DECmate Information and Frequently Asked Questions File Last edit: 20-JUL-1994 Charles J. Lasner (cjl) This file will attempt to answer most of the questions people, especially new users, have about DECmates, and guide them towards a first cut at getting good performance out of these popular machines. I. What is a DECmate? The DECmate is actually a family name for four products produced by Digital Equipment Corporation between the years 1981 and 1990. Each model has unique characteristics, although a few features span as many as all models and even some predecessor machines. DECmates are an outgrowth of the PDP-8 family of computers and share some measure of compatibility with the PDP-8 series. This allows certain software packages to be run on DECmates even though designed for the PDP-8. Not all packages fair so well, and certain specific systems cannot be run on DECmates for galling reasons. The hardware is just enough incompatible that only certain programming techniques are allowed, generally those found in unsophisticated packages that don't fully exploit keyboard and screen handling as is often practised on DEC systems in general, etc. Towards the end of bridging this gap, certain people are developing replacement software that either runs exclusively on the DECmate series, or perhaps runs generically on all PDP-8's and DECmates. This will be pointed out where relevant. Note that certain aspects of this file may be obsolete as certain packages are updated; the information is accurate as of the time of writing (July 1994). II. Acquiring a DECmate DECmates are available seemingly everywhere. Part of this phenomena is their perceived worthlessness, no doubt caused by some frustration on the part of the (former) owner. Digital caused this problem in large measure by promoting the machine as virtually a dedicated word processor. Additionally, since word processing is largely a matter of preference, DECmate word processing tends to turn off a large percentage of users, who are more likely to prefer packages that run on IBM-PC compatibles. On such machines, word processing variations are available to suit a multitude of tastes from the simple-minded to the most sophisticated. DEC just didn't learn the lesson that in word processing "one size does NOT fit all" etc. An additional factor is likely the lock-in effect in that only DEC's printers are supported, and only then-current models. (Older hardware wasn't supported in newer versions forcing the purchase of newer printers.) DEC's prices on printers have never been competitive with the rest of the industry. For example, a mediocre printer such as an LA-210 could cost as much as an entire IBM-PC compatible computer, yet it had only modest performance when compared to PC-oriented printers such as Epson FX/286 etc. Thus, DECmates as typically sold by DEC quickly became unpopular, overpriced, underperforming, mediocre word processors that were largely replaced by PC's running something like WordPerfect, Microsoft Word, or even Wordstar, or one or more of hundreds of other packages, etc. The fact that they are general-purpose computers was totally ignored. Today, DECmates can often be obtained for free, certainly for a small fee plus the cost of shipping. Often pieces of the machine will be lacking, such as the WPS or other software disks, keyboards, monitors, cables, RX01/RX02/RX50 diskette drives, power supplies, boards, etc. Since there are a lot of DECmate owners/hobbyists/enthusiasts, it is always possible to salvage a DECmate into at least a minimal working machine. Many individuals stock DEC spare parts just to help out someone who is missing a vital part. Prices vary from free to paying for their time, etc. In any case, the ability to support the machines virtually indefinitely seems assured. Thus, the first order of business is identifying what hardware (and software) is present on a DECmate. III. DECmate model descriptions Each model will be given a complete description, including options where applicable. Also, the relative scarcity of the components will be indicated. a) DECmate I The DECmate I, aka VT-278, was originally just known as DECmate. The "I" was added retroactively when the DECmate II was introduced. A standard DECmate I consists of a VT-100 style box with a DECmate-specific motherboard and card cage to house certain options. All of the basic interfaces are built into the motherboard, but there are option boards, some of which are common. A DECmate-specific VT-100 keyboard variant is used that supports more than the the standard four LED's, but is otherwise just a VT-100 keyboard with a DECmate-specific legend. Power supply and monitor considerations are identical to any of the other systems DEC adapted from the original VT-100 design such as the VT-180 (Robin) and various small PDP-11 models. With absolutely no options, the DECmate I is useless as it attempts to work with either a boot device, or at the least act as a terminal with one of the optional serial port boards. DECmate I systems have been located that are as minimal as the basic box, keyboard, and the DP278A dual serial port option. This enables this minimal configuration to act as a quirky variant of a VT-100. A true VT-100 without AVO would have 12 lines in 132 column mode, while adding AVO gives 24 lines. But the DECmate I only has a fixed value of 