Not only did I have a CP/M machine back in the day (which I truly loved), but I also owned an MP/M machine for a while too. For those of you unfamiliar with MP/M, it was a variant of CP/M that allowed for memory bank switching. When multiple programs were run, they were each loaded into (as I recall) a 16KB bank of memory that could be switched on and off. When switched on, the memory was overlaid with the other switched banks, meaning that only one bank at a time could be enabled. When it was time to give some CPU cycles to a program, the processor would save the context of the currently running application, turn off its memory bank, restore the saved context from the next app to schedule, switch on its memory bank, and start executing. It was a very cool way of multitasking, even though it required specialized hardware and was not memory efficient (memory banks could not be shared). Anymore these days MP/M looks like a dinosaur, but at the time it was pretty cutting edge.
Follow up: I implemented CP/M, MP/M & ConcurrentDos-386 on hardware I designed/modified from 1979 to 1994. I also wrote & sold an MP/M-80 disk performance enhancing product that sped up disk I/O from 2 to 4 times. In addition I modified the CP/M-80 ver 2 & MP/M-80 ver 1 to access drives up to 512 megabytes (8 megabytes was the max for both factory shipped versions, MP/M-80 ver 2 supported 512 megabyte drives as shipped).
The hardware independence of the BIOS didn't just allow for differing floppy formats but in at least one case the use of a different medium altogether. The only requirement from the BIOS was that storage was directly accessible via track and sector. Unlike other home computers of the day the Coleco Adam used a track and sector format for its cassette tape drive. Each data pack, the term used for their cassette tapes, consisted of 2 tracks each containing 128 256 byte sectors. When Coleco released CPM for the Adam they did so on both floppy, floppy drives were available for the Adam, and tape. You could boot CPM off the tape drive and then access the tape drives as drive A and B as if they were floppy drives. I believe the Coleco Adam was the only CPM system which could boot from tape.
@@THEPHINTAGECOLLECTOR It allowed for random access via tape but it was very slow. Most people who used CPM on the Adam would have owned floppy drive units. I think Coleco asked "can we?" when the correct question to ask was "should we?", but that about sums everything about the Coleco Adam.
@@tss20148 would be interesting to play around with it anyway :) Just making my time wirh CP/M on rhe C64. This one would certainly qualify into questions like „should we“ and „but why?“ as well.
@@tss20148 Sounds like a really obscure thing. Just reading through online. It seems a good example as to why certain products failed for good. Still, these things intreague me. Would love to see it in action.
In the book, "They Made America", there is an entire chapter devoted to Gary Kildall. Tim Patterson tried to sue over it, but it was thrown out (the book even covers that). It pointed out something I had not heard anywhere else. The synopsis was that DR had a most favored nations agreement with their top licensees. If DR gave another licensee better pricing, DR would have to go back and give the same terms to the other MFN companies. For what IBM wanted, $50,000 flat rate with no royalty per copy (or whatever it was), that would certainly cause DR to have to refund a lot of money. That was the major reason Kildall couldn't accept the deal from IBM. Either that or Gary was doing some stunt flying while IBM was waiting were waiting patiently at his offices in Monterey.
another less important issue at the time was that smaller software houses did not want to work with IBM as they were seen as Corporate America ( Big Software if you like) that could crush everyone..and they wanted to stay independent..Microsoft was considered just a large independent software house at the time specializing in languages i.e. MASM, Basic etc DR specialized in Operating Systems...the 3 top hardware manufactures were Tandy, Apple, Commodore..and software companies bet on those 3 being the future... At the time i was a manager at a Tandy Computer Service Center..Tandy was betting on TRSDOS or CP/M and Xenix for businesses and did not see what was coming with IBM and MSDOS That delay of reacting to the market virtually killed Tandy (they did release a PC compatible but never regained their market share), Apple came close to bankruptcy, Commodore by the late 80s they were virtually done...so in a sense IBM/Microsoft did crush them...
Interesting video, thanks! I think you've already checked this out, but there's a series (2?) of videos on this topic where the creator meets DRI employees. They go into some detail about the terms IBM wanted in their contract. It was a bit unfair, typical corporate style.
Anyone seriously thinks a teenager BillG would have put IBM to their knees alone, without any help? That's like SteveJ Apple own stories, pure fantasy hollywood style brand propaganda. Bill Gates wasn't alone when he went to negotiate with IBM. He had all the support of dad and mom, who where working with the US government at time, along with the best lawyers Gates family could pay. TL;dr, Gates family has been always part of the club, which would allow negotiations between "equals". In contrast, Gary Kindall was what you could call a "new rich". The Steve Wozniak type. More of an engineer than a business man, and ofc he wasn't part of the club. He only wanted to sell his software, and never thought about "20 or 30 years plans" in the future. Since he wasn't part of the club, he couldn't negotiate "between equals" with IBM... And even he surely had at time the money to contract a better lawyer specifically to advice him in the IBM negotiations, as a "day living" man he didn't thought it was necessary. In the sense was a lot like HP in the day: "Who would take seriously Apple's and future IBMs? Why the average Joe would need a computer, at all?"
Really interesting video, i have used CP/M back in the early 90s but havent touched it since around 1992 so it was nice to see this video and see the history of the OS. I do like your historic OS videos like this. You explain them well and easy to understand.
