READERS' HINTS AND TIPS  Ed: Keep those hints and tips coming! This issue, we have many more letters than hints, which is fine but many people find a lot of useful stuff in this section, so if you know something useful, please send it in. It's a way of thanking others for having taken the trouble.  ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| INKING YOUR DOT-MATRIX PRINTER Dear Ed: WARNING!!! Dot-matrix printheads (in the main) rely on very special DOT-MATRIX INKS and the lubricants they contain for their lubrication. To illustrate this point a properly compounded DOT-MATRIX ink consists of 70% high quality lubricants and 30% of NON-pigmented dyes. The use of stamp-pad, printers or duplicators inks is courting disaster. Even if lubricants are added to these inks the pigments they contain would act like a grinding paste of the delicate pin shafts of the printhead. One other false economy is the widespread practice of pounding your ribbons till the printout is gray. As the printheads receive their lubrication only from the ink in the ribbon, this procedure is analogous to running your car without oil in the sump. Another side effect is that the ribbon fabric is so compacted, in the track area, that it is unable to properly absorb ink when re-inking is attempted. The use of unsuitable inks clogs the ribbon making it even more unsuitable for re-inking. Professional re-inking costs between $3 and $5 for an eighty column printer cartridge so why gamble. Printheads are horrendously expensive, costing as much as half the price of the original printer. Bill Sharpe-Smith. Perth WA  ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| SOME REPLIES TO READERS QUESTIONS FROM MD27 by Steve Bolton  TO: David Boyce Atari ST Emulators. * Some time ago I got the public domain ST emulator surprising called Atari ST Emulator V1.2a. This was just to see if I could run any of the ST programs I have on various magazine coverdisks that were dual- or tri-format. I have some success in running the programs from the coverdisks of Public Domain Magazine (No.s 2 & 3) which included a virus killer (ST VirusKiller V3.11) and an animation program (AniST V2.11). But unfortunately all the dual format disks from Zero magazine and ACE that I have wouldn't run the game demos. Now I know absolutely nothing about the ST and know of no-one who does have one so I can't vouch for its abilities. I would hazard a guess and say that software and NOT games (in general) would work. I would think that it is the way that games are programmed these days to stop piracy. I don't know if the Amiga joystick is supported by the PD emulator but it appears to use all available memory in the system so if you have a hard drive with extra memory etc it will accept it. This emulator is in German and is the only one I know of (though I believe there is an English version of the same emulator). I recently got Public Domain Coverdisk 4 with a program (ST) called Quick Index V1.5 which does performance tests on your ST much like the SysInfo program for the Amiga. This program will run under the ST Emulator and the following is the performance of the emulator compared to a standard ST. (These results were the same for TOS 1.0 and TOS 1.4 under the low, medium and mono monitor options.) CPU Memory 61% CPU Register 61% CPU Divide 62% CPU Shifts 63% DMA Reads 66% GEMDOS I/O 14% DISK (RPM) 151 (The following are for various combinations of the TOS and monitors) TOS: 1.0 1.0 1.0 1.4 1.4 1.4 MONITOR: Low Med Mono Low Med Mono ------------------------------------------------------------- BIOS Text 46% 64% 62% 44% 60% 58% BIOS String 41% 62% 61% 41% 61% 60% BIOS Scroll 82% 83% 85% 61% 61% 63% GEM Draw 100% 78% 80% 76% 62% 66% ------------------------------------------------------------- * Now for some good (?) news. I did at one time see in the News section of a magazine that an ST emulator that plugged into the side expansion port of an A500 was available but I was unable to locate the article (it may have been the one that you say). BUT while I was looking I found ANOTHER one. It was in an ad for Memory And Storage Technology (M.A.S.T.) on page 105 of the March 1991 issue of Amiga World. It is a soft/hardware combination called brainSTorm which is supposed to be 98% compatible. It supports 8mb Fast RAM, 2mb Chip RAM, all ST resolutions, Amiga drives, serial & parallel ports and the hardware has sockets for Atari ROMS. It is advertised at $US99 so it may be a cost effective solution for anyone changing from ST to Amiga and has a large software collection. The Australia contact number in the ad is (02) 281 7411. I haven't seen it advertised elsewhere but if they made it then I'm sure that they would have some around to sell. If the PD emulator I have is different from your own then you are welcome to send me a disk and some stamps for postage and I'll copy it for you. - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - TO: Hayley Hummerston DKBTrace  VIRTGIF & HAMSHARP Moria * I have