Open Apple 1987 02 February 1987 Vol 3 No 1

Open-Apple Releasing the power to everyone. February 1987 Vol. 3, No. 1 ISSN 0885-4017 newstand price: $2.00 photocopy charge per page: $0.15 Apple introduces an updated Ile ‘The Apple Ile now features an expanded keyboard with a numeric keypad, a platinum color scheme, and 128K of RAM memory, Apple announced in mid-January. The new model is otherwise little changed from its predecessor —even the price remains the same at $829, suggested retail. The new keyboard has the same layout as the Apple IIgs keyboard, but the heavier feel of the older Ile keyboard. The reset key has been moved above the ESC and 1 keys and the power light is now above the slash key on the numeric keypad. The numeric keypad’s CLEAR key, which retums a control-X on the Ilgs, is a duplicate ESC key on the new Ile. By means of a hardware- jumper cut-and-paste operation, you can make it match the IIgs, however. In addition to the visible changes on the exterior of the computer, the machine uses a revised motherboard that Apple has actually been puttingin new Iles since October. The new board includes 64K of RAM in just two chips instead of the previous eight and uses just one ROM chip instead of the previous two. The contents of the ROM are unchanged. The motherboard has also been revised slightly so that an additional 64K of memory can be placed on a very small card that is inserted into the auxiliary slot during manufacturing, Thus, all units now come with 128K of memory. The internal numeric keypad connector was retained. The shift-key mod. In order to allow mouse-based software to detect a “shift-click” (simultaneous presses on a shift key and the mouse button), the new motherboard comes with the “shift-key modification” already wired. This modification originated with the Apple II-Plus, which dosen't support lower-case characters. The II-Plus shift keys affect only the dual-character number and symbol keys. In order to make the shift keys functional with letters too, a scheme was devised whereby users could wire the shift key to the third pushbutton input on the game connector. By monitoring that “push-button,” word processors and other programs can determine whether the II-Plus shift key is being pressed. When the original Apple Ile was released, jumper pads were provided at motherboard location X6 to make this an easy modification. Now bow-ties are provided to make it easy to disconnect. Problems can arise because devices that actually use the internal game port's third pushbutton input will cause shift-key-modified machines to go into a permanent shifted state unless the modification is removed. In addition, simultaneously pressing the shift key and the third pushbutton input will cause a short circuit on shift- key-modified machines and shut down the Apple Ile power supply. Since the open-apple and solid-apple/option keys are wired to the game connectors first and second pushbutton inputs, this chart and program show you how to read all three keys: decimal -decimal hex description 49249 49250 49251 -16287 -18286 -16285 C061 sc062 50063 P80, open-apple PBL, sol id-option PB2, shift-key mod 10 IF PEEK(49249)>127 THEN PRINT “You're pressing the open-apple key.” 30 IF PEEK(49250)2127 THEN PRINT “You're pressing the solid-option key. 40 IF PEEK(49251)2127 THEN PRINT “You're pressing the shift key.” 50 GOTO 10 Ifthis program says you're pressing the shift key when you're not, then the “shift-key mod” is disconnected. This is always the case on the IIgs, which has another means, known as the “modifier-key status register,” of telling you not only the status of shift, but also of control, caps lock, option, and open-apple, as well as whether any key on the numeric keypad is down, whether any of the ASCII-code keys is down, or whether the keyboard is repeating, This register is at byte 49189 ($C025). Here's an Applesoft program that shows howto useit, with thanks to alittle subroutine explained last month on page 2.92: 10 D=PEEK(49189) : GOSUB 1510 : HOME 20 IF B(7)=1 THEN PRINT “The open-apple key is pressed.” 30 IF 8(6)=1 THEN PRINT “The option key is pressed.” 40 IF 8(5)=@ THEN PRINT “An ASCII-code key is pressed.” 50 IF 8(4)=1 THEN PRINT “A nuneric keypad key is pressed.” 60 IF 8(3)=1 THEN PRINT “The keyboard is repeating.” 70 IF 8(2)=1 THEN PRINT “The caps lock key is pressed.” 80 IF 8(1)=1 THEN PRINT “The control key is pressed.” 90 IF 8(@)=1 THEN PRINT “A shift key is pressed.” 99 POKE 49168,0 : GOTO 10 : REM clear keyboard strobe 1510 FOR N=? TO 0 STEP -1 1520 : BN=2"N : B(N)=@ : 1530 NEXT : RETURN IF DDBN-1 THEN D=0-BN : B(N)=1 The reason a program might want to knowifa numeric keypad key is being pressed is so that it can tell between, for example, the keypad’s slash key and the main keyboard's slash key. All of the keypad’s keys have main keyboard duplicates. Also notice that bit 5 is meaningful when it has been cleared to zero, while the others are meaningful when they're set to one. Maybe even more exciting than a new Ile to many of you is Beagle Bros’ new ProDOS-based Applesoft compiler. The Beagle Compiler, by Alan Bird, is the first ProDOS-based Applesoft compiler we've come across ($74.95, 3990 Old Town Ave, San Diego CA 92110 619-296-6400). Acompiler does to your Applesoft program what a human translator does to documents written in languages you don’t understand. As far as your Apple is concemed, a program written in Applesoft is in a foreign language. The program has to be translated before your Apple can execute it. Normally, Applesoft itself does the translation as the program is running, every time you run the program. Fenwa “HOLD ON, THATS NOT A PROGRAM ERROR, ITS JUST A BOOGER ON THE SCREEN.” 3.2. Open-Apple A compiler translates an Applesoft program into something your Apple can understand one time. No further translation is necessary. Consequently, the program will execute faster. How much faster depends on the program. String-intensive programs can run as much as 5 to 15 times faster than normal (compare this to the 2.5 to 3 times boost you get from a hardware accelerator). Programs that consist primarily of floating point calculations, on the other hand, don’t run much faster at all. In addition to being ProDOS-based, Beagle’s Applesoft compiler has several other important features—compilation is fast, compilation is simple, and compiled programs are shorter than the original. Ampersand routines without parameters will compile without problems and Beagle’s manual explains how to modify ampersand routines that do pass parameters. The Beagle Compiler works with any 64K or larger Apple II. It's an order of magnitude better than any other Applesoft compiler I've ever seen. Apple released an updated ProDOS system disk to developers in mid- January. It contains version 11 of ProDOS 16 and version 13 of ProDOS 8. Both are dated 2-DEC-86. Among other changes, the newversions fix the bug in the ProDOS floppy driver discussed here in October. Apple's newSCSI hard disk controller card has some problems running on the llgs. Apple is recommending that users back up their disk often while its engineers work out the bug. Bob Shofstall at Nite Owi Productions, who has a ‘SCSI drive on a IIgs, says he can hardly keep it running long enough to back itup. Apple has also discovered that at least one revision of its DuoDisk has one too many resistors to work correctly on a IIgs Smartport daisy chain. Tell your dealer to look on AppleLink for more information on this one if you're having problems — the DuoDisk can be made to work by cutting the extra resistor off the DuoDisk’s internal logic board. Two of the most frequently asked questions about the Ilgs at the momentare how to wire older Disk II-type drives so they can be connected to the Iigs Smartport daisy chain and how to wire non-Apple analog RGB monitors to the IIgs. We printed a diagram showing howto convert the 20-pin “header” connector used by older drives to the newer DB-19 connector in December 1985, page 90. It's all very straightforward — the only twistis in the drive enable lines. Pin 17 on the DB-19 carries the drive enable signal for drive 1, pin 9 carries the signal for drive 2. In each case the drive expects its own enable signal on pin 14 of the header connector. I know far less about RGB cables, but here, at least, is the pinout of the Ilgs video connector: 7654321 151413 121110 9 pin layout of IIgs fenale video connector fron connector side (not solder side) 1 ground ground reference and supply 2 red red analog video signal 3 comp composite synch signal 4 NC. no connection S green green analog video signal & ground ground reference and supply 7 NC. ‘no connection 8 412v #12 volt