6),...B(0) to one or zero depending on the value of the equivalent bit: 1500 REM * Turn a byte’s bits into an array * 1501 REN * —D=byte to analyze * 1502 REM * 8(7)=array with bit values = * 1510 FOR N=? TO @ STEP -1 1520 : BN=2‘N : B(N)=0 : IF D>BN-1 THEN D=D-BN : B(N)=1 1530 NEXT 1540 RETURN Now we have everything we need to find out everything that can be known about any device attached to a Smartport protocol converter. There's a lot of information here. Try this: 400 FOR SLOT=1 TO 7 410 : IF SWART(SLOT)=@ THEN 699 420 : FOR UNIT=1 TO NDEV(SLOT) 4302: 440 450 CMO=@ : SUBCHD=3 GOSUB 1000 IF ERR THEN PRINT “ERROR--"; : GOSUB 1900 : GOTO 698 460 470 DEVST=PEEK( 784) NBLKS=PEEK(785) + PEEK(785)*256 + PEEK(787)*65536 480 2 FOR I=1 TO PEEK(76B) : NAMES=NANES*CHRS(PEEK(788+1)) : NEXT £K(BOS) : SBTYPE=PEEK(805) £K(B07) + PEEK(B08) #256 510 : 2 PRINT “Device *;UNIT;” PRINT “ It has “3NBLKS: O=DEVST : GOSUB 1500 in slot “;SLOT;” is a “:NAMES blocks of storage capacity.” IF B(7)=@ THEN PRINT * IF B(7)=1 THEN PRINT ~ It is a character-orlented device.” sal It is a block-oriented device.” You can’t urite to this device.” Writes to this device are allowed.” 21 IF B(6)=0 THEN PRINT ~ zt IF (6)=1 THEN PRINT ” Vol. 2, No. 12 544 8(5)=@ THEN PRINT ~ You can’t read this device.” 545 8(5)=1 THEN PRINT * This device is readable.” 546 THEN PRINT ” This device is offline (disk ejected).” 547: THEN PRINT ” This device is online (disk inserted). 8(3)=0 THEN PRINT ” You can’t format this device.” 8(3)=1 THEN PRINT * Formatting this device is allowed.” 550 : : IF B(2)=@ THEN PRINT “ The device isn’t urite protected.” S81 8(2)=1 THEN PRINT * The disk is urite protected.” 852 B(1)=0 THEN PRINT * This device is not interrupting.” 553 B(1)=1 THEN PRINT * This device is the interrupt culprit.” 554 8(7)=1 THEN 560 : REM lines 555-6 are for character devices only 585 B(0)=@ THEN PRINT ~ This device is closed (available).” 556 : : IF B(@)=1 THEN PRINT ~ This device 1s open (busy).” 560: : PRINT ~ The TYPE byte of this device indicates it Is a “s IF TYPE=@ THEN PRINT “RAldisk.” IF TYPE=1 AND SBTYPE= @ THEN PRINT “UniDisk 3.5.” IF TYPE=1 AND SBTYPE=192 THEN PRINT “Apple 3.5. IF TYPE=2 THEN PRINT “hard disk.” IF TYPE=3 THEN PRINT “SCSI device.” S7a:: IF TYPE < 2 THEN 600 : REM see text about SUBTYPE byte S81: : PRINT ” The SUBTYPE byte of this device indicates that it: ~ S85 : : D-SBTYPE : GOSUB 1500 590 : : IF B(7}=1 THEN PRINT “ Supports extended Snartport calls.” 591: : IF B(6)=1 THEN PRINT “ Supports disk-suitched errors.” 592 : : IF B(5)=1 THEN PRINT“ Supports renovable media.” 600 : : PRINT * The version nunber of this device is “;VERS;”.” 698 NEXT : PRINT 699 NEXT More Smartport commands. In addition to Status subcommands 0 and 3, subcommands 1 and 2 have also been defined for the Smartport Status command. Subcomman4d 1 tells a device to retum its “device control block” or DCB. Since every Smartport device has its own unique format for this block it is difficult to say much about this subcommand without going into painful detail. Subcommand 2 is designed to allow character-oriented devices to retum their “newline” status. More on this when there is such a thing as a Smartport character-oriented device. The Smartport Control command has four device-independent subcom- mands. They are all pretty useless for most of us. Subcommand 0 resets the device to its default values. Subcommand 1 sends the device a new Device Control Block. Subcommand 2 sets the newline status. Subcommand 3 is for servicing an interrupt Ofsomewhat more interest are the device-dependentControl commands. The non-intelligent Apple 3.5 disk, for example, has seven of these, implemented as subcommands 4 though 10. The most useful is subcommand 4, which ejects disks from the drive. The others are mostly used for copy protection and deprotection, both of which are a waste of human resources, The UniDisk 3.5 implements four Smartport Control subcommands and one Status subcommand. Again, Control subcommand 4 ejects disks from the drive. Subcommands 5, 6, and 7 allow you to download programs into the drive and execute them. Again, this is mainly useful for copy protection and deprotection, however, it may be possible for some wizard to do something as magic as getting one UniDisk to automatically back up another without the host Apple II knowing anything about it. Notice I said “may.” The UniDisk 3.5 Status subcommanzd 5 allows an application to get more detail about read/write errors, including the contents of the registers of the UniDisk’s 6502 microprocessor. Just for fun, let's add a few lines to our program that will eject the disk from any Apple 3.5 or UniDisk 3.5 drive that is found: 610 620 630 640 650: : IF TYPE <> 1 THEN 698 : REM all 3.5 drives (so far) have TYPE=1 CND=4 : SUBCHD=4 POKE 784,0 : POKE 785,0 : REM H of items in Control List Gosus 1000 IF ERR THEN PRINT “ERROR--"3 : GOsuB 1900 : GOTO 698 Line 630 is quite important. Just as the Status command passes back a Status List, the Control command requires that you pass it a Control List. The first two bytes of the Control List are supposed to indicate the number of items the Control List has. In the case of Eject, there are none. So you must put zeros in the first two bytes of the Control List or strange things happen. January 1987 Ask (or tell) Last month indicated that | couldn't boot from my