CP/M actually lasted quite a long time. The Amstrad PCW, a Z80-based, low-cost business machine used CP/M, and was extremely successful in the UK and elsewhere in the 1980s and '90s, outselling PCs for a while and eventually became the best selling British computer ever. Amstrad's smaller home micros, the CPC range (which was also commonly used in small businesses) and the Sinclair Spectrum +3 also could run CP/M.
I grew up with APPLE ][e+/][e and IBM PCs. CP/M felt kind of like a less useful and clunky MS-DOS (Yes MS-DOS copied lots of CP/M features). Apple DOS/Prodos were extremely functional for the apples and MS-DOS had a major advantage over CP/M which was the introduction of standard file formats and other handy things like a fully working piping system. Also directories were really handy. I started on DOS v2.01a. DOS in terms of usability for the user was around the same until V3.3. Some extra commands were added. The first quality of like commands I really recall was Echo off and the @ symbol. The @ symbol meant you could finally 100% seamlessly make a batch file look like an executable. There's a nice superuser post about DOS 2 being the first to include pipes,directories and Daemons/background execution. DOS 3 added in a bunch of network stuff, better disk stuff (bigger HDDs) and some more odds and ends. CP/M didn't feel like it was really growing but I never really used it as it was the "for business" OS.
I spent my high school years and college years writing papers on WordPerfect for Apple II, which wasn't as functional as it's DOS counterpart. Anyways, I think I missed out a little because AppleDOS and ProDOS just got out of the way. In a way, MS and PC-DOS forced users to learn a little because of all the settings, the rewriting of AUTOEXEC.BAT and conflicting TSRs, etc When you could do these things there came a certain pride and sense of accomplishment. By the time I got into Macs, I was "spoiled", and expected all my software to just work, which for the most part it did. When you think of it, users should have been blissfully unaware of the OS and only know the computer through the applications thye used. DOS, then Windows kinda changed that, and becoming the office computer expert meant something.
The problem with the lots of MFM disk formats forced Commodore to add the WD1770/71 floppy controller into the 1570 and 1571 drives to be useable with CP/M on the C128. The C128DCR combined the C128 and the 1571 into one PCB, so it is one of the few boards with three different processors: 6502, 8502 and Z80. But Commodore also combined a 6526 and the WD1771 into the disk controller 5710. It required a new CBM DOS 3.1 which wasn't as compatible as DOS 3.0 of the original 1571. But I don't know how it affected CP/M compatibility.
I grew up with Amstrad CPC computers (CPC464 with DDI1 disk expansion) and then later with the CPC6128. Both had CP/M available for it. V2.2 on the 464 and V3 (aka CP/M+) on the 6128. Used it mainly for Infocom text adventure games, sometimes with terminal software to try out BBSs.
At least DR-DOS 5 or 6 (which came with an EMM386 like memory manager, transparent file compression and many more utilities) gave a good scare to microsoft as their MS DOS 4 and 5 were VERY lacking in features. Then windows came and all this was lost. Windows, the only non Unix (or Unix like) mainstream OS, hope Dave Cutler is happy. Best wishes.
Well, to this day he tells people he hates Unix, so... There's a little more to the story, since Windows 3.1 ran on top of DR-DOS just fine, although the installer warned against it. There was a settlement for hundreds of millions for Microsoft's AARD bug which pretended DR_DOS (and other DOS clones) weren't compatible. By that time Caldera owned DR-DOS. Anywho, if installing Windows 3 on top of DR-DOS caught on, history would be different. And I'm kinda glad it didn't, because that brought about "Free and Open Source Software" and especially Linux. If it not for Windows, we wouldn't have needed an open source movement, or at least it would have been delayed many years.
@@squirlmy Part II will go a bit into the AARD story, not really for the technicalities but for the overall story on the Caldera vs. Microsoft lawsuit.
As I remember it when I was at DR, Concurrent-DOS was multi user too. You could have up to 4 serial terminals attached to a pc. The product names was not as complex as you make it as much was under license for individual companies to ship with their hardware. Each company was driven by their hardware and the intended usage model. That then naturally filtered them down to the products that would run and the level of support they wanted to offer their customer in the whole package. It only seems confusing if one assumes DR sold its operating systems as Microsoft does today, namely direct to end-users. Until the Amstrad 1512/1640 hit the UK market and the GEM environment, very little was sold to end users. At this point DR had to ramp up software duplication, distribution and sales direct to consumers. (Also for the other Amstrad machines.) All my observations are related to my time at DR in the UK.
First-hands information is always the best. As mentioned, I meant no offense here. It's just an outside point-of-view perception, and not all such details may be readily available in online sources. Thanks for filling that in!
@@THEPHINTAGECOLLECTOR No problems, but If you jump from this directly to Novell (I had left by then); you miss out the massive transition GEM and Amstrad made for DR at the time. At that time the company created, a you may not believe this but, a camera mount video box which allowed GEM Graph to be photographed for slides to use in executive presentations... ;) Also GEM publisher was well ahead of its time. I would play with the advanced tools on the Amstrad 1640 that I had for support. Was a great time.
@@THEPHINTAGECOLLECTOR Oh yes before my time as tech support I worked in accounts and also typed up the legal documents for the licensing contracts with the OEMs. I trained as a secretary; but it made my programming later very fast. All 10 fingers :)
@@THEPHINTAGECOLLECTOR I can't find it (only a dead link on google). I never really grasped the actual history of dr dos and cp/m with all different versions and how they were related. The document you found was very nice. It is, of course, not a big deal if you can't find it again, but if you happens to find the link it would be nice. Very nice video, as your other videos!!! Really good work!
