DR-DOS rises again – rebuilt from scratch, not open source

DR-DOS is back, and there is already a test version you can download. But as of yet, it's not finished, not FOSS – and not based on the original code. The long-dormant DR-DOS.com website is alive again, and DR-DOS 9.0 is in development. There have been six preliminary releases so far this year. The current work-in-progress …

[Anonymous Coward]

I will never understand people who publish a hobby OS as closed source. I don't mean that outrageously, I don't feel bitter about being deprived of the code. But a hobby OS is by far one of the single most useless, redundant things you could possibly make. We already have FreeDOS, we already have Redox and ReactOS and Linux itself, and they all are far more supported, have more functionality and drivers than almost any hobby OS you could ever hope to cobble together on your own. I felt this exact same thing seeing how MinuetOS's x64 version was proprietary, which, ironically, is also coded in assembly (is this a running theme? do you really have to have a mental disorder to use assembly?) Nobody wants your OS, nobody cares, not a single soul on the planet except for your mother. Why not just share the code, show off what you did? Why are you coming up to your schoolyard friends hiding your new toy behind your back and taunting them that they can't see it? They don't actually care, they're having fun on their own. I don't even care if the license is something as useless as Business Source, why are you so ashamed of your own code that you can't even bear to show it? There's genuinely no good reason. Again, I don't care, I don't feel deprived of it, but it's just such a nonsensically antisocial thing to do.

[TonyHoyle]

And inevitably when the author decides it isn't worth it because nobody is buying it, all the work is lost forever..

[2Blockchainz]

There's a company called Diagnostica Stago who are currently manufacturing clinical testing devices running DR-DOS. Lab techs who use the device are now learning command line from several OS generations past. If it ain't broke, don't fix it!

[Cliffwilliams44]

Why? Why do this?

Linux can fit in a very small footprint, I see no reason in hell to do something like this! Other than to make it closed!!

[PM.]

Because it was certified to run with DR-DOS ages ago ? To make it run with Linux would require recertification ? Not free and not cheap ?

[graemep]

Would any such certification apply to a reimplementation?

[spuck]

They didn't say the company is starting to manufacture devices that run DR-DOS; they might have been building them on the same base since 1992, for all we know...

[Jaybus]

For the same reason that specialty companies still manufacture carburetors. Just as there are hobbyists who still tinker with cars that were made decades before electronic fuel injection, there are hobbyists who still tinker with PCs that were made before microprocessors had MMUs..

[pirxhh]

Reliability.

DOS has no multitasking, so any task switching is done directly under the developer's control. Same for memory management.

When your hard- and software has life or death impiications, you might want exactly that level of control.

[IGotOut]

The flip side is it's a clean room creation, so free from legal challenges.

If it goes open source and someone throws in say a line of code from Caldera's version, and it ends back in the main tree, then all sorts of shit could happen.

[tracker1]

I'm still not sure why not FreeDOS? What does this really offer over the top?

That said, my concern would be the trademark infringement

[that one in the corner]

> but it's just such a nonsensically antisocial thing to do.

Nonsensical - probably.

Antisocial - nope.

Releasing this as binaries only isn't causing you, or anybody else, any harm: nobody is using it, as you point out there is no reason for anybody to use, other than mild curiousity, so the whole thing can be ignored without any loss. The only possible damage, the only antisocial bit, is the waste of time reading the article. And there are far more actually antisocial publications than TFA[1]!

Leaving aside active harms, there is the description of "antisocial" as a pattern of simply not caring about the effects of one's actions on people, but if the effect is barely even a "meh" then it is hard to say that description applies. Otherwise we'd be railing against all the projects on SourceForge/Gitlab/other places, whose existence doesn't make one jot of difference to us.

[1] like the dross that came up this afternoon claiming to be a helpful article about Python but got literally everything wrong; luckily the error messages from CPython had some actual content...

PS

watching the oven timer tick down is so boring, have resorted to typing out irritations about word usage, just to occupy the time!

[Jimjam3]

Agreed, to call it antisocial was an exaggeration.

It’s still an odd project with minimal utility.

[Eric 9001]

Making software proprietary is a very antisocial act.

People always deserve the 4 freedoms and it's clearly antisocial to refuse to grant them, by using a proprietary license and withholding the source code (in this case, the assembly with comments), as people then cannot socially exercise the 4 freedoms.

[PM.]

Well, they possibly try to earn money on this. That's not antisocial

[Eric 9001]

It isn't antisocial to try to earn money in a legitimate fashion, by respecting the user and not trampling over the 4 freedoms.