Apple Memory Card (page 2.87). This month I disco- vered that my problem was that I didn’t read the manual. The card has to be initialized witha program such as FILER or Copy II Plus to make ita bootable device. This step may seem unnecessary, since the card will automatically format itself for ProDOS the first time you access it. The problem with auto- format is that it doesn’t put some necessary “boot code” into the RAMdisk’s block 0. The boot code loads and executes the filenamed PRODOS— without itnothing happens. Both FILERand Copy II Plus will install the necessary boot code in block 0. The Apple igs has created lots of mail and interest, although hardly anyone has actually been able to buy one. Robert Inwin in Federal Way, Washington writes that he paid for a ligs exactly one-half hour after dealers were allowed to display the machine on September 25, yet as of December 2 his dealer had not been able to deliver. From Ann Arbor, Michigan, Clyde Godfrey writes that his dealer is now quoting January/March delivery for the machine he ordered September 27. A few subscribers have gotten their hands on the machine, however, and this month we'll print a few of their letters and answer a few Ilgs questions. But first, a word from our conscience: Auxmem corrections and more I'l bet the author of your auxiliary memory tester (December 1986, page 2.83) writes programs while using 80-column mode. Ifyou run the program in 40- column mode it doesn’twork. I suggestyou add aline to the program to turn on 80-columns. P ‘ete Ross Wayne, Mich, Oh, embarrassment! Here | am, always insisting that programmers put everything back where they found it, and I left STORE80 and PAGE2 dangling. As you point out, this is no problem in 80-column mode, but in 40-column mode the program hangs when it tries to execute the PRINT in line 400. Worse yet, in trying to find the source of the bug I found yet another bug | added at the last minute last month. | was concemed that my original version of the program (which worked fine) would mess up any data stored in a RAMdisk in auxiliary memory, so I added some lines to save and restore the contents of the auxiliary memory bytes that the program over- writes. Because of the cheap way | did this, the program shows some auxiliary-slot cards have four times more memory than they really have. Here are some changes to the program that fix both bugs: 150 S8=PEEK(49176) : POKE 49153,0 160 P2=PEEK( 49180) : POKE 49237,0 DELETE 230 250 FOR BANK=127 TO 0 STEP-1 260 POKE 48267,BANK : REM $0073 270 POKE 6192,BANK : REM put bank number at $2000 280 NEXT 381 IF P2C128 THEN POKE 49236, 362 IF S6Q12B THEN POKE REM Fix PAGEL REM Fix STOREGO Open-Apple 2.93 Lines 150 and 160 now save the status of the PAGE1/PAGE2 and STORE40/STORE80 switches before messing with them; lines 381 and 382 restore the switches to their original status. With these changes the program will work in either 40-column or 80-column mode. The additions and deletions in lines 230 to 280 get the program to stop having quadruplets. Last month Dennis also worked up some notes for me on the differences between the “Apple II standard” memory cards that are currently available. | couldn't find the notes when | needed them, but I've found them since. This stuff is intrinsically interesting, so! thought I'd let everybody in on what Dennis discovered: “Applied Engineering's RamFactor comes with software that expands the AppleWorks 13 desktop and allows a few more than 5,000 database records and 5,000 word processor lines. The card has built-in RAMdisk support for Pascal 13, DOS 3.3, and ProDOS, Recognition of the RAMdisk by ProDOS and Pascal is automatic; DOS 3.3 requires an IN#n command to enable the card. The firmware allows for the RAMdisk to be divided into as many as nine partitions, each of which can use any of the supported operating systems. Software on disk allows placing DOS 3.3 into a partition that can be booted from the partition manager. PR#n calls up the partition manager. A 1 meg RamFactor can be partitioned into two 400K DOS5.3 drives. It connects into DOS 3.3 ata different location than the Apple Memory Card, so the two will operate together with DOS 3.3. Maximum memory on the card is 1 meg; it can be expanded past this but the cost is prohibitive at present. Supplied software allows running AppleWorks 13 on an Apple Il-Plus equipped with a RamFactor and a Videx-compatible 80-column card. The expanded desktop appears in the current partition as a file of type $00 named Known Bugs. The SUBTYPE byte returned in the Device Information Block is pretty strange. Apple’s documentation says that a 1 in bit 7 of this byte indicates the device supports extended Smartport calls; a 1 in bit 6 indicates the device supports “disk switched errors,’ whatever they are; anda in bit 5 indicates the device has non-removable media. However, few, if any, of the currently available devices actually follow this protocol. That's the reason for line 580. Also see lines 571 and 572. Apple's documentation lists only specific devices in relation to the TYPE byte and doesn’t claim there is any pattern; however, 0 appears to be reserved for RAMdisks, 1 for Smartport Bus 3.5 disks, 2 for Smartport Bus hard disks, and 3 for SCSI devices. Ask me again in ten years and we'll see how well I guessed this one. We've known for some time that there is a bug in