VMS/XDE is an OpenVMS x86 development environment for Linux and WIndows (via WSL). It provides a familiar user experience for OpenVMS developers working in Linux and Windows yet offers 100% binary and file system compatilibilty with OpenVMS.
VMS/XDE includes OpenVMS V9.2-3 user, supervisor and executive mode operating system environments and a set of x86 native compilers and layered products geared towards OpenVMS software development and testing.
↫ VMS/XDE website
VMS/XDE is a beta version, and comes with the usual annoying OpenVMS x86 time bombs, this time exploding on 3 January 2026. If you intend to use the finalised commercial version after the beta period ends, you’ll have to employ the same licenses as regular OpenVMS. It’s a bit of a mess, but that’s the OpenVMS way, sadly – and I don’t blame them, either, as I’m sure they’re hamstrung by a ton of agreements and restrictions imposed upon them by HP.
Regardless, VMS/XDE brings a zero setup OpenVMS environment to the operating system you’re already using, making it easier to develop and cross-compile for the platform. I still have absolutely no clue just how many people OpenVMS is still relevant for, but I absolutely adore the fact VMS Software Inc. is working on this. In a world where so many of its former competitors are being held hostage by corporate indifference, it’s refreshing to see VMS still moving forward.
I think this is the dilemma of more or less all alternate commercial operating systems.
OS/2 (eComstation), Amiga (what was the current name), QNX, and many others have the same unsolvable problem:
A. Do you offer a free “good enough” version, where you know people will be content with and use forever as “shareware”?
B. Or do you add artificial annoying restrictions so that people are not comfortable enough to use it?
It is a chicken and egg problem, where hardware manufacturers don’t have this issue
(For example Amazon can do FireOS, and people hack it for other systems. Amazon don’t care as it is a vehicle to sell products and services).
sukru,
Yeah, there’s always been a balance. IMHO shareware worked well with games; you don’t get the full content until you buy it, but you can enjoy the shareware content freely. Sadly game publishers don’t seem to favor the shareware model much any more. With utilities you often need the utility to work just to try it and if it works then many won’t actually buy it. So the shareware utilities end up with arbitrary limitations added to make sure it doesn’t fully do the job.
Regardless, I would not call anything “open” if it contains timebombs and other restrictions. Such anti-features are inherently more closed. I’d completely understand why a company wouldn’t want to release open software, but then they shouldn’t label it open.
Alfman,
Yes, utilities like WinZip and WinRAR had that issue, too. I eventually bought both of them (after no longer needing either). But I could very well never have paid at all and still got a lot of use.
Back then OS would be bundled with the hardware. Or rather all hardware manufacturers had their own OS (IBM/AIX, Sun/Solaris, SGI/IRIX, Amiga/AmigaOS, Amstrad CP/M), and some were even completely locked down (Atari, Famicom/NES).
Ironically, Microsoft opened it up with MS-DOS being available for many different vendors. Software become a norm.
Which… was easy back then. Hardware could be bought for “cheap”, and you’d pay Microsoft (or Novell, or Digital Research) for your operating system.
Today… I’m not sure how they would find a sustainable model where hardware is completely generic, and other full featured operating systems like Windows is already “free”.
(Free as in anyone can now download the official ISO and only has to endure “shareware” nagging)
===
This is for general purpose OS. You could still sell specialized ones like digital signage.
Amstrad provided the regular CP/M for Z80. And I’m not sure what you’re referring to regarding Atari… TOS was GEM running on top of GEMDOS and unlike the Amiga, the Atari ST was designed with off the shelf parts. It wasn’t especially locked down.
larkin,
Okay I might be misremembering some parts.
Atari and NES were gaming computers (were the called consoled back then), that were much more closed than regular ones. Which continue the tradition to this day.