One legitimate business method would be for the business to state that for a copy to be provided it costs $5 and when you pay $5, you get a copy, including the source code under a free license.

It is very antisocial to restrict the software by making it proprietary, just because you hold a belief that the software being proprietary guarantees huge profits (in this case, the software isn't even as functionally good as "Free"DOS, which is available gratis, and therefore only complete and utter fools would hand over money for it - and there's only so many of those - therefore there are little potential profits available regardless).

[Jimjam3]

If it’s VERY antisocial to provide closed source software. How antisocial do you regard the current activities happening in the world?

In my view at the most it’s MILDLY antisocial to not offer source code.

[Anonymous Coward]

"Antisocial" as in "I don't know how to socialize"

But I get it, everything has to be political, it's new and alien when someone uses a word outside of political context.

[that one in the corner]

That isn't any definition of "antisocial"!

You mean, at best, "asocial".

And what on Earth is that prattling about "political"? Is every word you don't know the meaning of "political" now?

Good grief.

[Anonymous Coward]

I think a lot of "closed source but free" comes down to developers who don't want the Internet to pick holes in their code. I get it. None of us are perfect, and it's easier to criticize someone else's code than build something cool from scratch.

The Internet can be really judgmental, and there's definitely an emotional appeal in insulating one's self from the criticism, even if the rational brain knows that open source makes a better product.

[Rob Daglish]

I dont really understand why you're letting yourself get so worked up over this.

If it's a hobby, great - let them get on with it and do whatever they want. If you can verify the claim it runs as DRDOS did, does it matter what the source code looks like?

It's much less antisocial than the wallies who go out hiking in the middle of winter armed with a pacamac, flip-flops and a bag of pork scratchings then expect the mountain rescue to get them down from Skiddaw/Ben Nevis/Snowdon when they run out of bravado...

[VeryRealHuman]

Sometimes people make things because they feel like it

[Irongut]

This can't be "actually DR DOS" and a new clean room implementation, those are mutually exclusive things.

This is in fact a spiritual successor or homage.

[Liam Proven]

> those are mutually exclusive things.

Agreed.

He talks a lot on Reddit about re-implementing the distinctive stuff that made DR-DOS _DR_ not MS-DOS, but honestly, DR-DOS was a relatively quick hack to have an entry in the DOS market. It was a low-end product, simple and single-tasking, while DR worked on the real stuff, the multitasking Concurrent DOS and real-time FlexOS.

The first version incorporated the big FAT16 volumes from Compaq DOS 3.31 which MS-DOS/PC DOS 3.3 couldn't do.

It sold. It made money. So they leapfrogged the lardy MS-DOS 4 and did DR-DOS 5, which integrated Quarterdeck QEMM386-style upper memory block handling. This was a DOS that could live in the High Memory Area (from 1024 kB to 1088 kB in RAM) and load its own drivers into UMBs. Result, a DOS that _easily_ gave you 635+ kB of base memory, just as DOS apps were getting huge and really struggling to fit into 640 kB along with a network stack, multimedia and mouse support and stuff.

And in a stroke of genius they sold it at retail. You couldn't buy DOS before that. It only sold to OEMs for inclusion with new PCs.

And it had the ViewMax GUI, a cut-down GEM.

DR-DOS 5 sold great. MS noticed and did MS-DOS 5, a copy-cat release.

So DR did DR-DOS 6, with bundled disk compression.

MS copied with MS-DOS 6. We know how that went... 6.2, 6.21, 6.22...

Novell bought DR and one-upped MS again: Novell DOS 7 had built-in peer-to-peer networking.

But now, the main thing that was distinctive about DR-DOS was colon column separators in `dir /w`.

There's not much for him to include.

[Albert Coates]

Quarterdeck QEMM386 - now that's when things started happening - multitasking DOS sessions with DESQview..

And let's not forget 4DOS, almost bash. Beers anyway ------>

[mdubash]

Please, don't mention QEMM386 again: it brings tears of nostalgia to my eyes...

[APro]

Same here.

I replaced MS-DOS with DR-DOS within a day of having my first "PC", and later put 4DOS on top of that. Then followed with QEMM386 and further DESQVIEW to top it off and friends were awestruck at the functionality. DESQVIEW/X was what Windows should have been and, being a child of SunOS and HP/UX, running an X window on your PC, or DOS window on SunView or OpenWindows was cool.