the ProDOS/UniDisk 3.5 relationship that causes bad things to happen when you write protect 3.5 inch disks (see January 1986, page 98 and February 1986, page 2.7). Here's the bug. Before doing a write to a disk, ProDOS makes a status call. Itwants to make sure there is a disk in the drive and that the disk isn’t write- protected. If the status comes back ok, ProDOS messes around a bit and then does the write. Remember that there is a difference between a ProDOS status call and a Smartportstatus call. When ProDOS does a status call, it puts a bunch of stuff in bytes $42-47 and calls the ProDOS entry point. ProDOS expects the device to send back a $27 (39) if an I/O error occurs, a $28 (40) if no device is connected, or a $2B (43) if the disk is write protected. If no error occurs, it figures a write-enabled disk is in the drive and ready to go. On the other hand, the Smartport Status comman4d tells you whether a disk is write-protected (lines 550-551) but it doesn’t return an errorif it is. Only a WriteBlock call will do that. The original UniDisk 3.5 Smartport-to- ProDOS routines forgot to dig the write-protect situation out of the Device Status byte and send ProDOS an error if write-protect was on. Consequently, ProDOS didn’t find out a disk was write-protected until it actually tried to write on the disk—and by then it was too late to prevent ProDOS from becoming quite confused. This situation was fixed on the IIgs by adding a few bytes to the Smartport- to-ProDOS routines that turn the write-protected bitin the Device Status byte into a real error for ProDOS status calls. You can figure out whether your Smartport controller has this fix or not by adding the following lines to our program: 660 PRINT 661 PRINT “Please open the urite-protect hole on this disk,” 662 INPUT “then reinsert it. Press <RETURN> uhen read 663 PRINT 670 POKE 66,0 : REM for ProO0S status command, put a zero at $42 671 S1=SLOT : IF UNIT>2 THEN S1=S1-4 : REM convert to ProDOS unit # 672 S1=S1*16 : IF UNIT=2 OR UNIT=4 THEN S1=S1+128 673 POKE 67,S1 : REM put ProDOS unit 4 at $43 (dsss0000) 675 POKE 809,32 + REM JSR DRIVER 676 POKE 811, (SMART(SLOT)-3)/256 677 POKE 810, (SMART(SLOT)~3)-PEEK (B11) #256 678 POKE 12,176 : POKE 813,2 + REM BCS 616 679 POKE 614,169 : POKE 815,0 + REM LOR 0 680 POKE 616,141 : POKE 617,52: POKE 818,3 : REM STA 820 681 POKE 819,96 2 REM RIS 682 CALL 809 684 IF PEEK(@20) <> 0 THEN 690 685 PRINT “This Seartport has the UniDisk 3.5 urite-protect bug.” 686 PRINT “ Eject this disk, cover the write-protect hole, and don’t” 687 PRINT “ever put @ urite-protected disk in this Srartport again.” 688 PRINT : INPUT “Press <RETURN> to continue. “:AS 689 GOTO 698 690 PRINT ~ You can safely use urite-protected disks on this Smartport.” Have I mentioned that Smartport calls require 30 to 35 bytes of stack space, can’t be used to transfer data into or out of the zero page, and, on the IIgs, must be made from emulation (standard Apple II) mode rather than native mode? Oh, and finally, ..don’t worry. I've left computers alone together for weeks ata time and not one of them has ever gotten pregnant. 2.94 Open-Apple SYS.DESKTOP. It takes possession of all available memory. Tom Vanderpool (ed note: a local wizard) and I figured out that you can save a dummy file of whatever size you want in the partition, then bring up AppleWorks. The desktop takes up the rest of the partition. From inside AppleWorks, you can delete the dummy file and use its memory for other files. Another possibility is to create a special partition just for SYS.DESKTOP, as it will not overwrite other Partitions. “CirTech’s Flipper (called ‘Flipster’ in the U.S.) supports ProDOS, Pascal 13, and up to two 400K DOS 3.3 volumes. CirTech’s version of CP/M Plus will automatically recognize and use the card. Supplied software allows up to four equal-sized partitions for support of Pascal LL. 1.2, and L3, ProDOS, DOS 3.3, and Microsoft Softcard II CP/M versions 2.20 and 2.23, and can create bootable RAMdisks for DOS, CP/ M, ProDOS, and Pascal. The partition manager loads into the Flipper itself. The partition sizes are not as flexible as those of RamFactor. Flipper version 103- L09 ROMs require a software patch to AppleWorks 13 to expand the desktop automatically as it is booted; version 2.00 eliminates this restriction. The Flipper software doesn’t expand the database or word processor, neither does it allow AppleWorks to munonall-Plus. “Apple‘s own Apple Memory Card includes RAMdisk support for ProDOS, Pascal 13, and DOS 3.3 (DOS is limited to a maximum of a single 400K volume). AppleWorks 13 will recognize the card and use its RAM for an expanded desktop but does not expand the database or word processor as does RamFactor. AST has a card called Sprintdisk that is similar to Apple's Memory Card. “Of the cards looked at here, RamFactor has the edge for users whose main concern is AppleWorks support; Flipper has the edge for those whose main concern is support of Apple Pascal or CP/M.” igs questions and observations Trecently received an Apple llgs, AppleColor Monitor, and 3.5 drive. It replaces an enhanced Ile, but not without some surprises. Having owned every variety of Apple II] computer to come out of Cupertino, I thought | could handle any problem, but your help would be appreciated. The ligs does not always act like a Super Serial Card, When I try to LIST an