14 lines. All 80 column modes of the VT-100 are supported to within picky variants which are sometimes lacking even in various other VT-100 systems. Assuming the DP278A (or DP278B) is present, terminal mode is a built-in option. There are two variant modes, one for a fixed terminal, the other for use with a DEC-specific "scholar" family modem. However, the terminal mode can generally be used with other modems (Hayes, etc.). The DP278B is the same as the DP278A, i.e., both are dual serial port option boards, except that the B version adds additional plug-in chips to perform bit-stuff protocols. Apparently DEC was envisioning using the DECmate I as an IBM mainframe terminal and thus wanted to support SDLC protocols, etc. Since this is enormously unpopular, it is extremely rare to ever find the presence of the B option chips, although the DP278A is a common option, and highly desirable. If the DP278 is present, the DECmate is able to communicate with other systems using a variety of protocols ranging from DEC's DX protocol for WPS to Kermit as implemented under OS/78 V4 or newer OS/8 family systems. It is conceivable that certain software packages could fail to function correctly on the B variant. The option is so rare that many programs have never even been tested! For example, Kermit-12 contains initialization code that ought to work on the B version, but it is totally untested. The program works fine on the A version, and tests for the B version changes using the documentation of the DECmate I programming manual only. Whether this is accurate or not is still unknown. Whether the DP278 is present or not, the usefulness of the DECmate beyond being a simple terminal depends on the existence of a disk subsystem. Like the VT78 before it, the DECmate I is compatible with one or two pairs of RX01 or RX02. With regard to RX programming, the DECmate I and VT78 program identically. (In turn, both are slight variants of the PDP-8 version of the RX interface; the only difference is in the ability to support two pairs of drives by a pair select interface not found in the PDP-8 systems. Better software tends to assume the selection hardware is present so that the software runs on all models, but not all software is that well implemented! As an aside, PDP-8 systems with two drive pairs must handle RX systems as two independent devices. DSD-210 systems can handle up to four drives, but address the second pair in ways incompatible with DEC's select mechanism. Yet, some software, such as the P?S/8 RX01 handler, is capable of addressing the second pair of drives whether configured on DSD-210 controllers, VT78 or DECmate I, or even DECmate II, which is is even more incompatible with the original PDP-8 method.) The RX interface is built-in on DECmate I systems, and requires only a specific cable and the drive pair(s). RX drives have been produced in a variety of physical arrays, some of which are to accomodate the specific nature of DECmate I systems, or VT78 systems, etc. Popular variants include rack mount or table-top standard RX01 or RX02, "teacart" style designs where the drive(s) roll around on a caster-based frame, and a specific pedestal design which includes the drives, an integral power controller for all hardware, and a flat workspace on the top designed to hold the DECmate I box. Some variants include a cantilevered keyboard holder and a smaller flat box space. There is a specific large pedestal design that holds two pairs of a special variant RX02 that is not quite mechanically compatible with the rack or table-top RX02. A common variety is the one pair version of the pedestal containing this variant drive. The easiest way to identify the variant drives is the presence of a red stripe across the disk access doors not found on the standard models. These drives are also found in other DEC systems such as PDT-1150, etc. In spite of the mechanical differences, these drives are 100% software compatible with the standard RX drives. When a DECmate I system includes at least a single pair of RX01/02, it becomes a useful system capable of running RX-based software. Of course the limitations of the limited RX hardware apply, but this a lot better than a diskless system! RX02 drives are more versatile due to the larger capacity, but even a single RX01 pair can make for a useful machine, etc. While DECmate RX cables are relatively rare, the pinouts are documented in the available DECmate Programming Manual, and replacement cables can be constructed if necessary. DEC used several variations of round and flat cables and various between-series adaptors on the RX hardware, thus connectors are found that use 40-pin flat cable connectors, DB-25 connectors, and DB-37 connectors. The DECmate I box outputs the DB-37 dual interface connector; it is necessary to obtain the appropriate additional cabling to support the intended drive(s). For example, the single-pair pedestal drives use 40-pin flat ribbon cable internally with Berg connectors. As the cable leaves the pedestal, it plugs into a bulkhead mount adaptor that changes the connections to a DB-25. DECmate I systems then use a DB-37 -> DB-25 cable to make the external connection. A standard option of a table-top RX drive is a DB-25 connector compatible with the same drive cable to the DECmate I box, etc. An obscure option of the DECmate I is the RL278 