supply 3 blue blue analog video signal N.C. no connection 11 sound analog sound output 12 NTSC/PAL composite video output 13 ground ground reference and supply 14 NAC. no connect Lon 1S NC. no connection Uncle DOS is anxious to hear from anyone who can shed more light on what kinds of RGB monitors will work with the Ilgs and how they should be connected. Apple Writer guru Don Lancaster has already come up with a patch that allows Apple Writer to work on the IIgs. As mentioned last month, ProDOS- based Apple Writer won't print through the IIgs serial ports. Lancaster's patches make Apple Writer realize that what appears to be a Super Serial Cardin the Ilgs isn’t really a Super Serial Card at all. After the patches, Apple Writer treats the ligs as if it had a third-party interface card (and treats Super Serial Cards themselves as if they were third-party cards, too). This means, among other things, that the patch must also disable the control-O J “set modem/printer interface” option, and it does. The patch is for 128K ProDOS Apple Writer version 2.0 only: from the Monitor, enter BLOAD AND.SYS,AS2000,TSYS verify that 4DB@ is AQ, 4080360 verify that 4F6? is Ol, 4F67:10 verify that 4FGE is 31, then 4F6E:13 then then BSAVE AKD.SYS,AS2000,TSYS Vol. 3, No. 1 Back in November I mentioned (page 2.77) that I had never been able to get Apple Writer's “peek” command, which allows you to see what's in a file without actually loading it, to work. The manual says you can do this by acting as ifyou’re going to (L)oad the file, but then adding a backslash after the file's name. Every time I tried this | g Pathname.” Subscribers Robert Hall and Robe tums out that what the manual re whatever character you are using for filename. The underline token ap] Apple Writer's underline token, | had! couldn't get the file peeker to work, it the message “ProDOS error: Bad Flickenger have come to my rescue. It ly means is that you must append in underline token onto the end of the on the (P}rint menu. Since I don‘tuse it defined as nothing at all. No wonder I The Wall Street Journal is by far my favorite newspaper ($119/yr, 200 Burnett Road, Chicopee, MA 0102( financial journal. For example, | December 30, 1986: WhenJohn Sculley, chief executit First Boston conference on high te Apple had recently bought a Cray $14.5 million and was using it to det When it came time for John Rollwé the conference, Mr. Rollwagen said a one-way street. So, he said, “sinc our machines, some of us have bou Mr. Rollwagen also said he told about how Apple was using the mat end of the line,” Mr. Rollwagen interesting, because I'm designing Center St, Marion, OH 43306 80 magazine is that it limits itself to 16 week after week. From the December of an experiment by Diana Deutsch Diego that concerns pairs of tones, 800-841-8000). It's far more than a k at this story, which appeared on of Apple Computer Inc., addressed the hnology earlier this month, he said search Inc. supercomputer for about lop its next-generation Apple. jen, Cray's chief executive, to address hadn't wanted Apple to think this was they were good enough to buy one of ht a few of theirs.” ymour Cray, the company founder, ine. “There was a pause on the other recalls, “and Seymour said, ‘That's e next Cray with an Apple.” it is Science News ($29.50/yr, 231 W 247-2160). The best feature of this pages, front and back covers included, '20, 1986 issue (page 391) comes news of the University of California at San constructed from a set of sinusoidal waves, that are a half octave apart. Spme listeners hear the second tone as higher in pitch than the first, while dther listeners hear the second tone as lower in pitch. Furthermore, if the different speeds, a listener who hears may hear it as descending at anoth pattern and speed will clearly and descending. “In visual terms,” Science News sa) as seeing a square instead of a circl location.’ (Or as paradoxical as one sees a circle in the same locatio important to any of you trying to suggested here in March and Sey Apparently the effect is limited to ce} don’t work on, say, a piano keyboar« complex structure than the experim| organ) and listeners apparently use pitch.” I suspect there are some amai of sound and | suspect the Apple Il them in. Those of you interested in le tones are taped and played back at the pattern as ascending at one speed r. Yet, for any single individual, a given nsistently sound either ascending or ¥s, “this musical effect is as paradoxical when a shape is shifted to a different person seeing a square where another , | might add.) The effect would be ise sound as a data-analysis tool, as tember 1986 (pages 2.9 and 2.58). in types of tones. “The experiments because each piano note has a more ntal tones (which sound more like an lose extra clues to determine relative ing discoveries yet to made in the area is the perfect research ship to make jing more about writing Applesoft programs that use ProDOS should take a look at the new bovk ProDOS Inside and Out, by Open-Apple lical consultant Dennis Doms. Dennis borrowed a few chapters from my Spftalk and Open-Apple articles (that's how I got my name on the cover, too)} but most of it is new stuff. Unlike most other books you'll find on ProDOS,|ProDOS Inside and Out deals almost exclusively with Basic.system (in fact, Dennis and | wanted the book to be February 1987 called The Basic.system of Apple’s ProDOS, but changing the mind of a publishing company can be difficult). The book has five major parts, which consist of six introductory chapters, five chapters on Applesoft programming, five chapters on programming with text files, five chapters on assembly language programming, and four chapters on using the ProDOS machine language interface. The book is available (or can be ordered) from your favorite bookstore (Tab Books is the publisher), or you can get it from us at the suggested retail price of $16.95, postage paid. We also have a disk with the book's 40-plus programs for $10, which your bookstore will never have. Open-Apple: Vol 2—a bound, indexed edition of our February 1986 through January 1987 issues, is also now available, directly from us, for $14.95. Book orders are handled fastest if you send them to our Overland Park address (see back page) unless they also have something to do with your subscription (payment, change of address, missing issue, etc.), in which case itis always fastest to use our Syracuse address. Open-Apple 3.3 Open-Apple has a new apologist and business manager, Sally Dwyer. Among her many jobs is to make sure all of our customers have been satisfied with our service. Unlike our circulation manager Sally Tally, who is an imaginary being whose chores are done by a number of people (none of whom are actually employees of Open-Apple), Sally Dwyer is a real person with the patience ofa first-grade teacher. If you every have problemswith your subscription or with a product order, contact Sally at our Overland Park address and she'll see that you're satisfied. (A note to those of you who got an invoice or renewal slip this month— please don’t complain to Sally that you didn’t get an envelope to return your payment in. Enclosing both the index for last year and a return envelope would have made the newsletter too heavy to send without more postage. Returm envelopes will return next month.) Do we have any Stats Pro users out there? Subscriber Jean McCollom would like to trade tips on using the program with you (this is how special interest groups get started). You can reach McCollom at 813-657-2531 Ask (or tell) Uncle 1 ve always thought that writing assembly language programs was one of the best ways available to develop the virtue of humility. That's because no matter how hard you try nor how long you work there's always at least one more bug. Now I think writing a newsletter about computers may be even better for humility than programming. Look, I'm bursting out all over with humility. Grab your pencils and your back issues, folks. This month I'll start with a correction to last month's corrections. On page 2.93, the line that begins with 382 should say POKE 49152 instead of 49236. Then, on page 2.96, take those dollar signs out of the BSAVE FID line in the first column—those are decimal numbers. Back in October, on page 2.71, Isaid the mnemonic meaning of the program FISHEAD had been lost. Rich Williams, the Apple programmer who wrote FISHEAD and named it, collared me recently and said that the meaning isn’t lost at all—it stands for File Shower, HElper, And Duplicator. Next time I see him he'll probably tell me I spelled it wrong, but I'm not touching this fish story again. And that's not all folks... Smartport lines missing It appears