been using DKBTrace V2.12 from Fish 513 & 514 and have had no troubles at all producing some impressive pictures from the includes data files as well as making my own. Since the version you have (from MISC 28) is on the single disk and the data files are .data files rather than .dat files then I assume it is one of the earlier versions. As I'm used to V2.12 I can only offer a few suggestions to your problem. Firstly the whole DKBTrace collection of files for your version may not have been fully copied to MISC 28 when it was set up (ie. some vital file may be missing) though I doubt that this is likely. Secondly DKBTrace has the ability to dump the rendered image to various formats (dump, raw, targa) as a .dis file which is then processed by a secondary program to produce the dithered IFF picture. Since this is the case it may be that you've produced .dis files which aren't readable on an Amiga (I believe that DKBTrace has been ported to Amiga, IBM, Vax, Mac, UNIX, X Windows (UNIX)). If your DKBTrace looks for the file "trace.def" then read it and see what options are set then carefully read the relevant part in the DKBTrace.doc file and make sure that the options are set correctly (the "dump" format is what you'll need : +fd ). If the trace.def file isn't required then you will probably have to enter the parameters from the CLI when starting the program, just make sure that +f or +fd are a part of them for the proper "dump" format. You may also be trying to produce a resolution not supported by the program. I suggest trying a simple file (a ball, ground and light) and rendering at 80x50 pixels size which will be fast and you can adjust a parameter and re-render until you finally find the parameters needed to produce a usable .dis file. This is just a guess since I don't know what version you are using and what its limitations are so if anything doesn't make any sense then it may not be implemented in your version. If you would like try version V2.12 then you can send me two disks and I'll copy F513 & F514 onto them for you (don't forget a couple of stamps for postage), if you do send me the disks would you be able to copy MISC 28 onto one so I can test it and make sure that the program does/doesn't work (it is possible that the program just doesn't want to play anymore).  Graphic Converters * Some good news though for your long lost VIRTGIF and HAMSHARP programs. They are available from Megadisc on Amiga Oz 6. You may have already read Paul Hayter's review of some PD disks on MD27a and noticed that Amiga Oz 6 was among them. If you have Kickstart V2.0x then you may also like to check out Fish 541 for GIF Machine V2.137. It sounds like a powerful and useful utility for GIF pictures but requires KSV2.0x to run.  Moria * ...And some goodish news. Moria (V3.00) runs fine from a hard drive (under V1.3). Your problem may be that you haven't assigned Moria to wherever you've put it. Mine is in dh0:Games/Moria directory. So if you have a directory called Games on your hard drive then a new directory inside that and copy all the files from your Moria disk into the new directory. Then assign Moria: DH0:Games/Moria and all should run smoothly. Just make sure that all the files (.pic, .fnt, .doc etc) are all in the same directory together. I have been using Moria V3.00 from Fish 194 on my hard drive without any hassles at all (under V1.3). I have now added the Kickstart V2.04 chip to my A500 and now Moria appears to put odd items in odd places in the dungeon (such as using food rations for walls!) and leaves a screen and the small window opened on the workbench and sometimes just GURUs after saving or quitting. But once again, yes Moria should and does work fine from a hard drive (under V1.3) and if you want to send me a disk and postage I'll copy F194 onto for you (if you copied MISC 28 onto the disk I could have a look at DKBTrace for you and see if I can find the source of your problems with it). - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - TO: Malcolm Capner Cubulus * I'm sorry to say that I've also tried everything I could to install this wonderful game on my hard drive but to no avail. I tried scripts, assigns, and even edited the program file to look for the pic and sound files in their respective locations on my hard drive but it still wouldn't work. I think that the game just doesn't like living in the fast lane of the hard drive and prefers the back waters of the floppy life. Still, someone may have a solution (maybe Tobias Richter would like to shed some light on the matter, it is his program!). For anyone interested, when run from the hard drive (from the Workbench or CLI) Cubulus opens a black interlaced screen and sits there staring blankly at you (silently laughing to itself?) and does nothing. You can still flip between screens (WB and Cubulus) so it doesn't lock the system, just uses chip memory and doesn't give it back. - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - FROM: Steven Bolton,  6 Domain Avenue,  Para Vista,  South Australia,  5093.  ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| MEGADISC 27B HARDWARE BBU_Clock_Fix ANOTHER VIEW In response to this article on `memory' of Ni-Cad batteries in an Amiga, there are several points that need to be made. The concept of `memory' in Ni-Cads has recently been challenged and speaking to various `experts', they will all give you a different answer about this issue. My interest in this subject is through my involvement with remote control model gliders and flying electric powered planes. I don't use those old fashioned noisy, messy model internal combustion engines!! From zero knowledge I have had a steep learning curve and can now understand in a simple fashion, amps, volts, current, differences between fast and slow charge, and why a 900mAh battery pack has less power than a 1300mAh for example. My understanding of the `memory' problem, (and the model aeroplane books and magazines are for ever discussing the issue) is to put the batteries under load to wipe out the `memory'. I use a 12 volt globe which I have soldered two wires, + and - . I attach the wires to the battery and when the globe has no light, then I know that it is totally discharged. The globe runs down the voltage until it is zero. Now you can charge them. If you believe there is still a memory and it needs it, repeat the process. In this way you will have your batteries in tip top condition. The method suggested in the article of wiping out the `memory', I believe could end up damaging the battery to the point that it was useless, but as the article in Megadisc says, the battery is out of warranty and what have you got to lose. As another less extreme option why not try my method, it's a little easier on the battery. I would add using a car battery in the fashion described may cause the battery to explode.  ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| COMMENT ON NICAD BATTERIES In Megadisc 27, someone described a method to rejuvenate (?) Ni-Cad batteries in battery backed RAM extensions. I recently had a chip failure, as I thought, because of the message "Battery backed Clock not found" in the startup-sequence, whilst the battery actually read 3.6 Volts. I wrote to HES, the distributor of the RAM extension, who wrote back to say: " * Disconnect Battery and remove Clock Chip. * When removed, short out legs for 10 minutes with conductive material, i.e. foil. * Re-insert Clock Chip and re-connect Battery. Should the above fail to work, we do sell the Clock Chip from here, at a cost of $15.00 including postage. " Thinking that the safest to disconnect the battery was to un-solder it, I soon found that the + terminal of the battery was very poorly welded to the PC board connection. I distinctly remember that when I measured the voltage, I had used the meter probes on the battery itself rather than the connection. The actual failure was a very high resistance between the battery and the connection. After carefully soldering the connection back onto the battery and re-soldering the battery to the PC board, everything then worked fine. Some food for thought! Normally Varta brand Ni-Cads are spot on with their welding. Perhaps it was a Monday or a Friday one. Beau Rice Pearce ACT  ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| AIRBUS A320 - WORKBENCH 2.04 - POSSIBLE PROBLEM ----------------------------------------------- Readers with an Amiga 500 modified by adding the WB2 ROM who are thinking of purchasing the great airliner simulation "Airbus A320" might be interested in the following:- We purchased this program recently and as is our normal procedure attempted to back up the 2 supplied disks. Firstly the "Simulator" disk can not be copied as it is fairly heavily protected - XCopy 3 would not copy it. Secondly the "Pilots-log" disk would not copy or load either because it appeared to have bad sectors on the disk. In the end we were able to copy it from 1 out of 4 disk drives we tried it on so assume it had originally been produced on a drive which was out of alignment? Once we got the Pilots-log disk copied, we booted it up on our WB1.3 machine and all was well. (Why they had bothered to put disk protection on this program is beyond me, to use it you need the documentation which consists of about 100 airport maps and maps of Europe for navigation purposes. It would be impossible to use without all the paper-work!!) At the end of each level of the simulation, your pilots results are saved to the pilots-log disk. We then attempted to use the program on our other Amiga 500 which has had the WB2 rom chip installed and runs under WB 2.04. The program ran fine, but when a level was completed and the program wrote to the pilots-log disk, disk errors were reported and the program would not run again. We have re-copied the pilots-log disk many times and it always works properly on the WB1.3 machine, but always corrupts the disk on the WB2.04 machine. So please be warned, if you must buy software which