board which allows the machine to be compatible with RL02 drives. The RL02 drives must be externally mounted either as table-top drives or within one of several large DEC racks. One variant rack requires an RL drive or similar shape device to form the top portion of the cabinet, while other cabinets allow the RL drive to mount more conventionally, etc. The RL278 is designated to be used with RL02 only. It is not known if the RL278 supports a special hardware jumper found on the RL8A controller for Omnibus PDP-8's. This jumper is used with RL01 systems to prevent software from improperly setting a bit reserved for the RL02 system. Also, it is not currently known how to correctly write new software for a system with mixed drives, i.e., how to set this bit correctly to avoid the need for the jumper to be set. Assuming this restriction can be overcome, the DECmate I with RL278 would be as capable as a PDP-8 system to run a mixed RL01/02 system. Until such a time, PDP-8's must strictly run RL01 only or RL02 only with the jumper set accordingly, and presumably DECmate I must run RL02 only, etc. All PDP-8 family machines from the VT78 forward including all DECmates always include a serial printer port. On some models, the baud rate is fixed at 4800 baud. On the DECmate I, the rate is programmable to any one of the DEC standard 16 baud rates up to 19200 baud. The programming of all of the ports is identical except that all DECmate peripheral flag handling is incompatible with the PDP-8 equivalents. (But the differences in the printer port are analogous to the differences in the keyboard handling, not any worse!) Thus the programming of the VT78's printer port is identical to a PDP-8 with the properly configured option board, except that there is a new instruction to set the baud rate to one of the 16 possible values. All of the DECmate printer ports program the same as the VT78 except that the flag handling is different and characteristic of all DECmate interfaces, and additionally, the DECmate III and III+ interfaces have a fixed baud rate of 4800 thus ignoring the rate set instruction. Certain software depends on the programming of the serial port to be as similar as it is across all models. For example, there is a serial port download utility to acquire Kermit-12 for the OS/8 family assuming that Kermit-12 is not yet present, but can be sent to the PDP-8 or DECmate via a serial link to another machine. The utility runs on the printer port of a suitably configured PDP-8, or any VT78 or DECmate without change. Other than WPS and little other software, most of the useful software for the DECmates supports any reasonable serial printer capable of supporting the DC1/DC3 (alternately known as XON/XOFF or ^S/^Q) protocol, not just DEC brand printers. This is merely the hardware restriction of the serial printer port. Most software is otherwise quite flexible in this regard. This makes it much easier to find a printer useful to a DECmate (or to allow another system to act as a printer in the DECmate's behalf!). b) DECmate II The DECmate II is usually considered the best DECmate model due to the large collection of popular options. The screen handling of the DECmate II and up is far superior to the DECmate I which tends to have a "packeting" effect at high update rates, etc. All models from DECmate II up use the same keyboard, a VR-201 proscribed to have WPS keycaps and a blue trim. The DECmate II requires a VR-201 monochrome monitor, while it is the usual option for the III series. The keyboard always plugs into the monochrome monitor. As a terminal device, the DECmate II and up resemble VT-220 terminals. To within limits, the software attempts to conform to this where possible. All DECmate II versions come standard with one RX50 drive pair. An optional RX50 pair may be plugged into the basic chassis requiring only the drive itself and a standard 34-pin cable. (DEC tends to use extremely shielded and molded/formed flat cables, but they are made from standard parts.) 5.25" drives are mounted in the DECmate II using DEC special "skid" plates, but an acceptable variant can be made from a piece of 1/4" thick plastic. The DECmate II uses the most common 4-hole variant of skid, which is in turn a subset of the newer 8-hole version, etc. There is a hard disk option board for the DECmate II which is fairly available, although not common. It is probably one of the few parts of the machine that demands a premium price. A standard cable with a 40-pin Berg connector on one end and two edge connectors on the other (one 34 pin, the other 20 pin) is used to connect a standard MFM disk such as ST-506. The original sizes sold were all Seagate models: ST-506 5 Meg (aka RD50) and ST-412 10 Meg (aka RD51), but eventually the ST-225 (aka RD31) was added, which was actually a 20 Meg disk being claimed as a 10 meg disk. Eventually, standard software supported it properly as a 20 meg disk. Software exists to support up to 64 formatted megs of a larger disk in spite of not being "standard". Thus, Micropolis 1325 (aka RD53) can be used for the maximum size. Stated capacities are somewhat smaller than usual because DEC formats for 16 sectors/track, not the more usual 17 as used in PC-type machines, etc. The Seagate ST-4096 can also be used, wasting the 9th head. There are reports of slightly larger disks being viable (Microscience, etc.) but these are rather obscure drives, etc. Unless the hard disk is cabled out the back of the machine in a non-standard manner, it is expected that the MFM hard disk is to be mounted within the DECmate II chassis. Since this is a full-height opening with a DEC skid plate, the opening can remain blocked, or optionally can be opened to reveal the disk front bezel. If a half-height disk such as the ST-251 is used, then the appropriate full-height bezel adaptor should be used to allow proper air flow, etc. Some of the disks sold by DEC lack a grounding lug. According to documentation, it can be left off. Most disks provide the ground lug, and the DECmate II hard disk cable supports using it as intended. (It appears to be an excuse for an ordering mistake, i.e., DEC mis-specified the ST-225 to be lacking a ground lug; the documentation then claims it really doesn't matter!) Assuming the hard disk is mounted in the chassis, the power supply must be capable of supporting the increased current draw of a stalled hard disk when starting up. A power supply ECO which raises the revision level of the supply was made to the DECmate II and Rainbow to support large disks; this should be used if required. The new-type supply can be recognized by the presence of a stainless steel strap on the left rear of the chassis that is not necessary to mount the original model. Newer chassis assemblies don't support the older type of clasp-lock power supply mounting, but the newer supply can always be added to the older chassis. To prevent rattling, it is recommended that the then unneeded clasp-lock mechanism be glued down, etc. Alternatively, the hard disk can be mounted in an external PC-type powered box, with the cable passed through the rear apron of the machine into the external case. In this case, the supply need not be upgraded. Additionally, this allows the upgrade to a second RX50 pair while the RD51D hard disk option board is also present. If the hard disk is *not* installed, an alternate option is the relatively obscure RX78 option. This board allows the use of one or two pairs of RX01 or RX02 on the DECmate II, possibly in conjunction with the second pair of RX50, thus up to 8 floppy drives can be simultaneously present! Programming the RX78 is slightly incompatible with all previous machines, and most user-written software for RX01/02 won't run here. However, OS/278 does provide handlers for this RX78 versions of the RX hardware, that also run on previous models. It should eventually become possible to support many user-written device handlers for RX01/02 upgraded to support the optional needs of the RX78 interface as well as the subset needs of the VT78 and DECmate I, etc. DECmate II also supports one of two co-processor options: The APU board provides a Z80 with 64K. CP/M-80 and its applications can run from this option. Support includes RX50, hard disk, and possibly RX01 disks depending on configuration. While CP/M-80 is fairly generic, DEC has written or obtained some DECmate-specific applications. (Some are actually only specific to the DEC-type terminal environment, and as such also run on some other machines, such as Rainbow, etc.) WPS software also uses the Z80 option for specific features, such as footnote formatting and DECspell, etc. The other option is the XPU board, which consists of a Z80 and an 8086 with either 256K or 512K depending on which daughterboard option is chosen. It appears that the 512K board is capable of being expanded to 1 MByte but was never sold populated. The Z80 can only address 64K as in the APU version. Apparently due to initialization variations, the generic CP/M install diskette bothers to produce bootable diskettes or bootable hard disk volumes that actually are *not* interchangeable, i.e., must be booted on the hardware type they were created on! (Yet, the install diskette itself runs on either variant!) The APU boards are fairly common, partially due to the usage by WPS. The XPU boards are rather rare, especially the 512K version. Note: WPS doesn't take advantage of any additional hardware not also found on the APU board! While CPM/86 was never released for the DECmate, MS-DOS 2.11 was. Many of the programs are identical to the Rainbow, and indeed some packages geared to the Rainbow do run on the DECmate XPU MS-DOS, most notably SEDT for PC-100 (Rainbow). A few packages require minor patches, etc. An obscure option for the DECmate II is the graphics board. This allows DMA bitmap and other graphics functions to be displayed on either the mandatory VR-201 monochrome monitor, or on the optional VR-241 color monitor, or both. (There are no official color-only DECmate II configurations, although there are rumored cable variations that allow a color-only configuration that is not fully supported by all software. The lack of support could merely be that the color of text could be somewhat at random under some circumstances, but it doesn't appear likely that there could be any other consequence. The only exception is that some software enabled a screen-saver option of the hardware, and it's not clear how this would be affected in such a situation, etc.) The graphics board is capable of outputting the so-called 16 colors as usually defined on IBM-PC color adaptors, meaning 7 colors in two brightnesses, white, and black. On the DECmate III