Atari VCS/2600/5200/7800 (consoles like Nintendo NES) is *NOT* Atari 600/800 XL/130 XE (8 bits computers like Amstrad CPC) is *NOT* Atari ST/TT/Falcon030 (16/32 bits computers like Amiga 500/600/1200/2000/3000/…)
TOS (The Operating System) is composed of BIOS (not like PC BIOS), XBIOS and GEMDOS (from CP/M) that deal with I/O, file system and VT52 terminal emulation, while VDI (graphic routines) and AES (desktop and windowing) deal with graphic interface.
Nothing is “closed” as everything was clearly documented and available from DRI as well. It was just “niche” compared to more successful alternatives (Windows, Mac, etc).
@Alfman
> I would not call anything “open” if it contains timebombs and other restrictions
I can agree with that. However, this is us projecting our understanding of that term late in the era of Open Source. OpenVMS comes from an earlier era when “open” simply meant that something would support “open standards” which mostly meant that it would be available on multiple platforms. It is the same definition of “open” found in “The Open Group”, the keepers of proprietary UNIX that inspired the founding of the GNU Project.
The name OpenVMS was meant to signify a version of VMS that would run on more than just VAX. For it’s first 25 years, VMS was available for VAX hardware only.
OpenVMS pre-dates all GPL Linux. OpenVMS was released for DEC Alpha in 1992. That was before Red Hat. It was before Debian. I think it was even before Soft Landing Systems. Linus had announced Linux kernel 0.01 the year before but Linux did not go GPL until after OpenVMS had been released.
When OpenVMS was released, “Open Source” did not exist although “Free Software” was already a thing of course. If OpenVMS was intended to be “open” in the way we use that term now, they probably would have called it FreeVMS for the same reasons we got FreeBSD (released in 1993). There was an Open Source VMS clone attempt called FreeVMS but it came way after OpenVMS and is already long gone.
The Open Source Initiative claims to have come up with the term “Open Source” in 1998, years after the first release of OpenVMS. That was the year they released the “Open Source Definition”. Even that was 30 years ago already so it is hard to remember what the world looked like before.
While I agree with your definition of “open”, the term has evolved over the years and OpenVMS comes by its name honestly.
> The name OpenVMS was meant to signify a version of VMS that would run on more than just VAX. For it’s first 25 years,
nope. vms became openvms when it became posix-certified.
@Anachronda
It is pretty easy to verify that the first use of the OpenVMS name was when it was released for Alpha in 1992. In fact, the VAX version was not called OpenVMS until 1993.
Also, there was a product called POSIX for OpenVMS that was sold as an add-on (also in 1992).
If you disagree, you can head over to Wikipedia to make some edits to their “POSIX compatibility” section:
https://en.wikipedia.org/wiki/OpenVMS
I am not sure that OpenVMS was ever POSIX “certified”. But we can agree that “open” is for open standards and better POSIX support would have been part of that same overlall push.
@sukru
You are of course correct but there is another level of nuance possible.
It is possible to have a non-commerical edition for enthusiasts, Open Source devs, and educational purposes in addition to the fully licensed version for commercial use. In addition to the “entitlement” the commercial version can come with support which means that the non-commercail version does not have to cost anything other than hosting.
OS/2 (ArcaOS these days) actually has a “Personal Edition” but they get it totally wrong. It is $150 first year and $50 a year after that. Totally insane. It is hard to tell if this is just Arca getting it wrong or if they are beholden to some royalty scheme that forces them to do this.
If there was a free or even reasonably priced version of ArcaOS, I think there would still be a thriving Open Source community around it that would make it far more viable for everyone (inclding the commercial users).
QNX is a good example of an OS that still has commercial valuethat would absolutely benefit from a free non-commercial license.
This OSnews story is about OpenVMS being free during the beta period. But OpenVMS does actually have a non-commercial license for the full version as well:
https://vmssoftware.com/community/community-license/
It is a bit annoying as it is timed and you have to frequently re-apply for it but honestly having to put in a bit of work to get access to it is not a terrible trade-off. It allows them to keep the “free” version from becoming widely available (still license controlled) while giving legitimate non-commerical users a path to evaluation. I find that totally fair..