Applesoft program with PR#1 and LIST, the printer does not perform a carriage return until the end of the Applesoft line; that is, if the line is longer than the width of the carriage, it simply prints the rest of the line over the first part, from right to left. Is there a command to correct this (e.g. PRINT CHRS(9);"80N")? For now, | use the control panel to turn on echoing, which slows down the printing but corrects the problem. The clock/calendar works differently from the previous Thunderclock standard. My files under ProDOS 111 are not being time/date stamped; instead <NO DATE> is printed in the catalog, My ancient Disk II drives don’t work with the IIgs, though third-party 5.25 drives work fine. The 3.5 works beautifully, except you have to remember to eject the disk before you turn off the machine. The 800K disks are great for backing up Sider volumes, but the Sider is still faster. My Sider works just as well on the IIgs, but the connector is too large for the IIgs‘s back panel hole. | called First Class Peripherals, who told me the problem is being worked on. In the meantime, they said, | should simply feed the cable out the hole and let the plug dangle. Other than this, however, the Sider works flawlessly with the llgs. MultiScribe (version 11) does not work. | reada IIgs version is due out soon. The color monitor is beautiful, although | get color fringes on some programs and other color programs. don’talways look good on the monitor. Itreminds me of Apple RGB Monitor problems. The Control Panel is great for controlling the monitor display and adjusting the bell’s sound. The keyboard is comfortable, but its snap-together construction is a bit flimsy. Also, the cables connecting the keyboard to the computer on one side and the mouse on the other detracts from its detached nature. Richard Katz Los Angeles, Calif. My Super Serial Card acts just like your description of the ligs. To get it to stop overprinting you have to do a PRINT CHR$(9);"80N" to set a line width and a PRINT CHR$(9);"C" to get it to start sending a Return when the designated line width has been reached. This is, incidentally, different from the IIc, which Starts up with both those commands in effect. The IIgs clock does not appear “in a slot,” thus it can’t be accessed in the same manner as the Thun- derclock that ProDOS LL1 expects. ProDOS 12 contains the driver necessary to read the ligsinternal clock. You should have gotten a Ilgs System Disk with your machine, although it wasn't in the box with the first units shipped. Complain —the place you got your machine from can get you a copy. It has ProDOS 12 on it along with other goodies. See the next letter for how to get it onto your applications disks. We know of people using Disk I drives on a ligs with the Disk II controller card installed in slot 6 (remember to go to the Control Panel and select “your card” for slot 6). We know of no theoretical reason that would prevent Disk II drives, with the proper cable adapter, from working when plugged directly into the end of the Ilgs Smartport chain, but we haven't actually seen that done. You don't really have to eject the 3.5 disk before turning the machine off if you're willing to store the disk inside the drive until the next time you turn the computer on. (The problem is that you can't get it out of the drive with the power turned off.) This is how Apple's 3.5 drives work no matter what machine they're connected to. Multiscribe isn’t the only word processor that doesn’t work on the Iigs. Incredibly, Apple's own Apple Writer has problems. The ProDOS version won't print through the Ilgs serial ports, although it will print correctly through a serial card installed ina ligs slot. The earlier “Ile” DOS 3.3 version won't automatically load the PRT.SYS and TAB.SYS files. The “2.0” DOS 3.3 version won't boot. No updates have been announced. More llgs observations The difference in cost between the IIgs with mouse and system disk ($999) and the IIgs upgrade ($499 + mouse + system disk) was not worth giving up my Ile. Apple’s $250 rebate on an Imagewriter Il made investing in a second system even more reasonable. Although the igs Owner's Guide said my dealer could upgrade ProDOS on my application disks to version 12, my dealer didn’t know the procedure. What you have to do.is copy the file /SYSTEM.DISK/ SYSTEM/P8 to the main directory of your application disk, then delete the old PRODOS file and change the Vol. 2, No. 12 name of P8 to PRODOS. Any copy utility could make the transfer, but I found the copy a file option on the IIgs Desktop program easy to use. Updating to ProDOS 1.2 lets me access the Control Panel, which is helpful especially when | need to change the RGB display from color to monochrome or back. (The Catalyst 3.0 Desktop is really blurred when the Iigsis configured for color, so being able to interrupt Catalyst to change the display to monochrome is Great. Then, when | use Catalyst to choose a program such as Beagle Graphics, | can change the display back to color.) Apple Pascal L3-based programs can bea nuisance on the IIgs. MECA’s Managing Your Money, for example, recognizes memory storage devices (disks or cards) only in slots 4, 5, and 6, and works best with two 3.5 drives and a memory expansion card. If I configure part of the IIgs memoryas a RAMdisk, the IIgs makes the RAMdisk appear to be in slot 5, drive 