series, the output is only capable of being to up to four colors taken from a palette of 64. There is little software for the graphics board, and apparently none take much advantage of these differences. Additionally, there are "old" and "new" variants of the DECmate II graphics board, but the few software packages apparently mostly support both variations correctly. Occasionally, an older program will be found that only supports the older variant card. There doesn't appear to be a feature difference between the cards, merely a programming incompatibility detail involving status checking, etc. Each DECmate II option demands a dedicated slot (the hard disk and RX78 use the same slot each to the exclusion of the other; the APU and XPU board set plug into the same slot thus preventing the use of both; the graphics card uses a separate dedicated slot and a small two-wire interconnect cable between the board and the motherboard.) DECmate II systems are extremely similar to Rainbow and PRO-325 systems, and share all of the obvious components such as case, chassis, power supply, keyboard, monitor, cables, RX50, hard disk (but NOT hard disk controller!), etc. Thus, defunct systems can often be a good source of spare parts even if not a DECmate! c) DECmate III The DECmate III is the "baby" of the line. It is the smallest physically, and supports only a single pair of RX50 as a disk subsystem. The same keyboard and monitor is used as in the DECmate II. Using an obscure graphics board which is slightly incompatible (described above), it is possible to produce a color one-monitor DECmate III using the same cable as the Rainbow under analougous circumstances. (There are rumors of a cable that allows a dual monitor configuration as in the standard DECmate II with graphics and a color monitor.) Since the graphics board is so scarce, virtually all DECmate III systems will be monochrome. There is also an APU board for the DECmate III which is occasionally seen. (There is rumor of an XPU board that never left the prototype stage.) An obscure option for the DECmate III is a plug-in version of a DEC "Scholar" modem capable of only 300/1200 baud. As in the DECmate II, the comm port is capable of speeds up to 19200 as a standard feature. However, the printer port baud rate is fixed at 4800. The machine is housed in a small metal case that is surprisingly heavy, which is also used in some VAX systems. Power supplies are used that vary from as little as 78W to as much as 150W or more. A basic DECmate III can run with the 78W supply. A strange quirk is that all of the supplies are single-voltage, thus there are 110V and 220V variations of all of them! (DECmate II has a voltage-change switch.) Most software considerations for the DECmate III are such that it is a proper subset of a DECmate II. However, slushware versions on bootable diskettes must be of a higher revision to support DECmate III and III+ systems. Version 422 is generally available to handle all models and is always recommended. (The system test diskette can always upgrade any bootable diskette to the same level as is present on itself. Version 4.5 supports version 422 slushware as distributed, etc.) d) DECmate III+ The DECmate III+ superficially resembles the DECmate III, but contains almost completely different components: There is a new motherboard which is apparently meant to replace the DECmate III board, but never does. There is always present a daughterboard that is a HD controller compatible with the DECmate II HD controller. (There is a rumor that the III+ controller cannot support disks with more than 1024 cylinders, while the II controller can support disks with up to 4096 cylinders. In any case, there are only a few MFM disks with more than 1024 cylinders ever manufactured, etc. Either controller, just as in the Rainbow, can only support up to 8 heads.) The new motherboard clearly has a demised area reserved for the now-obsolete Scholar modem option. The two option slots are precisely the same part as in the DECmate III. Apparently, either slot can house the APU or graphics board in either system; both support both options. All monitor considerations apply as in the DECmate III. The drives are quite different. There is always a half-height 20 MB ST-225 (aka RD31) disk, although using non-standard software, a larger disk can be accomodated. (Half-height disks with at least 40 MB are readily available such as ST-251.) The RX50 drive pair is not used, and is replaced with an alternate hardware controller based on the 8751 (an EPROM chip as opposed to the 8051 masked version. This seems to indicate a very low production run). This version supports one or two TEAC FD55GFV/GFR drives with the jumper set to slow the drive speed to 300 RPM. Thus, RX50 format is still used; the double-sided drive capability and HD capability are ignored. Due to the small size of the case, only one drive may be installed, but with external cabling, the hardware supports two drives using normal drive select of 0 and 1. In fact, certain software such as CP/M-80 actually largely depends on the availability of a second drive. Without this drive, it's impossible to install CP/M-80 on the machine; you must install the software on a hard disk volume on a DECmate II then move the volume from using Master Menu backup, etc. The drives are mounted on an 8-hole skid using