Funny, how much CP/M I have had in my hands over the years, but never utilized. I had a Z80 Softcard for my Apple IIe, didn't have the software to use it. Coleco Adam with CP/M datapack. Yes, that would be CP/M on cassette tape, I really had no use for this. Commodore 128 with the CP/M disk and utility disks. I booted it, and that was it. I saw no reason to fetch CP/M applications, when everything would have a C128 mode version. I would someday love to have an Altair 8800 to play with, then I would be forced to uses CP/M and you know it would feel natural. As for the other systems, it felt it was tacked on and unnecessary.
I can definitely recommend to go into an Altair. I don't have the real thing but an Arduino-based replica, the AltairDuino. It's really close, just that it's missing the S100 bus. But good enough to experience it (almost) the way it was 50 years ago.
D'er Doss? I think you'll find it's pronounced "Doctor Doss". Anyway, while I don't hear this said today, back in the 70's and 80's it was widely known that CP/M was originally just written so a disk drive manufacturer could demo their product. After popular demand, it got expanded. I had CP/M on my TRS-80 COCO 2 and unmodified TRS-80 Model I (with expander). LDOS was better.
Yeah, I whitnessed it actually being called "Doctor DOS" by some people here in Switzerland back in day, mistaking the "DR" as being "Dr" (or "Doctor" in short), instead of being the initials of "DR" = "Digital Research". It may well be that in other languages, even english, it might have been sometimes mistakenly called "Doctor DOS". I wasn't there to whitness. But if anything, that convention was IMO more by mistake and not by intention of the vendor. All video formats I ever saw where DR DOS was mentioned, only ever refer to it as "DR DOS" and not "Doctor DOS". I stick myself as well to mentions in the Computer Chronicles, like here: th-cam.com/video/CKf7Oc60fVM/w-d-xo.html Stewart Cheifet calls it "DR DOS" as well, and I believe, as someone who worked closely with Kildall on a regular basis, they would certainly have talked about DR DOS on some occasions. Had Gary Kildall called it "Doctor DOS", then I'm certain Cheifet would also have called it "Doctor DOS". Btw, good point about the disk drive. Kildall mentions this in his memoirs, how he wrote CP/M on an emulator first, and struggled how to get it running for a Memorex floppy drive, as he wanted to add mass storage support for PL/M. There's ever so much room for facts in those videos, and it's always a challenge, what to include, and what not.
In my opinion, the reason why MS-DOS forced IBM compatibility on everyone is due to Tim Patterson's faulty function 9 in the DOS API. Function 9 simply prints a string terminated by a dollar sign in MS-DOS. In CP/M, it has an escape code language to place the cursor in different positions on the screen, to clear the screen, set the colors, etc. That why you can have full-screen word processors like Wordstar that can run on any CP/M computer. In MS-DOS, you had to use a BIOS call or access video ram directly to do any of that. That's why computers like the Tandy 2000, which was MS-DOS compatible but not IBM PC compatible, failed miserably. The Tandy 2000 couldn't run anything complex that had to be designed around the IBM PC memory map. It almost seems like this oversight was intentional.
CPM had the same issue with software requiring to be patched for each machine. There were MSDOS compatible machines that had WordStar, dBASE etc. So I'm not sure it was on purpose. I don't think anyone thought the PC would even be a big deal. Everyone would just keep their PC for a while and then throw it out, along with their software, like they had done already.
@@phill6859 The software only needs to know the underlying terminal or terminal emulation codes, the rest is hardware independent from the perspective of the user program. And if you don't need to move the cursor around you can even ignore that. I am working on a little emu board that emulates computers from the 70s on a, currently, RP2040 and can say that most CP/M programs have several terminal types built-in, the most used where ADM-3A/31, the VT52 and the VT100 and compatible ones. The C128 emulates a ADM-31 with a color expansion for example. Most modern implementations will be attached to a PC that uses VT100/ANSI, my little boards software can translate between them by the way.
WHAT? Forced compatibility? Compaq cleanroom reverse engineered the BIOS, (no the first) to get to as close to 100% compatibility as possible. They had sales of $111 million the first year, their first year of existence, with the clunky "Compaq Portable", breaking all records for a first company. Compaq got together with 8 other companies (sometimes facetiously referred to as the Gang of Nine), name the PC bus "ISA Bus" wrote a plan for it to be succeeded by EISA, and I believe Phoenix was there, offering their IBM-compatible BIOS. Before Compaq, Microsoft very happily made many MS-DOSes incompatible for "IBM Workalikes", including offerings by Tandy/RadioShack, Xerox, Hewlett-Packard, Digital Equipment Corporation, Sanyo, Texas Instruments, Tulip, Wang, Olivetti... The Tandy 2000 didn't die of being IBM compatible (note the compatibility or lack of, was with IBM "PC-DOS") Very quickly Tandy made the "1000", compatible with PCjr, not the XT or AT, and by all rights they could've set the standard with Microsoft's blessings. They made sales blunders, not design mistakes. IBM and only IBM was looking to set the standards, not Microsoft. AFTER Compaq and Phoenix, clones became leaders and wiped the floor with IBM. Before that it was a gamble which 8088/86 machine would set the standard, and if it weren't Lotus 1-2-3 and WordPerfect courting IBM, the biggest and most monied, things could have gone differently. The thing about the IBM PC is that it was made with industry-standard components, and IBM let Microsoft license DOS to anyone else. There was no "forcing". If Apple didn't sue and win against Franklin in 1984 for copying their boot firmware, and custom parts in the Apple lIe, ironically that might have become the PC standard, although Apple might have lost like IBM. The world's industries WANTED compatibility in software, whether tech companies wanted it or not.