2,and moves my second 3.5 drive to slot 2, where Apple Pascal can't find it. To let both 3.5 drives appear in slot 5, 1 have to set the Ilgs RAMdisk size to zero and instead use my Ile’s Apple Memory Card in slot 4 or slot 6 of the Iigs. But putting it in slot 4 disables the desktop mouse, while putting it in slot 6 disables my 5.25 drives, Catalyst 3.0 will recognize the IIgs RAMdisk as being on line, but will not automatically load applica- tions into it. However, itwill load applications into an Apple Memory Card in any slot. Some applications recognize the old Apple mouse, but not the new Apple desktop mouse. | have to put my old Apple mouse in slot 4, and configure the Control Panel for "your card’ in slot 4, for applications such as Electronic Arts’ Music Construction Set. (Soon Deluxe Music Construction Set should be out and | can put my Apple mouse and Mockingboard back in my Ile.) AppleWorks 1.3 can access the IIgs RAMdisk, but the desktop doesn’t expand into it as it does into the Apple Memory Card RAMdisk. The price of the Apple IIgs Memory Expansion Card with 256K ($129) is lower than most of the third-party cards, | saved even more by ordering more sets of 256K chips from other sources and installing them myself. Sony speakers designed for the Walkman work fine with the Ilgs and are less than half the price of the Bose Roommate speakers that Apple recot ds. mmen¢s. Sam L. Pemberton, Jr. Gallup, NM. As you point out, the two main problems with running ProDOS LL1 ona Ilgsare that the files won't be date stamped, because ProDOS 1.11 can't read the Iigs clock, and that ProDOS LL1 disables inter- mupts, which means you can't get to the Control Panel with the usual three-key control/open-apple/ escape command. ProDOS 12 fixes both of these problems, (Another route to the Control Panel, surer but more disruptive, is control/option/reset.) The file called PRODOS in the main, or root, directory of the ligs System Disk will confuse lots of people, It's not really ProDOS at all, but a little program that runs when the disk is started up. It, in turn, runs P16, which is in the subdirectory called SYSTEM and is the real ProDOS 16. ProDOS 16 looks fora file called START in the SYSTEM subdirectory. If it can’t find START, it runs the first root directory program with a name ending in “.SYS16". However, if there is a system program with a name ending in “SYSTEM” before the first SYS16 program, then ProDOS 16 gives up and runs SYSTEM/P8, which is January 1987 really ProDOS 8. ProDOS 8 then runs the program with the name ending in SYSTEM. What this all means is that the technique you outline for moving P8 and renaming it PRODOS makes 8-bit applications boot much faster than they do if you run them from a ProDOS 16 system disk. ProDOS 12 (which is the current version of ‘ProDOS 8" or “P8") comes complete with the track 0 bug discussed at length in our November issue (page 2.74)—Apple’s developer technical support people have indicated to me that this bug will be fixed in the next releases of ProDOS 8 and ProDOS 16. To adapt November's program so that it works with ProDOS 12, copy P8 and rename it PRODOS. Then make the following changes: 200 FOR 22723 10 22732 rest of Line unchanged 300 ADR=20996 rest of Line unchanged 550 PRINT “This version of ProD0S not 1.2.” There may be several release dates of ProDOS 12 floating around. The official version bears the date 06-SEP-86. AppleWorks addendum Ithink your suggestion that Lawrence Pratt (Novem- ber 1986, page 2.77) change from General Manager to AppleWorks is a good one, but with one caution. Anyone who tries to use AppleWorks without added memory and without AutoWorks will never know the true pleasure and power of AppleWorks. Added memory allows AppleWorks to be completely loaded into memory and eliminates disk access. Those of us who use AppleWorks this way soon forget the awk- wardness and slowness of the standard, disk-based AppleWorks. Autoworks adds such power to AppleWorks as to make it almost a brand new program. In suggesting AppleWorks to new users, | am always careful to add the suggestion for memory and AutoWorksas well. When I use Copy II Plus to rename AppleWorks files, | use all upper case (which is the only way Copy II Plus will allow), When the files are loaded into Apple- Works, however, the names are a random mix of upper and lower case letters. AppleWorks loads the files just fine and the files are all intact, but the first time one looks at the AppleWorks directory, itappears a disaster has struck. By the way, | know a number of AppleWorks SIG's around the country are cataloging all the disks from ‘TAWUG and setting up data bases with descriptions of the files in the disks. We are doing the same, and if some of the SIGs would like to share their work so that we don’t all have to re-invent the wheel, please have them contact me. Thomas E. Militello, M.D. Chairman, The AppleWorks S.1.G. Original Apple Corps Los Angeles, Calif. (213) 541-2766 ProDOS officially allows only upper case letters in filenames. For the three AppleWorks filetypes, how- ever, the 16 bits in a pair of bytes that are part of every file's directory entry (the AUX_TYPE bytes) were defined as indicators of which characters of the file's name are upper case and which lower. In addition, “lowercase” periods are defined as space characters. This is a very creative use of