writes to an original disk - REMEMBER TO BACK IT UP FIRST. We are at a loss to explain why these corruptions occur but assume the program uses a non-standard system to write to the drive which corrupts the directory structure. The program "Fix-Disk" corrected the problem for us, but most games players would probably not have such a program available. Mike Simpson Whalan NSW  ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| SIDECAR AND WB1.3 Regarding Phil Soden's query re sidecar and WB 1.3: I have been running Sidecar with 1.3 since its release this works both on a 1000 with kickstart disk 1.3 and a 500 with 1.3 roms. Copy your WB 1.3 to jh0: Suggest also copying setpatch from 1.3 to your copy of A1060 disk as setpatch should be first command in startup-sequence. By A1060 disk I am assuming that your sidecar supplied disk is named the same as mine Modify the startup-sequence on a copy of your A1060 disk as follows (note; if you have expansion memory put fastmemfirst early in startup up to conserve chip memory.) c:Setpatch r ;use 1.3 setpatch BindDrivers echo "A1060 Workbench 1.2 V33.56 7-MAY-87*N" Assign WB: sys:s IF EXISTS sys:s/JH0 echo "Initializing Janus ...*N" Wait 25 Echo "Mounting Janus Hard Disk*N" DJMount IF EXISTS JH0: Assign WB: JH0: Echo "Transfering control to JH0: *N" ELSE Assign WB: sys: EndIF EndIF assign c: jh0:c cd jh0: execute jh0:s/ROS ;my shorthand for rest-of-start saves typing The above mounts jh0: and any other partitions. Then create a file called ROS in the jh0:s directory that contains the rest of the startup including any assigns needed for your programs. This is a quicker startup as well as allowing you to run all of the time on 1.3. Because jh0: becomes your system ROS would need to contain assigns for all necessary WB directories i.e. s/devs/l/libs/fonts/sys Sample assign s: jh0:s also add path to utilities etc. When using normal c: commands it would pay to make assign resident until the end of startup if using a lot of assigns. We are using Arp which accepts multiple commands on the same line. T.Tricker Brahma Lodge SA  ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| ANSWERS TO MD27 QUESTIONS - DKBTrace // ERRATA NOTES TO SIDECAR? Dear Ed: After reading of Hayley Hummerston's problem producing an IFF file from DKBTrace last issue I tried it for myself. First using traceffp to produce a ".dis" file and then the dumptoiff utility to cunvert the output file. All worked correctly. However, when I first unpacked the files from Fish 397 I did have a problem getting it to run. One of the files is BasicShapes.dat, but all the examples expect it to be called BasicShapes.data. Change its name and all is well. Also there is a dungeon mapping program on Fish 603 in which Hayley might be interested. Does anyone out there have the errata sheets for the Sidecar that I could get a copy of? Lindon Flood Hobart, TAS.  ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| HANDLING A MOUSE MOVE Dear Ed: In his doc file for Mandle 1.9 on MD27, David Taylor was asking how to handle Mouse Move (REPORT MOUSE) messages quickly. He could try a technique that I have used successfully in cutting computational times dramatically: set IntuiFlags to INTUITICKS as well, and only draw on the screen when you gey a REPORT MOUSE and an INTUITICKS event. For the intervening REPORT MOUSES, just throw them away (Reply Mag (), but take no action). This will result in drawing only roughly ten times per second, rather than trying to handle the 640 times that would be needed if you moved the pointer right across the screen in one second. I don't know what the visual effect will be exactly, as re-drawing at 10 Hz will be noticeable to the eye, but it may be a lot better than waiting for the draw-rate to catch up to the pointer. It's worth a try, anyway. Geoff Farrell Holder, ACT.  ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| POPULOUS II CODES ~~~~~~~~~~~~~~~~~ 0 DOEGAC 1 AAWOAK 2 LONEAG 3 ACMEAB 4 OMJIAD 5 AKSUAF 6 OOAC 7 AGIIAC 8 OPOPAK 9 AMLYAG 10 UMHEAB 11 EMDOAD 12 UBTUAF 18 TUMOAD 24 VEPEAB 29 WIUPAK 33 MOMNAF 38 AFEGAB 42 ALJIAC 47 UHUXAF 53 CCUXAD 59 LYPIAB 63 THLOAC 65 AAATAT 67 ACUPAD 72 OPAMET 76 UBNE 82 TUABAG 88 VESUAD 92 DDLYAK 134 OOAKAF 138 UMLEAG 143 LEUBAB 148 MGATAF 150 MMUPAB Steve Bolton  ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| VERSION 1.3 IS NOT DEAD. What follows is for users of the 1.2 and 1.3 versions of the operating system. There are a lot of us who have puzzled at the volume of software being written and provided by Megadisc which works only under version 2.? of the operating system and this version is not yet readily available. Many machines will never undergo conversion or it will be some time before owners afford conversion. Whilst it is obvious version 2 has much to offer there is one area 1.3 has always been able to handle but I believe has missed out because of lack of a standard. Maybe version 2 will suffer the same fate. I refer to the ability to have WorkBench