@@phill6859 No "forcing" at all. Really smart people might have figured it out, and I suppose they did. But also remember IBM was firstly and importantly a minicomputer and mainframe company; they didn't want the PC cannabilizing that market (it did anyways). So they made decisions like letting Microsoft give them a crappy CP/M imitation, giving it an 8-bit bus and a crippled 16-bit CPU, and most importantly, letting Microsoft give licenses of MS-DOS to other companies. Also using off-the-shelf parts (except for the BIOS of course) None of those decision make sense, until you realize most of the executives and sales force would rather have another terminal and not anything to compete. I just heard an interview with the Tandy person in charge of their PCs, and he said "business machines made something like 20% commission while home computer sales could be 80%" the main reason to have a dominant office PC, was to leap into the home market from there. IBM tried with PCjr. The IBM "clones" won the latter.
It's more complicated. People had been begging for cpm86 since 1978 when the 8086 came out. Gary could have ported it in a month or two, he did not. Bill even told IBM to talk to Gary, that still didn't work. Eventually they had to go back to Bill and he told IBM they could license 86dos. IBM told Bill they would prefer if he was responsible for licensing it and doing the modifications Nobody (probably even Bill) thought IBM would be priced out of the market so quickly or that the PC would endure so long.
@@phill6859 and IBM really played dirty offering CP/M for 199 and DOS for 49 bucks (I believe the prices were). Anyways Microsoft and IBM are horrific untrustworthy companies. And I worked for big Blue 🤮
@@phill6859 yeah, the success of the IBM PC was surely a surprise. But from what i can gather from history it was all thanks to the work of Compaq and their balls against IBM creating a compatible BIOS, the only thing not off-shelf in a IBM PC. After that it was off to the races as everyone and their dog could build a PC and sell it. And while in the 80s Commodore for example had a huge success with the C64 as the computer for the masses and a good success with the Amiga as a high-end machine they rested on their laurels and innovated too late and too little. Apple somehow succeeded with the something-with-media crowd, Sinclair couldn't replicate their success with the Spectrum as a low-end machine and Atari, while it had some fans with the ST, fell to the wayside too. Not mentioning all the other small computer makers here. And the nail in the coffin for everyone but the PC was Doom. Quake cemented that.
@@phill6859 I think IBM didn't want to negotiate with Gary, who would have had the upper hand in that exchange. They were already on good terms with Bill and his BASIC software business, and I'm sure they figured they'd make a technically superior OS if the PC caught on. And they did eventually create OS/2, and they were caught flat footed when MS Windows destroyed the OS, that people preferred the lower cost ISA bus computers from companies like Compaq, (and companies with Phoenix's BIOS), over their technically superior MCA bus PS/2s. I think people overstate Kindall's arrogance, he just wasn't a good businessman. The collective borg of IBM were the arrogant ones. After all, Kindall had CP/M-86 ready for the PC rollout, IBM just decided to screw him with a $240 price tag.
@@squirlmy When reading through all possible sources, I came to the same conclusion. Kildall was more a techie and engineer seeking for the next technological step, and not so much the business man. As opposed to Gates, who was more the business oportunist. Anyway, the IBM price tag was really cheeky, especially when considering that at least all sources I found mention CP/M selling for 70 USD (according to Kildall's memoirs), 80 USD (Wikipedia), and somewhere around 80-100 $ when looking at ads, or when putting known OEM details into perspective. The most expensive version I saw was the adapted CP/M for the TRS-80 Model 4, which sold for 169$.
![BOWKYLION - วิงวอน (ex-change) [Official MV]](http://i.ytimg.com/vi/cLdnZWDaO-w/mqdefault.jpg)
Not only did I have a CP/M machine back in the day (which I truly loved), but I also owned an MP/M machine for a while too. For those of you unfamiliar with MP/M, it was a variant of CP/M that allowed for memory bank switching. When multiple programs were run, they were each loaded into (as I recall) a 16KB bank of memory that could be switched on and off. When switched on, the memory was overlaid with the other switched banks, meaning that only one bank at a time could be enabled. When it was time to give some CPU cycles to a program, the processor would save the context of the currently running application, turn off its memory bank, restore the saved context from the next app to schedule, switch on its memory bank, and start executing. It was a very cool way of multitasking, even though it required specialized hardware and was not memory efficient (memory banks could not be shared). Anymore these days MP/M looks like a dinosaur, but at the time it was pretty cutting edge.
Follow up:
I implemented CP/M, MP/M & ConcurrentDos-386 on hardware I designed/modified from 1979 to 1994.
I also wrote & sold an MP/M-80 disk performance enhancing product that sped up disk I/O from 2 to 4 times.
In addition I modified the CP/M-80 ver 2 & MP/M-80 ver 1 to access drives up to 512 megabytes (8 megabytes was the max for both factory shipped versions, MP/M-80 ver 2 supported 512 megabyte drives as shipped).