that the dastardly and devilish denizen of DOS deliberations has done diddled with Uncle DOS'ss disk drive device disassemblies. On the very same page that you corrected the AUXMEM program bugs (January 1987, page 2.93), there appears to be something lacking in the "Known Bugs” paragraph (how apropos, a bug in the bug finder). First, in line 680 shouldn't it be POKE 817,52, not POKE 817,53? This is needed to agree with the PEEK(820) command. Secondly, after entering all those POKEs in lines 660 to 681, I could find nowhere in the code where the addresses were CALLed. Is perhaps a line such as “682 CALL 809" missing? Anyway, those are the changes that I made and the program tells me that my UniDisk 3.5 card is a baddy. If this is so, what can I do about it? | always write- protect disks that | don‘twant to lose, and I absolutely refuse to stop doing this, What a dilemma! Craig Peterson Santa Monica, Calif. Youare correct. Even though we tested that program extensively with line 680's POKE 817,52 fixed and a CALL 809 at line 682, those little fixes didn’t get into the file we actually typeset. You can still write-protect disks if you want to. Just don't attempt to save anything important on one of them. If you do, ProDOS will get quite confused and you might lose what you're working on. Often that isn‘tas important as losing the whole disk, however. It still makes sense to write-protect originals when making copies of disks, for example, so that you don't initialize the original. Smartport follow-ups | thoroughly enjoyed your article on Smartport in the January 1987 Open Apple. With regard to the number of devices that may be connected to the Smartport bus, Apple has set a limitation of four devices due to power supply and signal degradation effects. It may be possible to extend this limitation by designing Smartport devices that supply their own power and buffer all signals on the bus, but this hasn't been done yet. In addition to the limited capacity of the bus, the ProDOS entry point on Smartport protocol converters also has a maximum capacity of four devices. ProDOS calls to slot 5, drive 1 always go to bus unit 1; slot 5, drive 2 to unit 2; slot 2, drive 1 to unit 3; and slot 2, drive 2 to unit 4. ProDOS is being updated to use the Smartport directly, however, rather than using the Smartport's ProDOS entry point. The newversion will not have the four-device limit. An updated ProDOS, combined with self-powered and buffered bus-devices, would allow full use of the Smartport's capabilities. Ray Montagne ‘Smartport Software Engineer Apple Computer, Inc. Readingyour Smartport discussion was worse than eating salted peanuts—I had to have more. Some details cried out for attention. What, for example, does ProDOS do with a Smartport during the boot process? And how are device numbers assigned? Is there any way to detect a RAMdisk in the chain short of asking every device for its name? Finally, since ProDOS 1.2 knows enough to remap device numbers when more than two devices are found in the Smartport chain, you assertion that "ProDOS doesn’t know a Smartport from a Dumbport” seemed limited to the world of ProDOS 1.11. Just what does ProDOS 12 do during system initialization? With such questions in mind, I disassembled the ProDOS 1.2 initialization code and spent an evening going over it. What I found may be of interest to your readers, (Side note: the IIgs disassembler is of course more powerful than previous Apple II disassemblers, buta curious problem crept in: Since the 65816 has 256 legal opcodes, the venerable "???” has all but disappeared from disassemblies; hence a small data cache tacked onto the end ofa code module is much harder to spot. The ProDOS title page text, for example, disassembles as legal code.) Aswith ProDOS 1.11, ProDOS 1.2 does a slot scan as itbuilds the system global page. Immediately prior to beginning the scan, however, ProDOS 1.2 peeks at slot 2. The result is stashed away for later use, and, beginning with slot 7, the slots are checked in descending order for the presence of block devices. If ProDOS 1.2 finds a block device which is not a Disk II, however, it specifically checks for the presence of Smartport protocol converter firmware by testing the ID byte at $Cn07. Ifa Smartportis found, then its entry point is determined and a status call is made via a Cute little piece of self- modifying code. The status call gives ProDOS the number of devices accessible through the Smartport and the first two of these devices (if extant) are assigned device numbers corresponding to the slot in which the Smartport is found. Then an interesting thing happens—unless the Smartport is found in slot 5, ProDOS ignores any other devices in the chain. If the Smartportis indeed in slot 5, then the slot 2 status previously stashed away is exhumed and checked; ifa block device exists in slot 2, then ProDOS washes its hands of slot 5. Otherwise, the third and fourth devices in slot 5 (if extant) are given device numbers corresponding to slot 2 with device driver addresses in slot 5. ProDOS then moves on to slot 4. So what does all this mean? As | see it, there are several conclusions to be drawn: 1 ProDOS 1.2 will recognize up to four devices controlled by a Smartport in slot 5, where the ligs ‘Smartport lives. Extra devices in other slots are ignored. This is in agreement with the claim made in the igs Firmware Reference Manual. As far as | can tell, however, ProDOS 1.2 does not discriminate here in favor of the IIgs. Thus, connecting more than two devices to a Smartport in an older Apple II appears to be beneficial only as long as the Smartport resides in slot 5. 3.4 Open-Apple 2. Iffor some reason a block device exists in slot 2 (an extra memory expansion card, perhaps?) then ProDOS 12 will ignore anything beyond the first two devices attached to a Smartportin slot 5. 53. There doesn’t appear to be a simple method of detecting a RAMdisk attached to the IIgs Smartport short of quizzing the Smartport yourself. ProDOS knows only the number of devices attached to the ‘Smartportat initialization, and assigns device numbers accordingly. Thus, while an Apple memory expansion card is readily identified by the ID byte at $CnFB, a RAMdisk attached to the Smartport is just another anonymous block device. I suspect this is the reason that Catalyst 3.0 doesn’t recognize the IIgs RAMdisk. as such. Do you or your readers know of a cheap way to recognize the presence of a IIgs RAMdisk? Finally, a few observations about the Ilgs in general. I've been using a IIgs steadily since the end of November, and have concluded that a) it's an impres- sive machine (of course), b) the memory expansion card is a virtual necessity, and c) the genius who decided to attach the mouse to the keyboard appar- ently never twigged to the fact that detached keyboards tend to reside on laps rather than desktops. | am constantly detaching and re-attaching the mouse as I shift between text-and graphics-oriented applications. Paul Harwood NoetiComp, Inc. West St. Paul, Minn. I don’t know of any other way to identify a RAMdisk that's in a Smartport chain than to do the calls outlined here last month. The problem you mention with no “???s” in ligs disassemblies is alleviated somewhat by the fact that the IIgs memory dump includes an ASCII display. Try BLOADing ProDOS L.2at $2000, entering the Monitor, and typing 2000.5D00. You'll find alt kinds of interesting text embedded in ProDOS 1.2. I totally agree about the necessity of adding a RAM card to the ligs. The bare 256K machine is fine for older programs, but precious few ProDOS 16 pro- grams will run with that little memory. Escape route flaw fixed Itried the technique for embedding ESC sequences in AppleWorks files described in your January 1987 issue (page 2.95). | was able to modify color, density, and other printer features from text as described. However, the initial letters following carriage returns were deleted. The problem seems to be that AppleWorks sends the current boldface command at the beginning of each line. If that command is a simple ESC, the printer takes the first letters of the line as part of a printer command and does not print them. The printer sometimes even begins to do strange things such as reverse linefeeds, erratic color changes, and strange character sizes. My solution is to leave the “bold off” control sequence unchanged. | use control-B(oldface) to send ESC, type in the printer command I want, then use control-B again to make sure the boldface-off code, not the boldface-on code, will be sent at the beginning of the next line. ‘Anna Martin Houston, TX AppleWorks vs chemistry In the November, 1986 issue you claim that Apple- Workswill