the holes behind those compatible with the 4-hole skid of the DECmate II and III. The diskette drive is attached to the hard disk using metal brackets so that both drives attach as a pair. The reason for using the rear holes in the skid is that the diskette drive has the usual TEAC bezel as well as a DEC bezel proclaiming "PLUS" on it. (In fact, this is the only part of the chassis that indicates the machine isn't merely a DECmate III!) The hard disk is not provided with any form of bezel. By removing the DEC bezel and adding a standard Seagate half-height bezel, it is possible to use the 4-hole type skid (or forward holes of an 8-hole skid) while changing the cosmetic look of the drive area. This is required if any form of drive modification is attempted, such as using a 3.5" diskette drive in a 5.25" -> 3.5" adaptor. The power supply of the DECmate III+ is slightly different from the DECmate III supply. Essentially, the only difference is that two 4-prong power connectors are provided instead of the one in the DECmate III version. The supply is rated at 160W while the equivalent DECmate III supply is rated at 150W or less. Presumably the supplies are nearly equivalent, thus a 150W supply equipped with a standard "Y" cable can be used to run a DECmate III+ system board, etc. And of course, each supply is a single voltage source supply (110V, 220V) unlike the dual-voltage DECmate II supply. (Were it not for the revised DECmate II power supply, there would be only one version, not two. However, there are at least 6 DECmate III and DECmate III+ variant supplies!) Master Menu and WPS have been modified in their most recent versions to preclude the requirement of a second diskette drive, but other programs besides CP/M-80 may require a second drive. Work is being done on providing the machine with a second drive or even a pseudo-drive, such as using the second head of the drive as a logical second drive with a special cable, etc. Additionally, the DECmate III+ hardware requires drives that use DEVICE CHANGE, an option normally associated with PC's, while the RX50 uses DEVICE READY. DD 3.5" diskettes use DEVICE CHANGE, thus the possibility of using 3.5" diskettes that appear as RX50! Perhaps someone will even determine a way to install two 3.5" drives in the DECmate III+ chassis! (Also, there are smaller MFM hard disks that are 3.5" half/height. Perhaps the use of a specific set of drives to replace the standard ones will appear, etc.) As in the DECmate III, the printer port is fixed at 4800 baud. For many purposes, the DECmate III+ is essentially equivalent to a reasonably "loaded" DECmate II due to the presence of hard disk, etc. IV. DECmate ROM sets Each DECmate system comes with a characteristic ROM set. The DECmate I has a soldered chip set and is apparently only one revision in all viable machines. Since the ROM is quite small, and additionally there is no slushware add-on capability, there is little built-in support routines for much beyond the built-in terminal emulator, etc. DECmate II systems come with various ROM versions. The version will be reported by running certain software systems, including WPS, etc. Known versions include: i. 31Z (3732) This is the prerelease version found only in very old motherboards. Unless the boards themselves are severely upgraded, the graphics option is not possible to be added. The tell-tale sign of graphics compatibility is a two-prong connector under the board where the graphics board would mount. (It's used to connect to a similar connector on the graphics board itself.) Assuming the board is old, upgrading the ROM set will allow all other options to work. 31Z specifically cannot support the hard disk hardware! Supposedly, DEC promised to provide a free upgrade to any system containing a board this old, yet many have shown up recently, etc. ii. 18K (2213) This is a prototype version used to develop hard disk capable machines that were likely never sold while machines equipped with 31Z were being sold. It is barely capable of supporting all options, although the board it may be placed in could be inadequate hardware! iii. 19H (2310) This version is quite common, but has self-test bugs that fail to find certain known hardware conditions that should be caught, such as pressing a key during self-test, etc. iv. 19N (2316) This is the final ROM version and has no known bugs. DECmate III systems apparently come in one of two ROM versions: i. 32H (4010) This is the usual DECmate III ROM and has no known problems. ii. 32I (4011) This is an obscure version recently discovered. It is not currently known what the difference (if any) there is between this and the previous version. DECmate III+ systems generally have only one version of ROM: 48E (6005). Presumably there are no known bugs, etc. All of the ROM's can be obtained to allow upgrade of any DECmate system except DECmate I. They are all socketed chips. DECmate II systems use three 2716 chips while DECmate III series systems use a single 2764 chip. Additionally, both systems use a single 2764 as a character generator ROM. DECmate III systems use a slightly different character set which at least causes the vertical bar character to appear unbroken while DECmate II systems break the bar as in most IBM-PC systems, etc. While there may be other differences, the two chips have been successfully