I am pretty sure that the OpenVMS Community License gives you access to the full system (including dev tools) on x86-64. Honestly, it does not seem like that big a mess to me. It is certainly better than ArcaOS is these days.
LeFantome,
Yes, that “free for personal use” would be awesome. But that would also be shooting themselves in the foot.
Because that could easily cause a drop to 10%, 20% or even 50% in legitimate sales, as many small shops, “solo” developers, and “I just need a small part for my business” people would claim to use them for personal reasons.
vmware had this (before takeover). You could get “free” licenses for many products by signing up, and using their web service to download one update at a time.
(They even had VMUG getting free licenses, but that is also now more locked down):
The line between “hobby” and business is very thin, and if they don’t have an existing viable stream of money (like Windows), OS developers seem to push the benefit of doubt to the other direction.
The only purpose of OpenVMS is to make money off of legacy applications. Well it’s too bad there is no hobbyist option for VMS.
@AndrewZ
Sorry, I have another comment in moderation (I assume Thom did not like the browser I was using as he, I think wrongly, sees it as political).
There is a Community License for OpenVMS. It requires registration but is no-cost for non-commercial use:
https://vmssoftware.com/community/community-license/
On the front page of the VMS Software website, it says “OpenVMS V9.2-3 is available and running on hundreds of servers worldwide and the Cloud”.
VMS used to power something like a half a million servers. I was surprised they were openly saying that this is down to “hundreds” as if that it is a good thing.
I found this link showing that there are 297 commercial customers for OpenVMS:
https://data.landbase.com/technology/openvms/
Granted, those are some pretty big customers and I have no doubt that OpenVMS is still running some extremely important infrastructure. It is clearly still a big enough business to keep a company afloat.
I am of two minds. It is amazing to see that something that used to run a material fraction of the worlds servers is now down to “hundreds” when a single AWS customer might have hundreds of EC2 instances running Linux these days. Then again, the fact that an operating system first released in 1977 is even still being used commercially at all is a pretty astounding fact as well. VMS turns 50 in just over a year!
Unlike UNIX, which has highly fragmented over the years, OpenVMS is the same “distro” as it was when it was released back in 1977 with it being built from a single code base this entire time. There is probably some BLISS code left in there. BLISS pre-dates the adoption of the C language but is still available for OpenVMS (now with an LLVM back-end). UNIX has taken over the world in the form of Linux and BSD but, in many ways, OpenVMS is the real survivor.
IBM z/OS is the current version of OS/360 which is even older than VMS, and also still has some (generally rather large) customers running it.
Unix is even older, but is survived today by multiple clones rather than the original.
VMS never had any clones, although Windows NT shares a lot of similarities.
> Unix is even older,
No it isn’t.
PDP-7 Unix, “version zero”, was 1969:
https://gunkies.org/wiki/PDP-7_UNIX
Version 1 was 1972.
https://gunkies.org/wiki/UNIX_First_Edition
IBM announced OS/360 in 1964:
https://www.britannica.com/technology/IBM-OS-360
And released it in 1965:
https://users.cs.jmu.edu/abzugcx/Public/Student-Produced-Term-Projects/Operating-Systems-2003-FALL/IBMOS360-by-E-Casey-Lunny-2003-Fall.pdf
I have to admit, it’s closer than I thought, but no: OS/360 is a 1960s product and Unix is a 1970s one.
@bert64
I was going to argue that z/OS is not one continuous code base but, after some research, it seems that it is.
The OS/360 -> MVS -> OS/390 -> z/OS progression represents one continually evolving and expanding operating system. Truly amazing.
So you are correct, even older than VMS. And, in fact, older than UNIX.
Linux is a UNIX clone that has nothing to do with UNIX from a source code perspective. But AIX and Solaris (so Illumos) can be seen as evolutions of the original AT&T code base. They are not really “clones” so much as forks. I mean, you can say that about FreeBSD as well although it is truly a Ship of Theseus.