do 50 per cent of what advanced users want todo. Unfortunately, |am in the other 50 percent. The reason is that my Apple spends much of its time writing materials for my chemistry courses—tests, quizzes, etc. To do this I need to use a lot of special printer features—in addition to underlining, sub- scripts, and superscripts | use alternate user-defined character sets (on an Epson FX-80), expanded type, and italics. I realize that | can have one of these last three by redefining the boldface command, but | need all three. (At present I'm using a version of Screenwriter II that I patched up myself to allow access to these and other printer features). I can see, based on a little experience, that Apple- Works has a lot of advantages and | would certainly like to shift over to using it Hasn't anyone tackled the problem of fixing this most serious and obvious defect of AppleWorks, and providing a way to enter arbitrary printer command strings into AppleWorks data files? Itwould also be nice to be able to get into, and out of, the various printer modes more easily; in chemistry ‘one might use as many as twenty sub- and superscripts ina single line, which gets to be a lot of keystrokes. Maybe one ofthe utilities that allow macro definitions would help here. Walter K. Dean Huntington Woods, Mich. We get more letters about adding printer functions to AppleWorks than on any other subject. And we've printed quite a few of those letters here in Open- Apple. Now might be a good time to summarize and bring togetherall we've leamed about this subject so far. First, there is no way to embed control characters within an AppleWorks document, AppleWorks itself embeds control characters into word processor files and uses them for its own purposes. Even if you managed to slip a control character into an AppleWorks file using a disk zap utility, AppleWorks would not send it to your printer. This leaves you with the built-in AppleWorks printer controls, AppleWorks allows you to define three printers, only one of which may be a “custom printer.” The limit of just one custom printer is hard to live with. One trick that makes it somewhat easier is to use the “Silentype” printer, rather than the custom printer, for printing formatted text files to disk. See Open-Apple, August 1985, page 60b, for more information on formatted text files. Since the Silentype doesn’t support much of anything in the way of printer features, no control-codes will be appear in files printed to disk—which is what most people want, anyhow, ‘So now you can use your custom printer definition to tell AppleWorks lies about your printer's control codes; this allows you to get your printer to do italics, for example, when AppleWorks thinks it's doing boldface. AppleWorks stores the printer codes you define ina file called SEG.PR. Because SEG.PR isn’t kept in memory but is always loaded right before printing a document, you can effectively increase the number of custom printers by making several copies of SEG.PR with different control codes and switching disks when AppleWorks isn’t looking. Custom printer controls supported by AppleWorks include 21 settings for characters-per-inch, two Settings for lines-per-inch, and four bold, underlining, ‘superscript, and subscript on-off pairs. The four on- off pairs can be used within lines, but the per-inch codes can be used only between paragraphs. For printer features that you want to switch on and off for whole documents (such as printing a document in Vol. 3, No.1 near-letter-quality), rather than within a document, two other possibilities are to add the necessary control characters to a printer's interface-card setup string, or to print an otherwise blank spreadsheet after specifying the control characters youneed with open-apple;O's “Send Special Codes to printer.” As mentipned, the characters-per-inch and lines- per-inch settings can only be changed between Paragraphs, To use the per-inch controls effectively, you have tq turn on the fake per-inch (say you have your printer's code for italics stored in the 24 characters-per-inch control), print a sticky space and areturn, then tum the per-inch setting your printer is really using back on. If you don’t go back to the correct setting, AppleWorks will put too many or too few characters on each line or lines on each page. ‘The boldfpce, underline, superscript, and subscript on-off pairs can be used within a line or paragraph. The limit most people run up against is that, like you, they have flve or more features they want to turn on and off. Before giving up completely, one last trick is to define the on-code of one of the pairs as simply an escape chafacter (see Anna Martin's letter above for more detail on this trick). Your printer, the FX-80, uses ESC 4 pnd ESC 5 to turn italics on and off, ESCE and ESC F §o turn emphasized print on and off, and ESC Gand FSC H to turn double strike on and off. If you define boldface-on as simply ESC, try this (the carets are bold on and off codes): ehh “a"ttatics"#" aren't “€''s'bold You should get something like this: Italics aren't bold. The only remaining problem with this trick is that AppleWorks will count the 4, 5, E, G, F and H as characters when deciding how much to put on one line, but your printer won't actually print any of those characters. Consequently, lines with lots of codes will be “wrapped” before they should be. One of the AppleWorks macro programs would, of course, make entry of these codes much easier. One very nice feature that AppleWorks provides the Imagewriter that it doesn’t provide “custom” printers is support of justified proportional type. To get both proportional type and the ESC trick we've just presented, see the next letter. If combinations of these tricks still don’t provide you the printer control you want, it’s time to give up on AppleWorks. If AppleWorks could do everything, after all, there wouldn't be any need for the many other word processors that are available for the Apple Il. You might also consider a program, such as FontWorks from the Software Touch, that allows you to add printer embellishments to files created with AppleWorks or with other word processors. Zapping Imagewriter codes I've become (by default) the Apple answer man at the high school where | teach math. The November newsletter item about printing a whole AppleWorks document in bold was of interest because the Image- writer needs the double strike to make a good ditto master. Actually, since | use AppleWrite:. | had no trouble doing so but many of my colleagues use AppleWorks. I've combined Dean Shut’ tip with one that appeared in the December issue of A+ as follows. Use a disk zap program (I use Bag of Tricks 2) to search the AppleWorks program file SEG.PR for the February 1987 hex string 02 1B 21 02 1B 22. The $1B $21 and $1B $22 are the ImageWriter codes for bold begin (ESC!) and bold end (ESC “). The $02’s seem to tell Apple- Works how many characters are in the code sequence. Change both $02's to $01's and write the change to disk. This change causes AppleWorks to send just the ESC character when you embed, via a control-B, a bold begin or bold end in your document. This by itself does nothing. However, as was pointed out in the At tip, if you follow the ESC with a valid Imagewriter control code, it is executed and the characters are not printed in the document. So, to send a bold begin, (ESC !) you would type “control-B I’, Unless you tum off bold with ESC “, the whole document is printed in bold. More importantly, though, now you have the Apple Writer-like ability to send the sequences necessary for using color, Mouse- Text, half-height characters, and so on while not Giving up anything available from the open-apple-O printer options menu. Howard Jackson Albany, N.Y. Note Anna Martin's letter above. Your trick will work more reliably if you leave the bold-off code’s length at two and change it to something that's meaningful to your printer but that doesn't turn off something you want on (bold in your case)—for example, try using “ESC f’, the code for forward linefeeds. The trick also still has the problem with short lines mentioned in the previous letter. But what's neat about itis that it doesn’t use up the custom printer definition and it allows the ESC-code trick to work with proportional type. Using the same basic disk-zap trick, you could also do such things as embed commands in an AppleWorks printer definition that AppleWorks won't normally let you enter (such asacaret). Somebody might even be able to find the text that makes up the open-apple-O menu in the word processor and change the words “Boldface Begin” and “Boldface End” to “ESC start” and “ESC end.” More disk recovery