╘HE ADVERTIZED FEATURES OF THE HEAVY-DUTY ═╙─ DISK DRIVES ARE WELL KNOWN. ─ISK STRUCTURE IS IDENTICAL WITH THAT OF THE ╓╔├-1541. ╘RIPLE BAND CONSTRUCTION REDUCES ALIGNMENT PROBLEMS. ╘HE DESIGN IS INTENDED FOR EXTENDED USE PERIODS WITHOUT OVERHEATING. ┴ LARGER DATA BUFFER (4 TO 6╦) PERMITS MORE OPEN FILES. ╞ORMATTING CAN BE ACCOMPLISHED IN 17 SEC VS 80 SEC FOR THE 1541. ╘HE DUAL ╙─-2 WILL BACKUP A DISK IN TWO MINUTES. (┴ SINGLE DRIVE, THE ╙─-1, IS ALSO AVAILABLE, BUT WAS NOT TESTED.) ╚OW DOES THE ╙─-2 RATE COMPARED TO THESE CLAIMS? ╔N GENERAL VERY WELL.
╘HE ╙─-2 CONSISTS OF TWO VERTICALLY MOUNTED DISK DRIVES IN A SINGLE HOUSING. ╘HE DRIVES ARE OF THE LEVER LATCHING VARIETY, WHICH SEEMS TO AVOID SOME OF THE IMPROPER DISK SEATING OBSERVED WITH THE SPRING LATCHING MECHANISM OF THE 1541.
╘HE DRIVES OPERATE AS DRIVES 0 AND 1 INSIDE A SINGLE SERIAL ADDRESS DEVICE (8). ╘HIS MAKES POSSIBLE THE ╨╥╔╬╘#15,"─0=1" DISK DUPLICATION ROUTINE OF THE 1541 ─╧╙, WHICH DOES INDEED RESULT IN BACKUPS IN UNDER TWO MINUTES. ╒NFORTUNATELY, THE DUPLICATE ROUTINE CRASHES IF THE DRIVE ENCOUNTERS ANY ERRORS.
─UPLICATION ABILITY OF THE ╙─-2 IS STILL COMMENDABLE IN TERMS OF SPEED AND CONVENIENCE WHEN COMPARED TO A SINGLE 1541 OR TWIN 1541'S WITH READILY AVAILABLE SOFTWARE. ╥ESULTS OF A SPEED TEST SERIES WITH A TYPICAL FULL DISK (╫╔├╒ #8) ARE SHOWN IN ╞IG. 1.
╘HE DATA SHOW THAT THE DUPLICATE ROUTINE IS FASTEST WHEN IT CAN BE USED. ┴ BASIC, SINGLE BUFFER COPY PROGRAM FOR THE ╙─-2 IS LISTED IN ╞IG. 1 AS ═─ COPY. ╘HIS PROGRAM WAS CONSIDERABLY SLOWER THAN THE DUPLICATE ROUTINE, BUT WAS STILL MUCH FASTER THAN ALL AVAILABLE TWIN DRIVE ROUTINES.
╥ESULTS OF THE TESTS SUGGEST THAT THE MAIN ELEMENT IN THE DUPLICATION SPEED OF THE ╙─-2 IS THE USE OF A SINGLE DATA BUFFER IN THE DRIVE. ╔N MOST DISK COPY PROGRAMS WITH THE ├OMMODORE 64, A BLOCK OF DATA IS READ FROM THE DISK AND TRANSFERRED TO THE DISK BUFFER. ╘HE DATA ARE THEN TRANSFERRED FROM THE BUFFER TO THE ├-64 MEMORY ONE BIT AT A TIME--SOMETIMES USING A VERY SLOW BASIC ╟┼╘# ROUTINE, BUT MORE COMMONLY USING A MACHINE LANGUAGE ROUTINE.
║AI+0
╞IG. 1. ├OMPARISON OF ─ISK ─UPLICATION ╙PEEDS WITH ╫╔├╒ #8
╬OTES--* ╙INGLE BUFFER BACKUP PROGRAM MODIFIED TO INCLUDE FEATURES OF ┬ACKUP 228.
║MA-12
** ╨UBLIC DOMAIN SINGLE OR TWIN BACKUP PROGRAM INCLUDING ERROR SKIPPING AND FORMATTING WITH ╔─ READ FROM BLOCK HEADER.
║MA-12
*** 1541 ┬ACKUP MODIFIED FOR TWIN DRIVE OPERATION WITH ERROR SKIPPING.
║MA-12
**** ╘HE ╙─-2 DRIVE WAS SERIALLY CONNECTED. ┴ PARALLEL INTERFACE IS AVAILABLE FOR THE DRIVE.
║AI+5
║LM10
╘O WRITE THESE DATA TO THE TARGET DISK THE PROCESS IS REVERSED. ╘HE DATA ARE AGAIN TRANSFERRED TO A DRIVE BUFFER ONE BIT AT A TIME AND THEN WRITTEN TO THE DISK. ╘HE ╙─-2 AVOIDS THE TRANSFER OF DATA TO THE ├-64 MEMORY AND BACK BY HAVING THE SAME DISK BUFFER AVAILABLE TO BOTH DRIVES.