[billdehaan]

4DOS 4ever

And let's not forget 4DOS, almost bash

Forget it? I still use it daily.

4DOS begat 4OS2, which begat 4NT, which begat Take Command, which is still available from JP Software. Of course, Take Command includes both a gui terminal with numerous advanced terminal features (file tree, tabbed interface, drag and drop support, etc.) but the shell (named TCC for Take Command Console) is still there, significantly improved and advanced from 30 years ago.

And yes, both the terminal and shell run under Wine. Although I've migrated most of my legacy BTM scripts to bash, there are some that use Windows command line tools that simply aren't worth the effort, so I still run them in Wine. I've even written a few new BTM scripts in cases where it was easier and faster than writing the same functionality in bash.

While I wouldn't recommend Linux users go out to get it, I do recommend it for Windows users. It predates Powershell by decades, and although there is some overlap with Powershell, it's a lot easier to learn, and quick to bang out scripts than Powershell. I

[frankvw]

Re: 4DOS 4ever

I no longer use 4DOS but I remember it fondly. I used to run a BBS until the mid-1990s (when Windows caused an explosion in file size that made running it too expensive for me, both in storage requirements and dial-up line bills) and the whole darn system, including FIDOnet mail, point systems, nightly maintenance batch jobs and what not ran off one single 4DOS batch file that was only different from a full-fledged application in that it was readable script rather than an executable binary.

Over the years 4DOS has done more heavy lifting for me than anything else. Even in bash I have never built anything the size and scope of those 4DOS batch files.

[BartyFartsLast]

Manuals, actual, real, touchable, lay flat ring bound manuals.

Oh, and source code.

[Dan 55]

But the OSI won't call it open source, they don't like the licence.

[Eric 9001]

The statement is not a license, therefore it is not "open source", just like how software validly released into the public domain with source code is not "open source"; https://opensource.org/blog/public-domain-is-not-open-source

The statement also appears to restrict commercial distribution, which doesn't meet requirement 6; https://opensource.org/osd

If the statement was modified into a free one, like; "You are permitted to use, distribute, modify and/or otherwise make available CP/M and its derivatives." and CP/M was fully available in source form, it would be free software; (https://www.gnu.org/philosophy/free-sw.html#four-freedoms), but it still wouldn't be "open source".

Such statement is still not a good idea, as if you aren't disclaiming warranty, it could be argued that you have granted a warranty - licensing under a competently written weak license like the Apache 2.0 would be a much better idea.

[that one in the corner]

One can safely ignore the OSI.

They don't actually *do* anything, beyond lying about inventing the phrase "open source" and advertising themselves - mainly by convincing people to use a URL into their website as though it is in any way "authoritative". They don't host any useful examination of what the different licences provide for and/or where they are applicable.

[Eric 9001]

The OSI hasn't claimed to have invented the phrase "open source" and did develop the 10 requirements that make up the "Open Source Definition", which gives a definition for "open source software".

The OSI does publish a blog post that is not true, in which someone claims credit for having come up with "open source software" first, who didn't.

The 10 requirements of the "OSD" are mostly agreeable - but the requirements list is way too long winded and is in fact looser than the much shorter 4 freedoms, which allow the OSI to approve several proprietary licenses.

It seems recently the OSI has pivoted to pushing proprietary software and SaaSS with their "open source AI" definition.

[RobThBay]

DR DOS 3.41 was from 1989, not 1981

.....we recommend DR DOS 3.41 from 1981....

The link takes you to a photo of 3.41's boot disk. It clearly shows August 1989.

[IvyKing]

Re: DR DOS 3.41 was from 1989, not 1981

I was about to make a similar correction. In 1981, 86-DOS was in the process of being converted to PC/MS-DOS and DRI was still busy trying to get CP/M-86 out the door.

[Liam Proven]

Re: DR DOS 3.41 was from 1989, not 1981

> DR DOS 3.41 was from 1989, not 1981

Whoops -- I knew that! A plain old typo. I will notify the editors.

[PhilipN]

"without fancy memory management"

Gee, thanks, Liam for sending me into a fainting fit at this time of the morning.

P.S. Good article.

[nintendoeats]

If it's released as binary, and written in ASM, then...in a sense isn't it already source available?

[david 12]

Maybe not. Re-locatable ASM is significantly more difficult to decompile than static-location ASM. And although a real-mode boot sequence will start in a fixed location, the way it's built will have the blocks of code kind of randomly mixed. 8086 machine language instructions are mixed length,, you can't always see the start of a block and the start of an instruction

It will be a lot faster now than when I did it, but still a very manual process of locating and labeling jump points and block starts, then repeating.