AUX_TYPE and one of the many fine touches of AppleWorks. When you catalog a disk of AppleWorks files with Basic.system, however, the filenames will appear in upper case, with periods instead of spaces, just like all other ProDOS filenames. This is because Basic.system doesn’t know anything about Apple- Work’s special AUX_TYPE protocol. From inside Apple- Works, however, the names appear in upper and lower case and include spaces as well as periods. When you change a file's name with Copy II Plus or Basic.system, you change the letters in the name but not the AUX_TYPE byte. Consequently, you get the previous mix of upper and lower case laid over the top of your new name —and the strange results you encountered. Versions of Copy II Plus beginning with 7.1 will correctly handle the AUX_TYPE bytes on AppleWorks files. Another escape route Asharp AppleWorks idea I read about a while back isto makea custom printer and set one of the printer options to just ESC. Ifyou make the bold on and bold off codes “ESC”, for example, you can do a control-B followed by your printer's ASCII command code and the printer will see it as a printer command. Inside AppleWorks your text might look something like: The “K2APPLE IIgs"KO is a ‘K3great’XO computer! When you print this you can get special options that Appleworks doesn’t normally allow, such as color on the ImageWriter II printer. Mike Mitchelson_ Overland Park, Kans. Single sheet sticky space When | print on single-sheet letterhead from Apple- Works, my printer beeps and locks everything up because it “runs out of paper” while advancing the single sheet out of my printer. | used to insert something behind the platen to trigger the paper sensor so that it would let my Apple go. But my printer also supports an “ignore paper error” command. The only trick was getting AppleWorks to send the code. | decided to put the code in an unused characters-per-inch setting. My idea was to switch to 24 characters-per-inch at the beginning ofa letter, send a blank line, then switch to the character- per-inch setting I really wanted to use. ButAppleWorks wouldn't send the code for 24 characters-per-inch (really the ignore-paper-error code, of course) unless I putsomething on the blank line besides spaces or a carriage return. Since anything else would get printed on my letter, this was a significant problem. | finally discovered that a “sticky-space” (created with open-apple-spacebar) was the solution I needed. A sticky-space won't print anything, yet it forces AppleWorks to send the character-per-inch code. Other people trying to use a custom printer's character- per-inch codes for unsupported printer features may find the sticky-space helpful. Paul Lucas Levittown, NY. Keyboard problems explained David Grigg’s problem with spurious characters in AppleWorks (October 1986, page 2.69) is a hardware problem all right —andit's not limited to AppleWorks, or even to his IIc. Dave Cortesi discovered a similar problem a fewyears ago ona different computer, and reported on the problem in his Dr. Dobb‘s Journal column. The spurious characters appear because of the way that the Apple keyboard's switches are networked. It's like a grid of horizontal and vertical wires, which each key atan intersection. The grid looks something like this: Open-Apple 2.95 21234567869 e is escape tQWwERTYUIO tis tab AOSHFIJGK;L c is return ZXCVBAN,./ x is delete \ se- s is spacebar Pt] u is up arrow cous’ dis doun arrow x dir ris right arrow 1 is left arrow The system works fine for detecting one key at a time and it successfully filters out most cases where more than one key is pressed at a time. But when three keys pressed form three comers ofa rectangle on the grid, the key at the fourth corner of the rectangle will also appear to have been pressed. For example, hold down the H, E, and R keys ona lle or IIc keyboard and you'll get a spurious F on the screen. If you have a Laser 128 or a llgs you'll have the same problem, buta different grid layout. Soit'sa hardware problem —and the only answeris to make the human at the keyboard adapt Lift your fingers when you type. Frank Hayes 1. Computing San Francisco, Calif. Softdisk lives If Softalk “died” then does that mean Softdisk has “died” with it? R.Roehm Los Alamitos, Calif. Softdisk, a monthly Apple !I magazine-on-disk, is alive and well. You can contact the company at PO Box 50008, Shreveport, LA 77130. Prices are $49.95 fora6-month subscription, or $89.95 fora 12-month subscription. Each monthly issue includes two disks. A similar publication is UpTime (174 Belleview Avenue, Newport, R.I. 02840). The cost for UpTime is $66 for 12 months (1 disk/month). Readers beware Both of the methods for embedding machine language code in Applesoft programs mentioned in the September 1986 issue of Open-Apple ("Another place for code” page 2.63) are in danger from certain utilities. Obviously, a REM stripper will wreck havoc with machine language code inserted in REM state- ments. And most RENUMBER utilities will reset the end-of-program pointer back to where it “belongs,” thus deleting any code appended to the end of an Applesoft program. Regarding "How many drives have we here? (Novem- ber 1986, page 2.80): You can tell the difference between no drive and no disk with the UniDisk 3.5 (and Apple 3.5 on a IIGS). The 3.5 inch drives retum the offline error ($2F) with no