The hardware independence of the BIOS didn't just allow for differing floppy formats but in at least one case the use of a different medium altogether. The only requirement from the BIOS was that storage was directly accessible via track and sector. Unlike other home computers of the day the Coleco Adam used a track and sector format for its cassette tape drive. Each data pack, the term used for their cassette tapes, consisted of 2 tracks each containing 128 256 byte sectors. When Coleco released CPM for the Adam they did so on both floppy, floppy drives were available for the Adam, and tape. You could boot CPM off the tape drive and then access the tape drives as drive A and B as if they were floppy drives. I believe the Coleco Adam was the only CPM system which could boot from tape.
@@tss20148 Cool, I didn‘t know that.
So it did implement tandom access via tape?
I wonder how quick or slow that was in peactice.
@@THEPHINTAGECOLLECTOR It allowed for random access via tape but it was very slow. Most people who used CPM on the Adam would have owned floppy drive units. I think Coleco asked "can we?" when the correct question to ask was "should we?", but that about sums everything about the Coleco Adam.
@@tss20148 would be interesting to play around with it anyway :)
Just making my time wirh CP/M on rhe C64.
This one would certainly qualify into questions like „should we“ and „but why?“ as well.
@@THEPHINTAGECOLLECTOR No matter Commodore's sins they at least didn't put the power supply for the whole system inside the printer like Coleco did.
@@tss20148 Sounds like a really obscure thing. Just reading through online.
It seems a good example as to why certain products failed for good.
Still, these things intreague me. Would love to see it in action.
In the book, "They Made America", there is an entire chapter devoted to Gary Kildall. Tim Patterson tried to sue over it, but it was thrown out (the book even covers that). It pointed out something I had not heard anywhere else. The synopsis was that DR had a most favored nations agreement with their top licensees. If DR gave another licensee better pricing, DR would have to go back and give the same terms to the other MFN companies. For what IBM wanted, $50,000 flat rate with no royalty per copy (or whatever it was), that would certainly cause DR to have to refund a lot of money. That was the major reason Kildall couldn't accept the deal from IBM. Either that or Gary was doing some stunt flying while IBM was waiting were waiting patiently at his offices in Monterey.
another less important issue at the time was that smaller software houses did not want to work with IBM as they were seen as Corporate America ( Big Software if you like) that could crush everyone..and they wanted to stay independent..Microsoft was considered just a large independent software house at the time specializing in languages i.e. MASM, Basic etc DR specialized in Operating Systems...the 3 top hardware manufactures were Tandy, Apple, Commodore..and software companies bet on those 3 being the future...
At the time i was a manager at a Tandy Computer Service Center..Tandy was betting on TRSDOS or CP/M and Xenix for businesses and did not see what was coming with IBM and MSDOS
That delay of reacting to the market virtually killed Tandy (they did release a PC compatible but never regained their market share), Apple came close to bankruptcy, Commodore by the late 80s they were virtually done...so in a sense IBM/Microsoft did crush them...
Interesting video, thanks! I think you've already checked this out, but there's a series (2?) of videos on this topic where the creator meets DRI employees. They go into some detail about the terms IBM wanted in their contract. It was a bit unfair, typical corporate style.
Indeed, there‘s so many videos on the topic, it was actually told many times.
Anyone seriously thinks a teenager BillG would have put IBM to their knees alone, without any help? That's like SteveJ Apple own stories, pure fantasy hollywood style brand propaganda.
Bill Gates wasn't alone when he went to negotiate with IBM. He had all the support of dad and mom, who where working with the US government at time, along with the best lawyers Gates family could pay. TL;dr, Gates family has been always part of the club, which would allow negotiations between "equals".
In contrast, Gary Kindall was what you could call a "new rich". The Steve Wozniak type. More of an engineer than a business man, and ofc he wasn't part of the club. He only wanted to sell his software, and never thought about "20 or 30 years plans" in the future. Since he wasn't part of the club, he couldn't negotiate "between equals" with IBM... And even he surely had at time the money to contract a better lawyer specifically to advice him in the IBM negotiations, as a "day living" man he didn't thought it was necessary. In the sense was a lot like HP in the day: "Who would take seriously Apple's and future IBMs? Why the average Joe would need a computer, at all?"
Really interesting video, i have used CP/M back in the early 90s but havent touched it since around 1992 so it was nice to see this video and see the history of the OS. I do like your historic OS videos like this. You explain them well and easy to understand.
CP/M actually lasted quite a long time. The Amstrad PCW, a Z80-based, low-cost business machine used CP/M, and was extremely successful in the UK and elsewhere in the 1980s and '90s, outselling PCs for a while and eventually became the best selling British computer ever. Amstrad's smaller home micros, the CPC range (which was also commonly used in small businesses) and the Sinclair Spectrum +3 also could run CP/M.
I used to be support for the CPC and the Amstrad 1512/1640 at DR in the UK... funny times.
I would love to see an alternate-history video where CP/M crushed Microsoft.
Enjoying the episodic nature. Appreciate your videos and many thanks for the time and effort
@@hmichaelkraut7968 thank you, and you‘re welcome 🙏
I grew up with APPLE ][e+/][e and IBM PCs. CP/M felt kind of like a less useful and clunky MS-DOS (Yes MS-DOS copied lots of CP/M features).