╘O TEST THIS HYPOTHESIS, A 121 BLOCK PROGRAM (╨APER├LIP 64├) WAS TIMED DURING A LOAD WITH BOTH THE ╙─-2 AND A ╓╔├ 1541. ╥ESULTS WERE:
═╙─ ╙─-2 1:25 MIN
╓╔├ 1541 1:18 MIN
║MA+0
╘HE RESULTS INDICATE THAT A PROPERLY ALIGNED 1541 MAY ACTUALLY BE FASTER THAN AN ═╙─ DRIVE. ╚OWEVER, IT HAS PREVIOUSLY BEEN FOUND THAT AN OUT-OF-ALIGNMENT 1541 CAN BE MUCH SLOWER DUE TO ITS LOGIC SEEKING HEAD ROUTINE.
┴ QUESTION OFTEN RAISED WITH THE ═╙─ DISK DRIVES IS THAT OF COMPATIBILITY WITH COMMERCIAL SOFTWARE. ╙UPPOSEDLY, SOME HEAVILY PROTECTED SOFTWARE IS DEPENDENT ON THE 1541 ╥╧═ AND WILL NOT LOAD PROPERLY WITH THE ═╙─ DRIVES. ┴ SERIES OF COMMERCIAL PROGRAMS WAS TESTED FOR COMPATIBILITY. ╔N THE MAJORITY OF THE CASES TESTED, THESE FEARS WERE GROUNDLESS. ╧F COMMON PROGRAMS, ONLY ╙UPERCLONE COULD NOT BE LOADED SUCCESSFULLY WITH THE ═╙─. ┴ LIST OF PROGRAMS FOUND TO LOAD SUCCESSFULLY IS SHOWN IN ╞IG. 2.
║AI+0
╞IG. 2. ╙OFTWARE ╞OUND TO ╠OAD ╙UCCESSFULLY WITH THE ╙─-2 ─ISK ─RIVE
┴ MORE SERIOUS DIFFICULTY OCCURRED WITH STANDARD BACKUP PROGRAMS. ═OST OF THESE WILL NOT ACCOMMODATE DUAL DRIVES, BUT ARE INSTEAD WRITTEN FOR TWIN DRIVES. ╔NDEED, THE ONLY KNOWN DUAL DRIVE COMPATIBLE SOFTWARE IS ╨APER├LIP ╫ORD ╨ROCESSING. ═OST OTHER SOFTWARE DOES NOT USE TWO DRIVES. ┬ACKUP PROGRAMS WRITTEN FOR THE ═╙─ DRIVES ARE BEGINNING TO APPEAR, AND SOFTWARE SUPPLIERS SEEM WILLING TO CERTIFY THAT THEIR PROGRAMS WILL LOAD ON ═╙─ DRIVES. ╘HEREFORE, THE DRIVE COMPATIBILITY PROBLEM WITH THE ═╙─ DRIVE APPEARS TO BE TEMPORARY.
╔F THE ═╙─ DRIVES HAVE A WEAKNESS, IT IS PERHAPS THE MANUAL PROVIDED WITH THE DRIVES. ╫HILE THE MANUAL SUPPLIED IS WELL WRITTEN, AND SEEMS TO BE FREE OF ERRORS, IT DOES NOT INCLUDE THE DEMONSTRATION DISK FOUND WITH THE ╓╔├-1541, NOR DOES IT INCLUDE THE MANY EXAMPLE PROGRAMS IN THE 1541 MANUAL. ╞OR THESE REASONS, THE ═╙─ DRIVE MAY NOT BE THE IDEAL FIRST DRIVE FOR THE BEGINNING USER. ╔T IS, HOWEVER, AN EXCELLENT CHOICE FOR A SECOND DRIVE, OR FOR THE EXPERIENCED USER.
┴ SECOND WEAKNESS OF THE ╙─-2 MAY ACTUALLY BE AN ADVANTAGE. ╘HE ╙─-2 APPEARS TO BE LESS FLEXIBLE IN ITS ABILITY TO READ A DISK. ┬ECAUSE THE 1541 DRIVE IS LOGIC SEEKING, DISKS WRITTEN ON AN OUT-OF-ALIGNMENT DRIVE WITH THE 1541 CAN OFTEN BE READ WITH A PROPERLY ALIGNED 1541. ╘HE ╙─-2 SEEMS MORE SENSITIVE TO THIS ASPECT. ╙OME DISKS WRITTEN ON A 1541 CANNOT BE READ ON AN ═╙─. ╘HIS POSSIBLY SUGGESTS A NARROWER READING RANGE FOR THE ═╙─. ╧F COURSE IF YOU HAVE BOTH A 1541 AND AN ═╙─ DRIVE, COPYING THE DISK TO THE ═╙─ DRIVE IS POSSIBLE, AND THE PROBLEM WITH MIS-WRITTEN DISKS CAN SLOWLY BE CORRECTED.
┴ FINAL AREA YET TO BE EVALUATED IS THE LONG TERM RELIABILITY OF THE ╙─-2. ╥ESULTS SO FAR SUGGEST NO DIFFICULTIES WITH OVERHEATING, BUT ONLY TIME WILL TELL IF THE ╙─-2 IS TRULY RELIABLE.
╔N SUMMARY, THE ═╙─ DISK DRIVES APPEAR TO BE A GOOD SOLUTION TO THE RELIABILITY PROBLEMS OFTEN EXPERIENCED WITH THE ╓╔├-1541. ╘HE ═╙─'S AND IN PARTICULAR THE ╙─-2 ARE RECOMMENDED FOR THE SERIOUS ├OMMODORE 64 USER. ╘HE NOVICE MAY NOT WANT TO BEGIN WITH AN ═╙─ AT PRESENT BECAUSE OF SOME SOFTWARE AND INFORMATION LIMITATIONS. ╞OR A SECOND DRIVE, THE ═╙─ IS AN EXCELLENT CHOICE.