disk inserted. Even though this is not a defined error under ProDOS, the Kernal passes it on to Basic.System. Unfortunately, it is not pigeon-holed in a global page location and it is converted by Basic.system into the umbrella I/O ERROR. If you're willing to search for it, the ProDOS error number can be found in a scratchpad location —in the Iicat $4F8; in the Ile at $4F8 + the number of the slot the interface card is in; and in the IIgs at $EQOFB1 In your answer to that same letter, you state that third and fourth drives hooked to one controller are “supposed” to appear to be in slot 1 or 2, but Apple has never suggested that higher slot numbers can’t be used. However, the Smartport interface will auto- 2.96 Open-Apple matically access the third and fourth devices in one of its daisy-chains only if they have been assigned a phantom slot lower than four. ‘Speaking, finally, of four drives, you can easily patch FID, as came up in October (page 2.71), to handle up tonine DOS 3.3 drives. Here's howyou do it: ORIVES = 4 BLOAD FIO POKE 2419,176 + ORIVES + 1 POKE 2495,176 + DRIVES + 1 IF DRIVES > 7 THEN POKE 2431,15:POKE 2507,15 SAVE F10,ayzost.L4se7 Tom Vier Reston, Va. Another restart trick Here's another trick like the one for restarting AppleWorks that you published in June (page 2.33). If you crash into the Monitor from a ProDOS system file, there is usually no way to restart except by rebooting. Basicsystem supports the old 3D0G warmstart vector, but few other system programs do. However, ifyou reboot ProDOS, itwill erase any files you have stored on the /RAM volume. Instead of rebooting, enter the following, It will execute a ProDOS quit call. Ifyou have a program selector such as ProSel, it will run automatically. Otherwise, you will at least get the ProDOS quit prompt. 's wrten, edited, published, and © Copyright 1986 by Tom Weishaar Business Consultant Richard Barger Technical Consultant Dennis Doms Circulation Manager _ Sally Tally Most nghs reserved All rograms publshed n Open-Apple ate pubic daman and may be coped and dstrbued wihout charge {rast are avalable nthe MAUG Inraryon CompuServe) Appe use Youps and sghteantahers may obtan permsson repent aris fom tme to tme by specie witen request, Request and her etal matealnuding eters oUree 00S, shoud be sent. ‘Apple P.O. Box 7651 Overland Park, Kansas 66207 U.S.A. ISSN 0885-4017. Published monthly since January 1985. World-wide prices (i US. dolars. armal deivery included! at no. addtional Charge} $24for 1 year. $44t0r 2 years SB0{or 3 years Alback ssves 4atecurenty avaliable ior $2 each. abound, ndexed edtion ol Volume 111s $1495. Index mailed wih the February issue. Please send all subscuiption-related correspondence to: Open-Apple P.O. Box 6331 Syracuse, N.Y. 13217 U.S.A. Subscnbers in Australia and New Zealand should send subscripion corespondence to Open-appe cl Cybernetic eseneh i576 Malvern Foad, Pranan Vc B18). AUSTRALIA ‘Open-Appie's available on disk or speech synthesizer users trom ‘Speech Enterprises. PO. Box 7986, Houston, Texas 77270 (713-461. 16685) Unlike most commerical software, Open-Apple is sold in an Unprotected format for your convenience You ave encourgaged to ‘make back-up archival copies or easy-to-read enlarged copies for your own use without charge. You may also copy Open-Apple for ‘istnbution to others. The distnbuton fee is 15 cents per page per Sop dibuted WARRANTY AND LIMITATION OF LIABILITY. | warrant that most of the information in Open-Apple is uselu and correct, although drivel and mistakes are included trom time to tme, usually unintenvonally Unsatsiied subscribers may return issues within 180 days of delivery fora full elund Please include a nate fom your parents or chidren Conkeming natal archivalcopeeshave been destoyed The ntulileg onion of any paid subscription wil be refunded on request. MY LIABILTY FOR ERRORS AND OMISSIONS IS LIMITED TO THIS PUBLICATIONS PURCHASE PRICE. In no case shall | or my Contributors be lable for any incidental or consequential damages, or for any damages in excess ol the fees paid by a subscriber ‘Open-Apple's neither atiated with nor responsible forthe debts of Apple Computer. Inc. tmaja questing’ 1s a trademark of Don Lancaster Source Mail: TCF238 ‘CompuServe: 70120,202 800:20 00 GF G5 07 08 00 04 00 20 00 00 00 00 N Boag The source code looks like this: 20 00 BF JSR SBF@ =; call MLT 65 0B $65 3 MLI QUIT call e708 OW } pointer to parm table cr) RK 3 BRK if ProDOS won't quit 4 0B $04 3 pare table - 4 paras ee 0B $00 3 all 4 parameters are $0’s ee e2 = ON «$0000 ee 0B $00 00 Ou $00 Ihave found this trick very helpful while program- ming in Kyan Pascal. dim Luther Kansas City, Mo. Users battle tech support If you charged for your newsletter according to what the information was worth, I'd have to cancel all of my other Apple magazine subscriptions, pawn my wife and kids, and maybe even put off buying a new Ilgs. Once more, your November issue proved that Open-Apple is an absolutely indispensable source of solid advice. Towit, | had been experiencing track 0 disk crashes when using ProDOS on a floppy drive for nearly six months. The problem began when we purchased a new Ile system at work, complete with an Applied Engineering TransWarp card. After purchasing two surge protectors, replacing the power supply with a heavy-duty unit, swapping out all peripheral cards, exchanging disk