Apple DOS/Prodos were extremely functional for the apples and MS-DOS had a major advantage over CP/M which was the introduction of standard file formats and other handy things like a fully working piping system. Also directories were really handy.
I started on DOS v2.01a. DOS in terms of usability for the user was around the same until V3.3. Some extra commands were added. The first quality of like commands I really recall was Echo off and the @ symbol. The @ symbol meant you could finally 100% seamlessly make a batch file look like an executable.
There's a nice superuser post about DOS 2 being the first to include pipes,directories and Daemons/background execution. DOS 3 added in a bunch of network stuff, better disk stuff (bigger HDDs) and some more odds and ends. CP/M didn't feel like it was really growing but I never really used it as it was the "for business" OS.
I spent my high school years and college years writing papers on WordPerfect for Apple II, which wasn't as functional as it's DOS counterpart. Anyways, I think I missed out a little because AppleDOS and ProDOS just got out of the way. In a way, MS and PC-DOS forced users to learn a little because of all the settings, the rewriting of AUTOEXEC.BAT and conflicting TSRs, etc When you could do these things there came a certain pride and sense of accomplishment. By the time I got into Macs, I was "spoiled", and expected all my software to just work, which for the most part it did. When you think of it, users should have been blissfully unaware of the OS and only know the computer through the applications thye used. DOS, then Windows kinda changed that, and becoming the office computer expert meant something.
The problem with the lots of MFM disk formats forced Commodore to add the WD1770/71 floppy controller into the 1570 and 1571 drives to be useable with CP/M on the C128.
The C128DCR combined the C128 and the 1571 into one PCB, so it is one of the few boards with three different processors: 6502, 8502 and Z80. But Commodore also combined a 6526 and the WD1771 into the disk controller 5710. It required a new CBM DOS 3.1 which wasn't as compatible as DOS 3.0 of the original 1571. But I don't know how it affected CP/M compatibility.
Ran Windows for Workgroups over DR DOS 6. It was more stable than running Windows over any MS-DOS version.
I grew up with Amstrad CPC computers (CPC464 with DDI1 disk expansion) and then later with the CPC6128. Both had CP/M available for it. V2.2 on the 464 and V3 (aka CP/M+) on the 6128. Used it mainly for Infocom text adventure games, sometimes with terminal software to try out BBSs.
At least DR-DOS 5 or 6 (which came with an EMM386 like memory manager, transparent file compression and many more utilities) gave a good scare to microsoft as their MS DOS 4 and 5 were VERY lacking in features.
Then windows came and all this was lost.
Windows, the only non Unix (or Unix like) mainstream OS, hope Dave Cutler is happy.
Best wishes.
Well, to this day he tells people he hates Unix, so... There's a little more to the story, since Windows 3.1 ran on top of DR-DOS just fine, although the installer warned against it. There was a settlement for hundreds of millions for Microsoft's AARD bug which pretended DR_DOS (and other DOS clones) weren't compatible. By that time Caldera owned DR-DOS. Anywho, if installing Windows 3 on top of DR-DOS caught on, history would be different. And I'm kinda glad it didn't, because that brought about "Free and Open Source Software" and especially Linux. If it not for Windows, we wouldn't have needed an open source movement, or at least it would have been delayed many years.
@@squirlmy Part II will go a bit into the AARD story, not really for the technicalities but for the overall story on the Caldera vs. Microsoft lawsuit.
Impressive overview
As I remember it when I was at DR, Concurrent-DOS was multi user too. You could have up to 4 serial terminals attached to a pc.
The product names was not as complex as you make it as much was under license for individual companies to ship with their hardware. Each company was driven by their hardware and the intended usage model. That then naturally filtered them down to the products that would run and the level of support they wanted to offer their customer in the whole package.
It only seems confusing if one assumes DR sold its operating systems as Microsoft does today, namely direct to end-users.
Until the Amstrad 1512/1640 hit the UK market and the GEM environment, very little was sold to end users. At this point DR had to ramp up software duplication, distribution and sales direct to consumers. (Also for the other Amstrad machines.)
All my observations are related to my time at DR in the UK.
First-hands information is always the best.
As mentioned, I meant no offense here. It's just an outside point-of-view perception, and not all such details may be readily available in online sources.
Thanks for filling that in!
@AlexJacksonSmith While we're at it, let me know you guys actually called it: *DR DOS* or *Doctor DOS* ...?
@@THEPHINTAGECOLLECTOR D.R. the initials.
@@THEPHINTAGECOLLECTOR No problems, but If you jump from this directly to Novell (I had left by then); you miss out the massive transition GEM and Amstrad made for DR at the time.
At that time the company created, a you may not believe this but, a camera mount video box which allowed GEM Graph to be photographed for slides to use in executive presentations... ;)
Also GEM publisher was well ahead of its time. I would play with the advanced tools on the Amstrad 1640 that I had for support.
Was a great time.
@@THEPHINTAGECOLLECTOR Oh yes before my time as tech support I worked in accounts and also typed up the legal documents for the licensing contracts with the OEMs.
I trained as a secretary; but it made my programming later very fast. All 10 fingers :)
Very nice! However, where did you get the timeline of cp/m and dr dos (at about 11:04)? I cant find it when I google it.
@@erikp6614 it‘s linked via the wikipedia of CP/M. Need to check, I thought I had added it to the video description as well.