A good AI should be able to do it, I don't have any idea how to set that up.

[Eric 9001]

No - even hand written assembly is incomprehensible when de-assembled - as comments, formatting and the original macros are a hard requirement to understand even assembly you yourself wrote a month ago.

[joeldillon]

I don't know about 'hard requirement' given that people (virus writers, and people analysing those viruses) disassemble stuff all the time. But it's not like getting actual source code, agreed.

[Eric 9001]

Yes, if you waste many hours reverse engineering, you can work out some of the functionality and then maybe make some trivial modifications.

With the actual source code, you can just read the sources to understand the functionality and implement whatever wanted modifications.

[nintendoeats]

Guys, this was a joke.

[LBJsPNS]

FreeDOS handles my DOS needs

And it has 4DOS available. And a lot of other DOS utilities. Even GEM, if you're into it.

[david 12]

more than two partitions

MS-DOS 3.3, which was the first release that supported more than two partitions on each hard disk

DOS 3.3 included support for booting from high partitions. Using the block driver interface, you had to boot from floppy or partition 1/2, then load the block driver.

DOS 2 had a documented block driver interface. It was one of the IBM requirements. When I say 'documented', it was in my user manuals, not like the DOS API or the information for building TSRs.

The documented block driver interface was how you ran the common 40MB disks on DOS 2. And since the standard partition size was 10MB, the common 40MB disks were commonly represented as 4 10MB disks.

IBM sold 10MB drives, but the IBM PC/XT were never cutting-edge hardware (apart from the keyboard).

[Esso]

SpinRite

Maybe Mr. Gibson will decide on a project to get his famous disc software EUFI-compatible as soon as 2039.

[Dante Alighieri]

RISC OS PCem

Ran/provided DR DOS - my first version of a DOS. still matched/outperformed ?286s in emulation.

[Bebu sa Ware]

Not vibe coding in asm86

The contrary doesn't bear thinking about. ;)

This reminded me of a product (from back when) which was rather clever for its time: Wendin-DOS: the forgotten-multi-tasking, multi-user MSDOS clone from the 1980.

While the toolkit used for for its construction was sold as separate products ie source available, the actual OS products (DOS, VMS, Unix) were only offered as binaries. You can still find copies on the net in the usual murky places.

Apart from glue code (mostly to get round the PC's non-reentrant BIOS) it was all written in C.

The only thing Wendin-DOS didn't support was TSRs — although I am sure DR-DOS 9.0 will, surely. ;)

_{Young chaps banging away on keyboards in their dorm rooms never produced a significant OS - did they ? ;)}

[IvyKing]

Re: Not vibe coding in asm86

1987 not 1980 per the linked article.

The Wendin logo looks familiar, probably running across it in computer mags from the 1986-87. For even more trivia on multi-user/multi-tasking DOS, the literature for 86-DOS specifically mentioned a multi-user version and when DOS 2.0 came around, comments about the new device driver interface suggested some features were put in to support multi-tasking.

As for ASM coding, one of the big helps in understanding what's going was Stephen Morse's book on the 8086, specifically how some instructions were intended to be used.

[Eric 9001]

Unfortunately 386MAX is proprietary software, as it is not licensed.

Just dumping a copy of the GPLv3 into the root doesn't legally mean anything.

[Liam Proven]

> Just dumping a copy of the GPLv3 into the root doesn't legally mean anything.

Then go file an issue.

I dimly remember seeing the mailing list post, where the author found that the company no longer existed and rights had reverted to him and he responded by putting it on Github. I think _he_ thinks uploading it and putting the GPL in there counts. If it does not, go tell him, not us. Explain what he needs to do.

[Eric 9001]

I am not running proprietary software from microsoft.

On further inspection, it seems in 2022 he has edited the .ASM & .C & .H files to say; "(C) Copyright 1984, 1988, 1992-7 Qualitas, Inc. GNU General Public License version 3." - which is ambiguous whether the license is GPLv3-only or GPLv3-or-later and that only applies to the marked files - all the other files, like the .BAT, .DEF, MAKEFILE's, .SRC etc are not licensed.

There are several binaries that contain the string "All Rights Reserved" and it's hard to tell if the source code is included.

I would go and email him about it, but in the git log, his "email" is; [e-mail], so I can't.

[Blackjack]

Freedos is enough for me thank you. Maybe I would have got interested a long time ago but I lack any working machine old enough to run this anyway.