running in eight colours. On Megadisc 8 appeared Nic Wilson and Andy Clay's program, WBench8, together with two supporting programs by David Pochron. The facility was provided to add four additional colours chosen as the user saw fit. I used these programs for a time until I found David Pochron's other program, Colorful Workbench, on a public domain disk from Megadisc. It had been wisely renamed, Bitplanes, a name more descriptive of its function. I had already chosen four colours to use and these were based on the ability to mix them on a pixel basis to represent a wide range of apparent colours. They were full green, full cyan, full red and full yellow, in that order. It should be obvious that together with the original Workbench colours a vast range of subjective colours can be produced and icons made this way look great. I found that David had coded his choice of colours into the foot of the program and by using a byte editor I was able to change the defaults to suit my requirements. The advantage offered by using Bitplanes is that an entry in the Startup-Sequence reading "bitplanes 3" will turn on the four extra colours at a cost of 20K of Chip memory but it can be turned off again within the Shell/CLI with an entry of simply "bitplanes" if memory is tight. You can in fact toggle it on and off. It will also reveal that Workbench 1.3 was set up with more than four colours in mind. Try the menus in the title bar for example. It performs best when included in the startup-sequence as there is a tendency for rubbish to appear in a CLI or shell if it is run later. I do not know how WB2 handles its 8 colours but I have read reports that some people are reverting to the colours of 1.3. This may in part be due to the swap of black with white between the two operating systems and the old colours were better anyway. Is now a suitable time to try for standardisation of colour to the benefit of all? Maybe the ability to have such a wide range of colours has prevented more than the original four being used. This is unlike the limited but more widely used colours of other machines which will remain unnamed and they don't even provide a screen on which you could attempt to mix them. I-nhibited B-y M-entality. I have included some icons which might be convincing, together with the modified "Bitplanes" for you to put in your startup-sequence and join the rush by readers to enjoy the emerging standard. Why not get Image-Ed from Jonathon Potter and show Megadisc what you can do with these eight colours. P.S. There is a lot to be derived from going over your earlier Megadiscs as things that did not appeal may now suit your increasing knowledge or you may have just missed things. You will also find many icons had colour you did not know existed, and some unfortunately where it should not have. Rod Humphreys Moorebank NSW  ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| QUEER ICONS Oh hang on, I just remembered a suggestion I had concerning the icons on MD. As the icons look a bit queer under WB2, I think it would be nice to include the freely distributable program called MapIcon in a future edition of Megadisc. You can set this up to "invert" all the icons on a whole disk in one hit, so both 1.x users and 2.x users can have a reasonably pretty workbench. I have included MapIcon and its (very brief) documentation on this disk if you don't already have it. If you do include MapIcon on Megadisc, make sure you suggest to the readers that they diskcopy the original MD to rad:, run MapIcon on rad:, possibly optimise the disk in rad: and then diskcopy it back to floppy. MapIcon is incredibly slow working directly on a floppy disk and it sounds pretty bad too. Richard Booth Randwick NSW Ed: Check out "Icons_&_WB2" in the ARTICLES drawer this issue for a full discussion of this problem and how you can get the new ICONSTARTER Theme disk by Leigh Murray which looks at many ways you can improve icon use.  ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| MORE ON STEPRATES About Ron Williams' letter in MD27: The steprate bootblock was mentioned by Lesley Ayling in his article about speeding up the A500 by installing a faster CPU, and made it possible for floppy drive users to avoid the Catch-22 situation of being unable to change the steprate, because they couldn't load the steprate alteration program, because the steprate was too fast for loading to work... I have a hard drive, so when my modification works (it is very intermittent in fast mode) I can load a steprate program from hard disk and thus re-enable use of floppy drives, but if I wanted to boot from a floppy disk, I would have fatal problems with this method. A 'supercharged' A500 can at least read the bootblock successfully however, and if the bootblock slows down the steprate, everything is hunky dory. I have tried to obtain the disk mentioned in Lesley's article, but my letters to the address he gave (P.O. Box 334, Glebe 2037) have not received replies. If you have another address for Lesley, or can give me Ron's address, since his modification sounds to be more reliable than mine, I would again be very grateful. The knowledge that my Amiga would be going twice as fast if something was changed is extremely frustrating. Yours sincerely, Sean Flanigan Grafton NSW Ed: I haven't heard from Les Ayling for at least a year, and rumour has it that he has moved to Queensland. ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| HARD WORDS ON KINDWORDS For users of Kindwords word processing software: A word of caution to readers of Megadisc 27. Under Tutes_&_CLI, there was an item called Disk Customising by Doug Groat. Doug advocated the discarding of Clipboards from the newly customised workbench disc he was creating. If you use Kindwords software I would advise against this. Kindwords users may find (as I did), that on using the menu or mouse selection of text and trying to delete the text, causes the dreaded GURU. If so, then check that `Clipboards' (a directory) and Clipboard.device are in the Devs Directory. That could be the problem. Copies will be in the Devs directory of your original workbench if required. It took a while to work that one out. Talking of Kindwords: Sometimes with larger files it may seem that the computer is `frozen', that nothing is happening. Be patient. It most likely will come good if you wait a few seconds. Hitting control keys in panic will ensure all work is lost. Kindwords also seems to duplicate text (incorrectly) when you are moving about the text and correcting it. Kindwords is a bit slow and you have got it confused. In both these instances, just hold off for few seconds. When it appears to have settled down, if there is incorrect duplication of text, try moving back or forwards a page, and come back and check. Or press Ctrl-1 (1 on the numeric pad) to get to the start of your file, then go back to the duplicated section. In just about all cases Kindwords corrects itself. Don't you try to correct it before doing this because the text is not really there. After all, you didn't put it there. I think it is to do with the Clipboards again, as Kindwords keeps a copy of your last change for the `undo' command. The lesson is clear, do not ask Kindwords to do a second thing when it has not completed the first. Be kind to Kindwords week, and don't use files that are too long. THE SORT PROGRAM A reference to Superspell review on MD27 concerning use of Disearch and a handy hint for using Amiga DOS `Sort' programme. Handy hints on using AmigaDOS SORT which is memory hungry. Under Reviews on Megadisc 27 was a review of Superspell, a PD spelling checker. You rightly pointed out that on the Scrazzle disc (GA71) is the programme Disearch, which can be used to create a suitable dictionary for Superspell with the aid of the Amiga SORT programme, used from the CLI. The reason for the need to use SORT is that Disearch only sorts on the first three letters as it goes along making CULLING lists. Someone pointed out to me that even on a 1meg machine one can have problems. SORT needs a lot of memory to work in. A particular problem if you are trying to create as big a file as possible. Say the dictionary for Superspell. So for all the users of SORT, here are a couple of ideas for use in situations where AMY gets indigestion when SORTing. First the easy way. Make a separate file for each letter, 26 files. SORT each separately. If any files are still too large, then see the next, not-so-easy way for the too large files. Now use the AmigaDOS JOIN programme from the CLI to join all your files into one beauty. The not-so-easy SORT. It's really messy. But I suggest that after you have saved your words: using your favourite editor, split the file into, say, two or more other files by selecting and saving. Let's say we've split and saved the file containing all the letter `S' words into four, each split at the start of a new second letter. That is we have SA.... to SH..... and SI.... to SL... and so on. SORT each of those files. You now have four files each sorted. Now use AmigaDOS JOIN to join all, in the correct order, to a new complete file. Help! you say, I have a large file or files in NO order. OK. Using you favourite text editor split and save your file(s) into size files you think SORT can cope with, say 5000 words. Then SORT. If SORT can cope with files of this size, then back to your editor, taking each file in turn, split them into the same number of files as SORT coped with. For instance, assuming four files. With the first file select say, A to E and save to a new file ` a' and F to L to new file `b', and so on. Now the same for each of the other files, one at a time, only APPENDING or JOINing to `a' `b'`c' and `d'. Now SORT these files. You now have four sorted files A to E, F to L, M to S, and T to Z to be joined, in the correct order, into one file. I told you it was messy, but fun. Yours sincerely, Colin J. Edwards Grenfell NSW  ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 ||| 28 |||