@@THEPHINTAGECOLLECTOR I can't find it (only a dead link on google). I never really grasped the actual history of dr dos and cp/m with all different versions and how they were related. The document you found was very nice. It is, of course, not a big deal if you can't find it again, but if you happens to find the link it would be nice.
Very nice video, as your other videos!!! Really good work!
@@erikp6614 I saw, it was missing from the video description.
en.wikipedia.org/wiki/DR-DOS#cite_note-Paul_2000_History-26
is the source.
@@THEPHINTAGECOLLECTOR Thank you very much!
@@erikp6614 you‘re welcome!
Great video, very informative. I have subscribed. One small point of info: The opposite of "lightweight" is "heavyweight", not "hardweight"
To get around different disk formats, XModem was used a lot between to machines connected with a null modem cable.
Funny, how much CP/M I have had in my hands over the years, but never utilized. I had a Z80 Softcard for my Apple IIe, didn't have the software to use it. Coleco Adam with CP/M datapack. Yes, that would be CP/M on cassette tape, I really had no use for this. Commodore 128 with the CP/M disk and utility disks. I booted it, and that was it. I saw no reason to fetch CP/M applications, when everything would have a C128 mode version. I would someday love to have an Altair 8800 to play with, then I would be forced to uses CP/M and you know it would feel natural. As for the other systems, it felt it was tacked on and unnecessary.
I can definitely recommend to go into an Altair.
I don't have the real thing but an Arduino-based replica, the AltairDuino. It's really close, just that it's missing the S100 bus.
But good enough to experience it (almost) the way it was 50 years ago.
D'er Doss? I think you'll find it's pronounced "Doctor Doss".
Anyway, while I don't hear this said today, back in the 70's and 80's it was widely known that CP/M was originally just written so a disk drive manufacturer could demo their product. After popular demand, it got expanded. I had CP/M on my TRS-80 COCO 2 and unmodified TRS-80 Model I (with expander). LDOS was better.
Yeah, I whitnessed it actually being called "Doctor DOS" by some people here in Switzerland back in day, mistaking the "DR" as being "Dr" (or "Doctor" in short), instead of being the initials of "DR" = "Digital Research".
It may well be that in other languages, even english, it might have been sometimes mistakenly called "Doctor DOS". I wasn't there to whitness.
But if anything, that convention was IMO more by mistake and not by intention of the vendor.
All video formats I ever saw where DR DOS was mentioned, only ever refer to it as "DR DOS" and not "Doctor DOS".
I stick myself as well to mentions in the Computer Chronicles, like here: th-cam.com/video/CKf7Oc60fVM/w-d-xo.html
Stewart Cheifet calls it "DR DOS" as well, and I believe, as someone who worked closely with Kildall on a regular basis, they would certainly have talked about DR DOS on some occasions.
Had Gary Kildall called it "Doctor DOS", then I'm certain Cheifet would also have called it "Doctor DOS".
Btw, good point about the disk drive.
Kildall mentions this in his memoirs, how he wrote CP/M on an emulator first, and struggled how to get it running for a Memorex floppy drive, as he wanted to add mass storage support for PL/M.
There's ever so much room for facts in those videos, and it's always a challenge, what to include, and what not.
@@THEPHINTAGECOLLECTOR You certainly seem to have done your research! I am persuaded about the pronounciation. 😎
In my opinion, the reason why MS-DOS forced IBM compatibility on everyone is due to Tim Patterson's faulty function 9 in the DOS API. Function 9 simply prints a string terminated by a dollar sign in MS-DOS. In CP/M, it has an escape code language to place the cursor in different positions on the screen, to clear the screen, set the colors, etc. That why you can have full-screen word processors like Wordstar that can run on any CP/M computer. In MS-DOS, you had to use a BIOS call or access video ram directly to do any of that. That's why computers like the Tandy 2000, which was MS-DOS compatible but not IBM PC compatible, failed miserably. The Tandy 2000 couldn't run anything complex that had to be designed around the IBM PC memory map. It almost seems like this oversight was intentional.
CPM had the same issue with software requiring to be patched for each machine. There were MSDOS compatible machines that had WordStar, dBASE etc. So I'm not sure it was on purpose. I don't think anyone thought the PC would even be a big deal. Everyone would just keep their PC for a while and then throw it out, along with their software, like they had done already.
@@phill6859 The software only needs to know the underlying terminal or terminal emulation codes, the rest is hardware independent from the perspective of the user program. And if you don't need to move the cursor around you can even ignore that.
I am working on a little emu board that emulates computers from the 70s on a, currently, RP2040 and can say that most CP/M programs have several terminal types built-in, the most used where ADM-3A/31, the VT52 and the VT100 and compatible ones. The C128 emulates a ADM-31 with a color expansion for example. Most modern implementations will be attached to a PC that uses VT100/ANSI, my little boards software can translate between them by the way.