But I will probably see whatever video LGR eventuality makes from this.

[Vader]

Oh well to have a with the DR I guess.

[Liam Proven]

> Oh well to have a with the DR I guess.

... Pardon?

[Dr Kerfuffle]

Amstrad?

Didn't some of the early Amstrad home computers come with a version of DR-DOS? together with a windowing environment called GEM ? or is my memory of this wrong?

Paul

[Liam Proven]

Re: Amstrad?

> Didn't some of the early Amstrad home computers come with a version of DR-DOS? together with a windowing environment called GEM ? or is my memory of this wrong?

Your memory is wrong, I'm afraid, but it is close.

They came with DR's DOS Plus, a different spinoff from CP/M-86 and much less compatible with MS's DOS.

https://en.wikipedia.org/wiki/DOS_Plus

[Soruk]

Re: Amstrad?

The BBC Master 512 also used DOS Plus, and someone wrote an add-on dubbed the PC Compatibility Enhancer, which allowed many additional items of software to run on it.

[Mythical Ham-Lunch]

So it has perfect legal clarity to... disappear whenever the current owner gets tired of it, and be impossible to upgrade or maintain after that since there isn't any public source code?

[Rich 11]

Fond memories...

...that make me feel old. Again.

[Acrimonius]

Why?

What is the point in expnding this huge effort on a DOS OS? In assembler as well!

[J.G.Harston]

Re: Why?

What's the point of expending effort on an old car/bike/boat/radio/sideboard when you could just buy a new one?

[doublelayer]

Re: Why?

It's a question worth asking. If the answer is "no reason, it makes me happy", it provides information. If the answer is not that, then there might be things the unacquainted user would like to know which convinces them to use it or, if source was available, help out.

But also, the analogy fails. Some people work on old vehicles because they have historical value, so making them operate can let people see how technology existed back then. That quite often is for the general public when they're put on display or maintained by museums or it can be for the benefit of a smaller group. The analogy in this case would be building a new car, though one with few features, and designing it to look like an old car. You can do anything you want, but if you have another reason, it might be compelling to someone else, so they ask to see if there is one.

[spuck]

Seems a long way to go...

Does it work? I've tested DOOM, Warcraft, SimCity, Stronghold, Commander Keen, Oregon Trail, and plenty of other period-accurate titles. Lots works. There are still gaps.

I guess it depends on what "works" means to CheezWeasel, but there are plenty of corner cases ahead beyond "loads executable from disk into RAM and turns over execution to it" to be 100% (or even 80%) compatible to an existing DOS means.

Everyone needs a hobby so hack away and enjoy, I say.

[EvaQ]

"it seems that it isn't vibe-coded." ...

As vibe-coding is so 2025, why not let Claude create the complete DOS? Just like the anthropic person who let Claude automagically create a C compiler able to compile Linux source code:

"Anthropic's Claude Opus 4.6 AI models recently developed a C compiler, named CCC (Claude's C Compiler), written entirely in the Rust programming language. This project was an experiment in autonomous software engineering, where 16 AI agents worked in parallel over two weeks to produce the 100,000-line codebase"

eva@zwarte:~/git/claudes-c-compiler/target/release$ ./ccc hello_world.c -o hello_world

eva@zwarte:~/git/claudes-c-compiler/target/release$ ./hello_world

Hello World

eva@zwarte:~/git/claudes-c-compiler/target/release$

[GrahamCracker]

I loved DOS

I hope that someone creates a DOS that allows for easy access to video mapping like the original DOS did. You could write really decent games in QuickBASIC, QBASIC, BASIC PDS, and Turbo C. Except that the screen modes like SCREEN 12 (640x480 16 colors) and SCREEN 23 (800x600 256 colors) be expanded to 1920 x 1080 @65000 colors. I'd give up DirectX and OpenGL (And C++) in a flash to be able to work quickly again in modern resolutions.

[that one in the corner]

Re: I loved DOS

DirectX allows you to create a direct memory mapping to a portion of your video card's memory, which can then be displayed anywhere on the screen (e.g. overlapping a window or, if you so desire, overlapping the entire display). If your video card allows it, you can even choose the colour depth/ format and have it mapped on the fly to match the rest of the screen (e.g. you deal in YUV and it displays in 32-bit RGB). So you can have a VGA-sized area and run your old code (recompiled) on that.

Just be warned that cutting out the GPU, even if it is just an oldie stylee 2D accelerator card, and relying on the CPU for all your image generation may end up feeling rather slow...