drives, and purchasing an expensive disk drive analysis program, | felt | had traced the problem to the TransWarp card. The disk erasures only occurred when using a ProDOS-based program on one of the floppy drives and only when the TransWarp card was active. The problem never occurred under DOS 3.3, never involved the UniDisk 3.5 drive, never affected the IIc system plugged into the same power strip, and never happened when the TransWarp card was either removed from the computer or deactivated. With the TransWarp activated, the problem occurred approximately every third disk access. Furthermore, the problem occurred with varying frequency in any of three Ile systems that the TransWarp card was plugged into. Pretty convincing evidence implicating the ‘TransWarp card, wouldn't you agree? Applied Engineering didn’t agree. I exchanged the card twice with them, wrote detailed accounts of my bug-hunting treks, and paid for three lengthy phone calls to their technical assistance line in Texas. On each occasion, | was told that the problem could not bereplicated and that the problem was unequivocally a “user error.” When I mentioned that your newsletter had been carrying on-going descriptions of the problem since January, their terse response was that Applied Engineering receives hundreds of calls each day so if there was any problem they would know about it. In mid-November, | called them once more to read select portions of your article "ProDOS bug found in Australia” (November 1986, page 2.73). The techni- cian’s reply was that the article was simply wrong and that the problem could only be user-related. | thanked him for his input and requested the return of my ‘TransWarp card. After reinstalling the TransWarp afew days later, | verified that the track 0 problem was still occurring and then made the ProDOS patches outlined inyour newsletter. Voila! We have not losta single data disk since. Vol. 2, No. 12 What | have leamed from this experience is that some companies work under the assumption that their products are bug-free but that their customers are idiots. Perhaps Applied Engineering should stop acting like it's God's gift to the Apple user and spenda little time debugging their "public interface.” Thanks for your fabulous publication. And, most especially, thanks for communicating with your readers from a position of respect and appreciation. I don’t understand some of your articles the first time I read them, but I have never felt intimidated by your descriptions nor belittled by your language. Maybe that's why I genuinely enjoy re-reading the articles time and again. Pat Mamett Tempe, Ariz. Open-Apple is always happy to play the role of the hero, but I think we should point out that the credit here goes to other people. SubscriberJ. Emest Cooper was the one who convinced me the problem was more than “user error” and who got me to report the problem in Open-Apple. Stephen Thomas was the Australian programmer who listened carefully to his customers’ feedback and who consequently found the ProDOS bug. Subscriber David Grigg made the connection between what Cooper had reported and what Thomas had written for an Australian user group newsletter, and who wrote us about it. This kind of world-wide information exchange is the core around which Open-Apple is growing. Having been on both ends of technical support lines, I think you're being too harsh with Applied Engineering. As it turned out, after all, fixing ProDOS, not the TransWarp, is what solved your problem. On the other hand, any company that gets hundreds of support calls a day should have gotten enough reports of track 0 crashes by now to suspect some- thing was amiss. Technical support people have got to be more cautious with exotic problems. It's easy to blame bug reports that you have no solution for on “user errors” and forget them. For example, my own brother-in-law called me this week to ask why so many of his AppleWorks data disks were being destroyed. When | told him I had a ficfor him, he said, “That's funny, last year you told me it was my fault.” don’t even remember talking to him about crashed data disks before. How come, by the way, we're not laying any blame on Apple's shoulders? It's Apple's own operating system that has been crashing all these data disks after all. However, Apple's technical support group has a world-wide network of dealers that ensure it will never hear about, much less find, any bug that has even the slightest chance of being blamed on “user error.” Formore perspective on the battle between troubled users and defensive technical support see the note from Call -A.P.P.L.E.'s Charlie Stillman on page 55 of the November 1986 issue of that magazine. “Pinpoint was driven to the edge by Dennis LaMonica’s report of a serious bug in Pinpoint's Graphmerge accessory in the July ‘Write A.PPL.E. Outrage, hurt, and denial eventually gave way to a patch to fix the problem....The Pinpoint technical staff is courteous, responsive, and effective....But Pinpoint almost wrote off Dennis LaMonica as a trouble maker. Publishers need to recognize the LaMonica’s of the world as their best allies. For each LaMonica who takes the trouble to report a bug, there are ten other people who quietly file the software in a drawer and spread the word that it's unreliable. Keep up the good work Pinpoint and LaMonica! You are on the same team!”