interchanged with no ill effects, etc. V. DECmate software Software for the DECmates also varies by model. a) DECmate I DECmate I software generally follows PDP-8 guidelines. Due to the DECmate I being the first model of DECmate past the compatible VT78, there wasn't much committment to making software that works on the incompatible hardware. Thus, a lot of then-current software is quite buggy. There are apparently versions of COS-310 that were released allegedly functional on the DECmate I. Undoubtedly some of these releases also claim to be functional on PDP-8 and VT78 systems, assuming all have RX01/02 drives, etc. Since there are no other processors, only PDP-8 software is applicable. The usual software was a version of WPS that could be as high as version 2.0 on 8" floppies. This is quite confusing considering that versions started again at 1.0 on the DECmate II RX50 series! Additionally, there is a clandestine in-house version of *actual* version 2.1 WPS for the RX02 on DECmate I produced at the request of a DEC employee. (This is the same version 2.1 as would be released normally for DECmate II on RX50.) The applicable version of an OS/8 family system for this hardware is OS/78 Version 4. There are many bugs in this system, and largely it is replaced by configuring OS/278 V2 for an RX01 or RX02 boot device. (A small few programs from OS/78 V4 should be retained, etc.) Note: Kermit-12 and its utilities runs on ALL OS/8 family systems by avoiding any hardware-specific dependencies of console handling. (Note: Kermit-12 doesn't support ^C as an exit feature, thus skirting much of the issue!) OS/78 V4 must be used if any attempt is made to use any of the several user-written RX handlers for non-standard format, since this is the last OS/8 system that uses the standard done flag convention for idle/busy detection, etc. OS/278 V2 provides its own handlers that conform to its own quirky flag convention. In the future, OS/8 Version 5 will restore this convention making it possible for existing handlers to again work where possible. (Note: DECmate II hardware is incompatible requiring a rewrite regardless!) Assuming the user has a bootable copy of an OS/8 family system that can run on the DECmate I, it is possible to run the ENCODE/DECODE utilities associated with Kermit-12. This allows obtaining an entire RX01 or RX02 image file as well as individual files. Since often the only device available is a pair of RX01 or RX02, the ENCODE utility can break the device logically into two roughly equal parts to allow even a minimal system the ability to restore an entire disk image, etc. (Note: the ENCODE utility is dependent on the OS/8 handler's ability to read/write the raw image of the relevant data. Thus, the transmitted system image need not be OS/8. However, it may not be possible for the current system to accomplish the transfer depending on the specifics. For example, a P?S/8 RX01 system can be moved using the standard OS/8 RX01/02 non-system handler because both systems use the same internal format for data, each physical sector on tracks 1-77 only is used, and only the first 3/4 of each sector. A WPS8 document disk cannot be moved using this method because it requires that each full sector of the tracks 1-77 be used. If a user-written RX02 handler is available at both ends that does use the entire disk, then this transfer can be accomplished. There are several that support this transfer, one that even accesses track 0, but only for RX01. Additionally, both ends must be using the same user-written handler. As of this writing, there is no equivalent handler for RX02, but there may be a way of working around this, etc.) b) DECmate II The DECmate II can read any of the disks geared to VT78 and DECmate I, but not as boot devices, and only if the RX78 option and RX01/02 are present. Otherwise, files need to be communicated between the DECmate II and another system. The usual medium of software exchange for the DECmate II and up is the RX50 diskette. Many utilities exist that run on IBM-PC compatibles to support RX50 format. Read the separate file "rx50faq.doc" for details, etc. Copies of most of the major software packages for the DECmate II are available in the prevailing archives (sunsite, etc.). At present, a PC-based method is required to access the disk images. Eventually, a more universal handler will be available to allow DECmate II (and up) OS/8 family systems the ability to directly access disk image files with ENCODE/DECODE. At present, only an entire non-bootable OS/278 diskette image can be transferred using this method. (The slushware information will not be copied!) Many of the available utilities are geared to the hard disk subsystem, where Master Menu is recommended. The hard disk itself must be formatted using the System Test Diskette, possibly in conjunction with the "Clandestine Disk Formatter" which is also available. The slushware must be transferred with the Test disk regardless of method; it is automatically transferred to a standard size disk when the Test disk is used to accomplish the format. Note: non-standard formatting causes the Master Menu "X" (compress) option to be non-functional at this time! To accomplish disk data compaction, it is necessary to carry out the laborious diskette-based backup/restore through Master Menu deleting volumes after backing them