tricks Inthe September 1986 Open-Appleyou mentioned POKE 47246,24 during a discussion of methods for recovering a crashed disk (page 2.59). This POKE changes a set carry instruction inside the DOS 3.3 RWTS routines to a clear carry instruction. At this point the carry bit is being used to tell RWTS whether there was an error in the targeted sector's “address trailer” or “data-checksum trailer”. Your POKE makes RWTS ignore these errors. The following patches take more time but have a better chance of reading bad disks: 8989:ER EA Ignore address checksum 992:EA EA Ignore address trailer B99C:EA EA 8920:EA EA Ignore data checksum BEOF:EA EA Ignore long gap betueen address and data 8936:ER EA Ignore data trailer 893E:A9 00 8940:ER EA Ignore ‘ ‘can’t find address’* (keep trying) For really hard cases only: 8960:EA EA One byte address header 896B:ER EA BeF2:EA EA One byte data header BOFO:ER EA These patches should cope with most disk errors and allow them to be read. However the data is probably corrupted and if this happens in a track/ sector list the problem is not gone yet. | hope this helps save wasted hours of re-typing. Charles H. Putney Dublin, Ireland POKE 47246,24 handles a fairly common problem caused by a Slow disk drive. This is that the last few bytes of a sector are overwritten by the first few bytes of the next sector when the next sector is written onto the disk. More severe problems can be attacked with your additional patches, but everyone should realize, as you point out, that the more severely RWTS is patched to ignore errors the more likely that any data you read will be corrupted. If you patch RWTS to completely ignore errors, you have to suspect the integrity of everything and you may get stuck inside an infinite loop trying to read an unreadable sector. Those who make severe modifications to RWTS should do so with a copy of Beneath Apple DOS at hand, Use it to try to understand what exactly these Patches do. Any error-checking you remove from DOS you will have to account for yourself. Protection errors Why can’t my disk sector editors always read track 0, sector 0? It seems to be unable to deal with disks that have funny sync bytes. Obviously the Disk II controller can read it What's different? Jeff Root Lancaster, Calif. Yoursector editor is using DOS 3.3's RWTS routines to attempt to read the sector. RWTS expects to find some byte patterns that mark the beginning and the end of the “header” and “data’ fields for each sector. Italso does some other error checking. The disk controller card, on the other hand, doesn’t use RWTS. The routines it uses to read track 0, sector 0 are right there on the controller card. Because of limited space, the program on the controller card has far fewer abilities and does less error checking than RWTS itself. People who do copy protection take advantage of this by introducing errors that stop RWTS but that don't stop the disk controller. ‘Tryjudicious use of a few of the POKEs mentioned in the previous letter. I'm sure you'll be able to read the sector you want to read by eliminating the right error. This is why copy-protection is such a waste of time—the computer has to be able to read the copy- protected disk and if the computer can do it so can any human who wants to bad enough. If this topic interests you I suggest you get a subscription to Computist, PO Box 110846-T, Tacoma, WA 98411 ($32/yn). General Manager resupported Inregard to the letter in your November issue (page 2.77) about the General Manager data base program, I'd like your subscribers to know that my company has taken over marketing and support of the program from Sierra Online and from PC Manager, owners of the program. 1 am upgrading the General Manager and will be releasing version 2.1A in January 1987. There are still both hard and floppy disk DOS 3.3 versions of the program. Upon reaching my office, any individual receives technical support at no charge. I also write custom applications for the General Manager for Open-Apple 3.5 individuals and organizations. Upgrades are $25.00 for the system disk and the cost ofa new program has been reduced to only $179.95. David Spooner Aspen Data Systems Box 567 Driggs, Idaho 83422 208-354-8185 Color=confusion What causes Applesoft memory to get scrambled occasionally and give me garbage such as “COLOR=" when I try to LIST a program? Link Keur Edison, NJ. Memory can get scrambled any number of ways. “COLOR="and similar garbage appears when Apple- Soft thinks there's a Basic program ina certain range of memory when there really isn’t. Fora lot more on this, see “Applesoft meets the RAMdisk” in our April 1986 issue, pages 2.17-2L Pascal RAMdisk loader In the December 1986 Open-Apple, page 2.87, there's a letter and answer about moving large files into a RAM disk on startup. You wrote that you didn’t know how to automatically load large files under the Apple Pascal operating system. | am writing to letyou know how to do it. The Pascal Filer will copy files of any length from a floppy disk to a RAM disk. The trick is to get this done automatically on startup. Here are the steps to take: 1 Make an Exec file that invokes the Filer and transfers the Filer from disk to RAM disk. This file is best saved to the boot disk. Type "M” for "Make Exec” and answer the prompt “New Exec Name:" with “*:st” to make the Exec file ST.TEXT on the boot disk. Answer the prompt “Terminator = %, change it?” with “N”, When the command line retumns, type “f” (for Filer), “t’ (for transfer) and at the prompt “Transfer what file?” type the volume and filename (including suffix) of the file to be transferred. At the prompt “To where?” type the volume name of the RAM disk followed by the file name (or “$” if you want to keep the same name). Then type "q" to quit the Filer. Type "%%" to end the Exec file and save it Check the boot disk directory; ST.TEXT should be there as a 2 block text file. 2.An Exec file cannotitselfbe a startup file. In order to execute ST.TEXT automatically on startup, a special program called SYSTEM STARTUP must be written which will chain to it. Here is such a program: program sysstart; uses chainstuffs { needed to chain } begin setchain (‘exec/#:st’) { chain to (exec) ST } end. Save this file as *:SYSTART.TEXT and compile it as ” to keep the same name. Upon successful compilation, use the filer to change SYSSTART.CODE to SYSTEM.STARTUP. That's it. When the computer is booted using the boot disk, the program SYSTEM.STARTUP will run the 3.6 Open-Apple Exec file ST and copy the file or files you specified from the floppy to the RAM disk. My boot disk has an Exec file on it to transfer most of the stuff on the system disk (the Filer, Editor, Compiler, Assembler, Linker and other files). The system works considerably faster having these files (and program text files) on an Apple II Memory Expansion Card than on floppy disk, and makes Apple Pascal 13 a joy to use. Keith J. Bernstein The Bronx, NY. After reading your letter, Dennis found the directions for Pascal Exec files in the Apple brochure “Addendum to the Apple Pascal Operating System Reference Manual.” The brochure cautions that (as with DOS and Basic.system EXEC files) an unexpected occurence can throw the Exec commands out of sync with the prompts. For example, trying to Transfer a file toa RAM disk a second time may cause the prompt “Remove old <filename>?” to appear, which may send your Exec file off into the weeds. Another consideration is the loss of the original ‘SYSTEM.STARTUP file. Rather than replacing the original SYSTEM. STARTUP with the new file, you may want to rename the original and then ‘chain’ to itat the end of the Exec file by adding a command sequence such as “X" (eXecute), and then “*:OLDSU.CODE" (to reply to the prompt “Execute what file?’) to the end of the Exec file. Incidentally, the “*” prefix is used to tell Apple Pascal that the specified file is to be found (or created) on the boot diskette (see the Apple Pascal Operating System Manual for particulars). May | have your name? The RWTS routine in the DOSTalk Scrapbook is great. I even use it to find volume names on ProDOS disks. How about a similar routine for ProDOS or even justa way to get the volume name? On page 3-17 of Beneath Apple ProDOS is a chart converting sectors to blocks. It indicates that sector 11 on track 0, which is where | find a disk’s volume name, is the second half of block 6. But on page 4-8 it says that the volume name is in block 2. What gives? Here's a tip you may be able to use in exchange for your help. You can load any ProDOS file into Apple- Works, not just text files. AppleWorks will change all control characters (ASCII 0 to 31 and 128 to 159) to the symbol “#”, All other characters appear as themselves. This makes it easy to convert files from word processors that use binary files. Guy Forsythe Westerville, Ohio Here’s a couple of simple ways to find out a disk’s name. Put the disk you're unsure about in a drive (slot 6, drive 1 for these examples) and type any of the following: CAT,S6,01 (volume nane appears at top of catalog) PREFIX,S6,01 (set prefix to what’s in S6,01) PREFIX (ask for current prefix) PREFIX / (slash by itself voids prefix) CAT,S6,01 (make 6,01 the active drive) cat The final example is useful if you have several disks whose names you don't know. After the first one, all you have to do is insert the next disk and type CAT to leam the names. Gary Little has a simple Applesoft block read/write Program in his book Apple ProDOS: Advanced Features for Programmers, pages 36-40. The chart on page 3-17 of Beneath Apple ProDOS confuses just about everybody who tries to use it. See our July 1985 issue, page 55, for an explanation and forauseful ProDOS to DOS 3.3 sector conversion table. We tried your AppleWorks trick, and it works nicely. Applesoft ONLINE call While using ProDOS, | often have as many as four volumes on line at any one time. Out of necessity, | wrote a STARTUP program for my hard disk in which | included a subroutine to check each slot and drive and display any volume names (see listing below). This program can be run independently or you can make ita part ofa larger program that needs to know what volumes are online. It's possible to put ProDOS volume names into an array for easy manipulation. | thought your readers might enjoy knowing how to “capture” a ProDOS volume name. 5 REM * — VOLUNE.FINDER —* GREK 5 =slot x 7 REM * © -D =drive * BREN * — VS=volune name * 10 HOME : S=1 : De1 : OS=CHRS(4) 20 ONERR GOTO 50 30 PRINT DS;“PREFIX,S”;5;,0":0 40 PRINT D8;“PREFIX” : INPUT “*3V8 50 POKE 216,0 : CALL -3288 : REN fix ONERR 60 PRINT “SLOT 3S: VS="" 70 SeS+1 : IF $= 80 IF 0C3 THEN 20 9@ POKE 216,0 : REM clear ONERR flag Joseph Kline Lubbock, Texas Readers curious about line 50 should take a look at “Digging into DOS” in the January 1985 Open- Apple, page 2. Try adding these lines to your program: 25 POKE 222,0 55 IF PEEK(222}=3 THEN 70 ligs system utilities We have had an Apple Ilgs with 3.5 drives for several weeks now and it is super. But we don't like the long, interminable System disk boot up necessary to make backup data disks (it takes over 80 seconds just to get there). The QUICK COPY program itself is magnif- icent, but getting to itis the drag, Is there any way, using the file moving options in the System Disk, to transfer a startup file and just the copy program to another disk, so that when we want to backup our 3.5 disks, we can have a separate, fast attack copy program? Or does such a program already exist on 3.5 disks? Preston Boomer Santa Cruz, Calif. Interestingly, there are two complete system utility programs on the IIgs system disk. One is in the subdirectory called SYS.UTILS. This is an updated version of the IIc system utilities. It was written by Apple's own software engineers. The other is in the subdirectory called DESKTOP. It's a renamed version of MouseDesk, which was developed by a French company called VersionSoft and was once distributed in the U.S. by International Solutions, which has since gone out of business. Both of them are ProDOS 8 programs. Both of them can copy disks, but the disk-copy portion can’t Vol. 3, No. 1 be separated from either. To make a disk that directly boots into one of these programs, start by initializing anew disk and copying the ligs system disk program /SYSTEM.DISK/SYSTEM/P8 into the new disk’s main directory. Then rename “P8" “PRODOS”, for that is what it really is. For Apple's system utilities, next copy all the files in /SYSTEM.DISK/SYS.UTILS into your new disk’s main directory. That's it. For the MouseDesk look-alike, copy the files / SYSTEM.DISK/DESKTOP/DESKTOP1 and /SYS- TEM. DISK/DESKTOP/DESKTOP2 into your new disk’s main directory. Rename “DESKTOP1” “DESK- TOP.SYSTEM”. If you have a 3.5 disk, you could consider moving the other files from the DESKTOP directory to your new disk, too, but the program will start up faster without them. The other files won't fit ona floppy. Besides the programs on the Ilgs system disk, there are several stand-alone copy programs that have been designed to speedily copy both 5.25 and 3.5 disks. ProSel has one, as does Living Legend's Misk Disk #1 (see September 1986, page 2.64c). Bill Basham has already updated his DiversiCopy to use IIgs memory ($30, Diversified Software Industries, 34880 Bunker Hill, Farmington, MI 480018-2728, 313/553-9460). File lost, sort of I've got my word processor data disk set up using very efficient subdirectories and I've had no problems. But I just saved a document as /DATA/LETTERS/ APPLE and it's disappeared. This should have been the thirteenth entry in the directory but it only shows twelve. | know the file's there because I can use any other program and it shows in the catalog, but I can’t find itwith Apple Writerand | can’t load it even though its there. Michael C. Tiernan North Reading, Mass. You've been sorting directories with a buggy sort utility. See “Don't try to sort ProDOS directories” in our March 1986 issue, page 2.10. What you need is Glen Bredon’s ProSel. Aprogram called MRFIXIT on that disk will fix your damaged directory. A program called CATDOCTOR will sort your directories correctly. You get an excellent program selector free ($40, 521 State Road, Princeton, NJ08540). Booting ProDOS from DOS 3.3 Whats the best way to print a path tree? I have a 20 meg hard drive and can get lost easily. My hard drive is in slot 2. The kids play on the computer a lot, and if it is in 7 either 6 won't boot (hard drive off) or they tum the hard disk on and off. When I boot my Sider with a PR#2, DOS 3.3 works fine, but I can’t get into ProDOS without booting it elsewhere. | can’t boot ProDOS from the Sider's menu. PS. The tip on Glen Bredon’s ProSel is worth many years of your subscription price. The article on the llgs is priceless. | spent $20 on magazines for a lot less information. Bob Wilson The answer to your ProDOS path tree problem is easy if you already have ProSel; try running the program on that disk called INFO.DESK. And don't be embarrassed you didn't already know it was there — there's so much good stuff on that disk it takes February 1987 weeks to find it all. Likewise, getting ProDOS to boot from DOS is pretty simple. First, we have to get PRODOS into a BiNary file that we can convert into a DOS 3.3 file (PRODOS is normally a SYStem file, which can’t be converted). Do a CATALOG to find the length of the PRODOS file you want to convert, (PRODOS 111 is 14848 bytes long, PRODOS 12 is 15485 bytes.) Then: BLOAD PRODOS,ASZ000, TSYS BSAVE PRODOS.0BJ,A82000,L77777 Now you can use CONVERT or Copy II Plus to transfer the new file PRODOS.OBJ to a DOS disk. To get the file onto the Sider, you'll have to convert it onto a 5.25 DOS 3.3 disk and then use a FiD-type utility to get it to the Sider, since neither CONVERT nor Copy II Plus is equipped to deal with DOS 3.3 on devices other than 5.25 floppy drives. Once you have the file on the Sider, change its name from PRODOS.OBJ to just PRODOS. Now the only problem is that you can’t BRUN the PRODOS file without a few considerations. After the PRODOS image is loaded and executed, it will relocate itself up into the top of the Apple memory. PRODOS by itself in memory is “all dressed up with no place to go,” as it knowsa lot about managing the disk drives but doesn’t do much useful work by itself. PRODOS will therefore attempt to load a system Program (such as BASIC.SYSTEM) to tell it what to do with the knowledge it has; PRODOS looks for the first system file in the “boot disk's root directory that has aname ending in ".SYSTEM”. If such a system file is not found, ProDOS throws up its hands and stops with an “UNABLE TO LOAD SYSTEM FILE” error. This means that in addition to the DOS disk from which we are going to BRUN PRODOS, we also need a ProDOS volume on-line with a suitable system file so that ProDOS has something to load and run. And we have to be able to tell ProDOS where this disk is. It tums out that ProDOS assumes the boot drive's slot (times 16) is stored in zero-page location 67 ($43). Therefore, we need to POKE the desired slot number times 16 into that location before executing PRODOS. For your hard disk in slot 2, a program to start PRODOS would look like this: 10 REM ** Boot ProD0S from DOS *** 20 SLOT = 2: REM slot nunber of ProD0S boot disk 30 PRINT CHRS(4);“BLOAD PRODOS,AS2000" 40 POKE 67,SLOT#16 : REM pass slot number 