WHAT? Forced compatibility? Compaq cleanroom reverse engineered the BIOS, (no the first) to get to as close to 100% compatibility as possible. They had sales of $111 million the first year, their first year of existence, with the clunky "Compaq Portable", breaking all records for a first company. Compaq got together with 8 other companies (sometimes facetiously referred to as the Gang of Nine), name the PC bus "ISA Bus" wrote a plan for it to be succeeded by EISA, and I believe Phoenix was there, offering their IBM-compatible BIOS. Before Compaq, Microsoft very happily made many MS-DOSes incompatible for "IBM Workalikes", including offerings by Tandy/RadioShack, Xerox, Hewlett-Packard, Digital Equipment Corporation, Sanyo, Texas Instruments, Tulip, Wang, Olivetti... The Tandy 2000 didn't die of being IBM compatible (note the compatibility or lack of, was with IBM "PC-DOS") Very quickly Tandy made the "1000", compatible with PCjr, not the XT or AT, and by all rights they could've set the standard with Microsoft's blessings. They made sales blunders, not design mistakes. IBM and only IBM was looking to set the standards, not Microsoft. AFTER Compaq and Phoenix, clones became leaders and wiped the floor with IBM. Before that it was a gamble which 8088/86 machine would set the standard, and if it weren't Lotus 1-2-3 and WordPerfect courting IBM, the biggest and most monied, things could have gone differently. The thing about the IBM PC is that it was made with industry-standard components, and IBM let Microsoft license DOS to anyone else. There was no "forcing". If Apple didn't sue and win against Franklin in 1984 for copying their boot firmware, and custom parts in the Apple lIe, ironically that might have become the PC standard, although Apple might have lost like IBM. The world's industries WANTED compatibility in software, whether tech companies wanted it or not.
@@phill6859 No "forcing" at all. Really smart people might have figured it out, and I suppose they did. But also remember IBM was firstly and importantly a minicomputer and mainframe company; they didn't want the PC cannabilizing that market (it did anyways). So they made decisions like letting Microsoft give them a crappy CP/M imitation, giving it an 8-bit bus and a crippled 16-bit CPU, and most importantly, letting Microsoft give licenses of MS-DOS to other companies. Also using off-the-shelf parts (except for the BIOS of course) None of those decision make sense, until you realize most of the executives and sales force would rather have another terminal and not anything to compete. I just heard an interview with the Tandy person in charge of their PCs, and he said "business machines made something like 20% commission while home computer sales could be 80%" the main reason to have a dominant office PC, was to leap into the home market from there. IBM tried with PCjr. The IBM "clones" won the latter.
Thanks!
@@hmichaelkraut7968 likewise! Thank you for your appreciation!
I used to call it Doctor DOS lol.
I certainly never heard it called "D'er Doss", lol!
Don't forget KP/M, also known as TIKO
I hope it runs on windows ME
@@judewestburner it does ;)
One meeting, and the future would have belonged to DR-DOS and GEM instead of MS-DOS and Windows
Novell tea back then
MS-DOS and Microsoft happened :/
It's more complicated. People had been begging for cpm86 since 1978 when the 8086 came out. Gary could have ported it in a month or two, he did not. Bill even told IBM to talk to Gary, that still didn't work. Eventually they had to go back to Bill and he told IBM they could license 86dos. IBM told Bill they would prefer if he was responsible for licensing it and doing the modifications
Nobody (probably even Bill) thought IBM would be priced out of the market so quickly or that the PC would endure so long.
@@phill6859 and IBM really played dirty offering CP/M for 199 and DOS for 49 bucks (I believe the prices were). Anyways Microsoft and IBM are horrific untrustworthy companies. And I worked for big Blue 🤮
@@phill6859 yeah, the success of the IBM PC was surely a surprise.
But from what i can gather from history it was all thanks to the work of Compaq and their balls against IBM creating a compatible BIOS, the only thing not off-shelf in a IBM PC. After that it was off to the races as everyone and their dog could build a PC and sell it. And while in the 80s Commodore for example had a huge success with the C64 as the computer for the masses and a good success with the Amiga as a high-end machine they rested on their laurels and innovated too late and too little.
Apple somehow succeeded with the something-with-media crowd, Sinclair couldn't replicate their success with the Spectrum as a low-end machine and Atari, while it had some fans with the ST, fell to the wayside too. Not mentioning all the other small computer makers here.
And the nail in the coffin for everyone but the PC was Doom. Quake cemented that.
@@phill6859 I think IBM didn't want to negotiate with Gary, who would have had the upper hand in that exchange. They were already on good terms with Bill and his BASIC software business, and I'm sure they figured they'd make a technically superior OS if the PC caught on. And they did eventually create OS/2, and they were caught flat footed when MS Windows destroyed the OS, that people preferred the lower cost ISA bus computers from companies like Compaq, (and companies with Phoenix's BIOS), over their technically superior MCA bus PS/2s. I think people overstate Kindall's arrogance, he just wasn't a good businessman. The collective borg of IBM were the arrogant ones. After all, Kindall had CP/M-86 ready for the PC rollout, IBM just decided to screw him with a $240 price tag.
@@squirlmy When reading through all possible sources, I came to the same conclusion. Kildall was more a techie and engineer seeking for the next technological step, and not so much the business man.
As opposed to Gates, who was more the business oportunist.
Anyway, the IBM price tag was really cheeky, especially when considering that at least all sources I found mention CP/M selling for 70 USD (according to Kildall's memoirs), 80 USD (Wikipedia), and somewhere around 80-100 $ when looking at ads, or when putting known OEM details into perspective.
The most expensive version I saw was the adapted CP/M for the TRS-80 Model 4, which sold for 169$.
Four hundred million licences for Windows 11, Microsoft is now at it's peak, Windows 12 will be less popular. I won't be using it.