up. This takes advantage of the fact that restored volumes are created without gaps, etc. Note: CP/M-80 hard disk install has a bug in it. It is not possible to install CP/M-80 on any portion of a hard disk that is located higher than about 10 MBytes into the drive. The best way to avoid this is to either create it on a smaller drive and then use backup/restore to move the entire volume, or alternatively to allocate dummy volumes past the first 10 MBytes to force CP/M-80 to use a lower disk address volume located entirely within the first 10 MBytes of the hard disk, etc. Note: MS-DOS hard disk install has a bug in it. While drive volumes can be specified up to 32 MBytes each, attempts to create bootable volumes larger than 10.7 MBytes cause a crash. Non-system volumes can be as large as the 32 MByte limit. Note: Setting the date and time within MS-DOS actually updates the system clock accessible through Master Menu. When WPS is run from Master Menu, the stupid questions regarding the date and time are faked out by Master Menu rigging canned responses to the queries based on the system clock. However, any attempt to set the date and time from within WPS does not change the system clock. OS/8 ignores the system clock, but it does run correctly and presumably a program can be written to access the system clock to set OS/278's date to it, etc. Note: Master Menu can give initial commands to bootable subsystems. Not all of the systems support this, but WPS and CP/M-80 do. OS/278 does not directly support it, but it's possible to write a program under OS/278 to access the command from Master Menu. Note: Master Menu can allocate up to 8 volumes where the lowest volume is the boot volume. OS/278 V2 provides capabilities to access only the first four volumes. If the MOUNT.SV and RD5x handlers are rewritten, then it will be possible to access the additional four volumes. Note: conversion programs such as WPFLOP can access hard disk volumes. Thus, it is possible to convert OS/278 <-> WPS files without using diskettes by MOUNTing the appropriate volume and accessing it in the WPFLOP command, etc. Note: MS-DOS CONVERT can be given command line options to perform a pre-arranged conversion while the CP/M-80 WPSCONV cannot. This can be useful to allow MS-DOS editing with SEDT followed by conversion to a WPS document followed by booting to OS/278 followed by WPFLOP conversion to an OS/278 text file to allow editing in one system while developing in another! c) DECmate III The DECmate III can use all of the DECmate II software unless it is geared to either the XPU board or a hard disk. (And could require either the graphics board or APU board.) It is necessary to confirm that a bootable diskette has the proper slushware level for DECmate III usage, as many older disks can only boot on DECmate II systems, etc. The System Test Diskette can always be used to correct this situation, etc. DECmate III series comm port is slightly incompatible with DECmate II. Some early code such as the standard OS/278 V2 comm port printer handler cannot run on the DECmate III as distributed. Hopefully this will be corrected. Software that attempts to change the printer port baud rate will be ignored as the rate is fixed at 4800. d) DECmate III+ Generally all of the software for the DECmate II can be used on the DECmate III+ unless it requires the XPU board. APU software is identical for both machines. Note: Most newer packages have been rewritten to allow for the standard DECmate III+ restriction of only one diskette drive. However, CP/M-80 is not updated as such. Thus, to install CP/M-80 requires a modified DECmate III+ that does support a second drive somehow, or a Master Menu backup/restore copy of a CP/M-80 bootable volume from an APU DECmate II may be restored onto the DECmate III+ etc. CP/M-80 requires the second drive in one particular function: the HD utility is essentially a CP/M-80 rewrite of the Master Menu HD utilities. The backup/restore function uses drive B: while the actual Master Menu routine now uses drive 0 (A:). Thus, the backup and restore functions must be avoided in a (normally) one-drive DECmate III+ etc. Disk copying on a DECmate III+ depends on the availability of an RX50-sized non-allocated area on the hard disk, as the source disk is copied onto the hard disk as an intermediary to allow writing onto a target diskette, etc. Note: it has been reported that due to diskette controller chip hardware differences, the DECmate III+ lacks the feature found in DECmate II and DECmate III systems where 48 TPI one-sided disks such as IBM-PC one-sided 160K and 180K diskettes can be read. This feature is actually supported directly on DECmate II MS-DOS, but in theory, a utility could be written for other systems including DECmate III, but not DECmate III+ This is likely attributable to the 8751/8051 chip used in each system to control the diskette drives. It is conceivable that the chips could be switched allowing the drives of the intended system to be used in unintended chassis! (Note: This could mean that DECmate II could then support two drives with DEVICE CHANGE, but loses the ability for MS-DOS to read IBM-PC one-sided disks, and possibly the ability to have two drive pairs! And the DECmate III+ could then support an RX50 drive pair and the ability to read IBM-PC one-sided diskettes!) [end of file]