50 CALL 6192 : REN $2000; start ProDOS After PRODOS starts up, it will now look in the ProDOS volume connected to slot 2, drive 1, for a system program. Make sure there's something there for it to execute and your problems should be solved. VCR backup Please expand on the subject of hard disk drive backup using a video cassette recorder. In the ‘Apple Ilgs Book’ (Bantam) by DuPrau and Tyson, the authors comment on page 138 about using a device by Corvus anda VCR. This would be significantly less expensive for someone that already owns a VCR than buying a backup device like B-Sider. William D. Blessing Tulsa, Okla. Back in the Olde Days of microcomputers when hard disks were rare and tape backup units were almost non-existent, Corvus used to manufacture a product called the Mirror that allowed the use of a VCR as a back-up device for Corvus hard disk drives. We called Corvus‘s sales department (at 408-559- 7000) to see if the device was still in production. We were informed that although Corvus still supports the unit, it is no longer being manufactured. The Sales representative told us that some surplus units may be available. Although we couldn't locate an old ad with the price, as we recall the unit cost several hundred dollars without the VCR, within the range of the B- ‘Sider (which includes the tape drive unit). And the Corvus unit may not support non-Corvus drives since backing up a disk device is a matter of combined hardware and software. No help available Here's a short question you might be able to help me with. What can I do when I get a NO BUFFERS AVAILABLE error? This often happens when | have GPLE and Double-Take up and running in ProDOS. Could | be the only person with this problem? John M. Sklar Brown Deer, Wisc. The NO BUFFERS AVAILABLE message typically occurs when youare trying to opena file or otherwise access a disk and there isn’t enough memory space available for a disk buffer (1K). You'll also get this message if you have the limit of eight buffers already open (rare), or if you attempt to BLOAD a file into a memory area that has been protected in the ProDOS system bit map (not rare). The protected areas are the zero page and stack ($0000-S01FF), the text screen ($0400-$07FF) and parts of Basic.system itself. GPLE and Double-Take may also set the map to protect the areas they are in. Here's a short program you can run to determine exactly what memory pages are marked as protected at any given time: 100 REM * ProOS System Bit Map Reader * 11@ PRINT “ProO0S System Bit Map" : PRINT LIS PRINT ~ Pages sarked X are protected” : PRINT 120 PRINT” 0123456789ABCODEF 125 ADR=48964 : REM where bit map is in global page 13@ FOR 1-0 10 11: 142 15+STRS(I) 150 IF I=10 THEN IS="A" 160 IF T=11 THEN [5="B~ 170 PRINT 18;”-"; : REM print label at left edge 180 FOR K=0 0.1: REN tuo bytes=16 pages 190 V=PEEK(ADR) : AOR=AOR+1 200 FOR L=7 TO 0 STEP -1 : REM bit magic 210 vse" * 220 IF V=>2"L THEN VS=°xX": V=V-2"L 230 PRINT VS5" “; 240 NEXT 250 NEXT 260 PRINT 270 NEXT 280 END The ProDOS versions of GPLE and Double-Take install beneath Basic.system and use up part of Applesoft's memory. If you load a long program, memory can start getting very tight. Execute a PRINT FRE(0) to determine how many bytes of free memory you have left at any particular time. If the answer is less than 1,024 or so, you simply don’t have enough memory left to access the disk. The only alternatives are to split your Applesoft program into modules (see April 1986, page 2.17) or to remove GPLE or Double-Take. Note that Double-Take is modularized; you may be able to keep your favorite features of REM do 12 sets of 16 men pages Open-Apple 3.7 that program and remove only the features you don't often use. A better MouseText trap Ihave a solution for those who desire to have the old MouseText-less character set on the IIc (November, page 2.78). The operation requires a hardware hacker's tum of mind, however. Whatis involvedis laying open the Ilcand unsolder- ing the character generator chip (Video ROM) from the motherboard. The squeamish need not apply as there is potential for disaster here. The chip in question is located under the disk drive. The clearance is such that it is unlikely that a chip socket can be installed here for a new chip. The new chip to be installed is a 2764 EPROM (approximately $4) into which you must program a copy of the old character generator data from a borrowed pre-enhancement Ile. This data goes into (let's arbitrarily say) the lower half of the 2764 and the IIc’s present data goes into the other half of the 2764. You'll also need to make a hole through the Ilc’s case for access toa miniature toggle or slide switch. A conveniently unused cranny is available just above the shelf’ near the speaker volume control. A generous blob of epoxy will serve to anchor the switch in place. A few things to note: The character generator ROM data is not copyrighted. Every time you trade in your IIc motherboard for the newest upgrade you have to make this modification all over again. Worst of all, your dealer may not accept a modified motherboard on an exchange basis, which could put your ability to get upgrades in jeopardy. And since broken Ilcs are usually fixed by means of a motherboard exchange, you may even have trouble getting a modified IIc repaired without unmodifying it first. Save those old parts! The Ile and IIc character generator ROMs are equivalent to 2732 EPROMS and are 24 pin devices. The 2764 is a 28-pin chip and contains twice the number of locations as the 2732. The early IIc motherboard that | worked with was laid out to accommodate 28 pin chips. Could Apple have been on the verge of installing a ‘dual’ character generator? Ifnot, why not? You must bend pin 2 out to the side before soldering the 2764 to the motherboard. Pin 2 will not be soldered to the motherboard. The select switch controls whether pin 2 is yanked low or pulled high, thus determining which character set is displayed. A pull-up resistor, perhaps 4.7K ohms, must be soldered between pin 28 and pin 2. Then pin 2 is wired to one terminal of the selector switch. The other terminal of the switch is connected toa convenient ground, Such a ground connection can be made by securing the bared end of the wire under the motherboard mount- ing screw at the left front comer. You can flip the switch at any time (powered up or not) to make MouseText go away. Also you can switch MouseText back on just as conveniently. The software doesn't (and won't) need to know what is going on. This same scheme, using exactly the same chip, applies in principle to Ile’s. The practice is slightly different in that the 2764 is inserted into a 24 pin socket with pins 1, 2, 27 & 28 hanging off the keyboard end of the socket. Again, pin 2 is manually controlled by a switch and pulled high through a resistor. Also, solder a jumper to connect pins 26, 27, 28, and 1 This provides 5 volt power from the socket’s power hole to the chip’s power pin and makes the other mentioned pins happy. Note also that the concept can be applied to 3.8 Open-Apple produce ROMs containing switchable English and Spanish (or other) character sets. Also, a 27128 or larger ROM could be employed to produce a killer character generator. All that is needed is a suitable multiple-position switch and the right connections to map in the appropriate 4K chunk. Those not wanting to go to the trouble of burning their own EPROM may want to check out a product from Bone Frontier in the November 1986 inCider, page 158. I haven'tseen their gizmo butit sounds just like what I've described here for the Ile. They don’t mention the llc. It's $50 price seems a bit steep, however, when you consider that's 5/6ths of the cost ofan EPROM programmer board from JDR-Micro. Incidentally, the EPROM burner | used for this exercise reports a checksum of $F800 when reading astandard Ile character generator ROM. The checksum. of an enhanced Ile character generator is $B595. Appropriately enough, combining these two chunks of data into a single 2764 yields a ROM whose checksum is reported as $AD95 ($F800 + $B595 = $AD95, with a carried bit lying around on the floor). Asan aside, | had originally offered to perform this modification for a friend, free of charge except for a few bucks to cover the cost of parts. After I told him the details, he had second thoughts and changed his mind about the proposed surgery. | then offered the same deal to another friend. The second friend accepted the offer and is still very pleased with the outcome. He also allowed me to take his modified IIc Open-Apple iswritten, edited, published, and © Copyright 1987 by Tom Weishaar Business Consultant Richard Barger Technical Consultant Dennis Doms Circulation Manager Sally Tally Business Manager Sally Dwyer Most rights reserved. All programs published in Open-Apple are public domain and may be copied and distributed without charge Apple user groups and significant others may reprint articles from time to time by specific written request. Requests and other editorial material, including letters to Uncle DOS, should be sent to: Open-Apple P.O. Box 7651 Overland Park, Kansas 66207 U.S.A. Published monthly since January 1985, World-wide prices (in US. dollars, airmail delivery cluded at no additional charge): $24 for 1 year, $44 for 2 years, $60 for 3 years. All single back issues are Currently available for $2 each; bound, indexed editions of Volume 1 and 2 are $1495 each Open-Apple P.O. Box 6331 Syracuse, N.Y. 13217 U.S.A. Subscribers in Australia and New Zealand should send sub ‘scription correspondence to Open-Apple, c/o Cybernetic Research Lid, 576 Malvern Road, Prahran, VIC $181, AUSTRALIA, Open-Apple is available on disk from Speech Enterprises, P.O Box 7986, Houston, Texas 77270 (713-461-1666), Uniike most commercial software, Open-Apple is sold in an unprotected format for your convenience. You are encouraged 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 distribution to others, The distribution fee is 15 cents per page per copy distributed. WARRANTY AND LIMITATION OF LIABILITY. lwarrantthat mostof the information in Open-Apple is useful and correct, although drivel and mistakes are included from time to time, usually unintentionally, Unsatisfied subscribers may return issues within, +180 days of delivery for full refund, Please include note from your parents or children confirming that all archival copies have been destroyed. The untulfiled portion of any paid subscription willbe refunded on request. MY LIABILITY FOR ERRORS AND OMISSIONS. IS LIMITED TO THIS PUBLICATION'S PURCHASE PRICE. In no case shall | or my contributors be liable for any incidental or ‘consequential damages, nor for any damages in excess of the fees paid by a subscriber. ISSN 0885-4017 Printed in the U.S.A. Source Mail: TCF238. CompuServe: 70120,202 to my user's group where | demonstrated it and offered my services to the group to perform the mod. The interest was underwhelming, Here is an idea every bit as good as sliced bread, and yet | feel it will take a marketing maven and nota marketing Marvin to make it fly. Maybe the $50 is not so out of line after all. Marv Wager San Diego, Calif. Slot for terminal needed Ineed to knowifit’s possible to use my Super Serial Card to drive both my printer and a modem. While I realize that it can’t be done concurrently, is there a way I can buy (or rig) a switch to do it? I have a fair understanding of wiring, and I realize that the jumper block on the SSC is a modem eliminator. To my understanding, this simply reverses some of the lines, and it doesn’t seem it would be so tough to have an external switch that could flip between either the modem or the printer, and reverse the lines depending on which peripheral was being addressed. So, can it be done? How? My reason for asking is, of course, because I'm out of slots. My printer is in slot 1, a clock card in 2, an accelerator in 3,a mouse in 4, two UniDisk 3.5sin 5,a DuoDisk in 6, and a RAMFactor in 7. Help! A possible solution I've read about is SRCG's Switch-a-Slot, but I suspect that to use it would have to tum off my computer to switch slots, This would be so inconvenient with my hardware/software setup that itwouldn’t be worth the effort. Of course, | may be wrong; the advertising doesn’t address it, and as usual my dealer is as useful as an extra nostril. Do you knowifl can use the Switch-a-Slot without tuning the computer off? Finally, | need to know if | can get software to emulate an IBM terminal connected to an IBM System 36 in our building. This, of course, is why I want to hook up a modem in the first place. Actually, I have to manage it, or my department may take away my beloved Ile and give me (ick!) an IBM PC. Finally, is there an accelerator card for the Apple III, or an accelerator for the Ile that can be modified to work with the III? Dean Esmay Flossmoor, Ill. RS-252 switches are common and available froma number of sources (at widely varying prices). For example, Moore Business Forms has a number of different switches in its computer products catalog, ranging in price from $80 for a simple A-B switch to $220 for a box that can switch between six devices. You would also need a null modem cable. If your ‘Super Serial Card is set to “printer,” use the null modem cable between the switch box and the modem. Conversely, if the SSC is set to “modem,” use the null modem cable between the switch box and the printer. If you decide to build the switch box yourself, don't attempt to put any crossovers in the switch itself—just run everything straight through. RS-232 wiring can get extremely complicated in no time at all. If a simple null modem cable doesn’t seem to work, you might want to look at “You Can Make the Right Connections”, in the June 1986 At, Pages 53-64. Southem California Research Group’s Switch-a- Slot does require you to power down before changing the selected card. Another product supposedly available is called ProSlot (from CGL Microsystems, 10407 Blue Ridge Blud., Suite 522, Kansas City, MO 64134, 816-385-6619). Dennis has talked to the Vol. 3, No. 1 company's answering service several times but has been unable to get anyone who knows anything about the product to call him back. According to an ad in the February 1986 Nibble, page 59, the device fits into slot 7 and provides two software-selectable slots; it costs $119.95 plus $4 shipping and handling. Dennis says that terminal emulation for the IBM is tricky if your system 36 is close-mindedly using the IBM “standards” for communications and character codes. IBM only recently decided to recognize ASCII devices; many older IBM systems use a different byte-to-character code known as EBCDIC. Hooking up to such a machines often requires that a rather expensive emulation device be connected to your computer to make it look like an IBM terminal. Most current IBM systems will allow you to define your terminal as an ASCII device and will communicate with you througha device called a protocol converter (not the same as the Apple Smartport gadget dis- cussed at length last month). You need to check with your data processing people and see what kinds of terminals they will support, then find a software Package that will emulate that device. ASCH Express, the Professional, for example, will emulate an IBM (ASCII) 5101 terminal. One thing you have to watch out for is that some “emulations” do not fully duplicate the function of the original terminal for some hardware-specific items such as special function keys; for example, AE Pro's emulation of the DEC VI-52 does not include generation of the keypad function key codes. Sof- tronics’ Softerm 2, which performs 24 terminal emulations (including an IBM 3101-20 in block mode), includes a hardware device to act as an ancillary keypad for some emulation functions. Softerm 2 is $195; you can reach Softronics at 303- 595-9540. We don’t have an answer to the “accelerator card forthe Apple III” question. The best answers to Apple III questions come from the folks at On-Three, P.O. Box 3825, Ventura, CA 93006, 805-644-3514. Communications software Doyou know ofa good (and inexpensive) commun- ications package that runs under ProDOS? | have been using Hayes Smartcom I that comes with my MicroModem Ile, but it captures files under DOS 3.3.1 have to convert them to ProDOS before using them. Also, | cannot install Smartcom on my UniDisk 3.5. Joe Movich Alta Loma, Calif. There are a large number of ProDOS communica- tions packages. Most of them have been included in comparative reviews in Apple magazines recently. Go to your local library and look at “Telecommunica- tions Software” in the April 1986 A+, pages 24-36 (14 Packages compared); “Telecommunications Software Review” in the January 1987 Call -A.P.P.L.E., pages 11-19 (five packages compared); and “Telecommuni- cations, The Software Connection” in the February 1987 inCider, pages 55-62 (nine packages compared). Inaddition, some word processors, such as Apple Writer, and most desktop accessory programs, such as Pinpoint, Jeeves, and Fingertips, allow rudi- mentary use of a modem and transfer of text data. There is also a public domain terminal program, called Kermit, that is widely used in the university environment. If you are associated with a college or university, your data processing people should be able to get you a copy of the ProDOS version of this program.