HP announced the release of OpenVMS 8.3-1H1. “HP is pleased to announce the release of OpenVMS version 8.3-1H1 for Integrity servers, the 30th anniversary [I got startled by the automatic video] edition of the OpenVMS operating system! OpenVMS version 8.3-1H1 includes all the capabilities of its predecessor, and introduces new features and hardware support.”
Posted from an OpenVMS 8.3 machine to boot 🙂
…that VMS was designed to run on systems from a company that did not believe in bringing computers to the masses. If there were any operating system worthy of putting on a pedestal for OS students to study, VMS would be it (rather than the unfortunate UNIX abominations we’re stuck with today). Sadly, VMS is no longer accessible enough or known well enough to be used in education. Maybe VMS would be a good candidate for Open Source.
You can get a license and the media for OpenVMS for peanuts from http://www.openvmshobbyist.org
OpenVMS is still a commercial system, and HP charges a pretty penny for the IA64 version of the OS. So I doubt they will be opening it anytime soon. Especially stuff regarding the clustering and volume mirroring/extension technologies.
Anyway, most of the system is not even written in C.I think it is mostly written in BLISS, and I doubt most people out there have even heard, much less used, the BLISS and MACRO languages.
Unix is not that bad as an academic tool.OpenVMS is really stagnant in a lot of areas, for example the shell needs a major overhaul. Plus the development tools are down right primitive compared to the stuff that you can expect for free from the *nixes out there. For example, trying to get a decent editor for VMS can be a bit of an ordeal… and it only goes down hill from there.
A lot of VMS bigots tend to make full of the lack of userfriendliness associated with Unixen. However VMS can be down right hostile for an inquiring developer.
I have an alpha system that I’d love to throw OpenVMS on, unfortunately getting a license requires that you are a member of an OpenVMS users group, which costs $90 here in the US:
https://www.encompassus.org/about/join.cfm
I’m just not willing to pay $90 for an OS that I’m just going to “play around” with.
“I’m just not willing to pay $90 for an OS that I’m just going to “play around” with.”
If you read the stuff, you can simply sign for a free associate Encompass account. It is free, as in zero, zilch, nada.
Once you get your free membership you can get the licenses no problem.
If you just want to putter around there are some freely accessible public shell VMS systems.
… from a company that did not believe in bringing computers to the masses.
Maybe. But to be fair, Digital got started when only the biggest companies could afford computers and only the high priesthood could touch them. They brought computers out a little further toward the masses, so they served a vital role as a stepping stone in bringing computers to the masses.
It’s not like VMS is completely dead. Many of the principles behind its operation live on in Windows NT/2K/XP/Vista/etc., at least at the lowest levels of the kernel such as the scheduler, the way it handles I/O internally, etc.
Well, VMS is alive and kicking really. I dunno how big its user base is though.
Cutler left for Microsoft in the mid 80s. So the snapshot of VMS that he took with him was less than 10 yrs old. The system has been on its own for another 20 yrs. So I am sure there is little similarities between the NT kernel of today and VMS’.
Furthermore, Cutler used some of the lessons learned as the basis for NT, rather than the actual theory of operation of VMS. There are some cultural similarities, but at the end of the day NT and VMS are two very different animals.
NT itself started as a portable microkernel research for RISC platforms, which was in some ways very antagonistic to what VMS had been up to that point: a monolithic kernel for a CISC platform. Furthermore, the commercial version of the NT system is far different to what Cutler’s team produced initially during their research. So that further dilutes the family ties between NT and VMS…
A lot of the stuff that makes VMS stand out came right after cutler left, esp the clustering stuff.
I can’t speak of VMS clustering and its relation to Cutler (he probably wouldn’t have been heavily involved in its creation since his focus is elsewhere), but I can say that NT was never a research OS. And the design is pretty similar to VMS. The major change which makes NT different is the Object Management subsystem. Otherwise, it was similar enough to spark a lawsuit and settlement between Microsoft and Digital.
University students could see this for themselves since universities can license NT sources (it’s called the Windows Research Kernel) along with a set of the original design documents written by Cutler and his band of engineers.
A close acquittance of mine worked under Cutler at MS. And honestly NT was a research OS during most of Cutler’s involvement. For most of its initial lifetime, NT actually ran on custom reference platforms designed by MS, first on the 80860 (that is where NT got its name BTW) and then on MIPS. The NT project was supposed to be a research environment for the OS (NT) and the platform (ACE/ARC) that MS wanted to push for the 90s. All of this was happening fairly deep between MS, and completely shielded away from IBM for obvious reasons.
If you were to use an internal version of NT from the late 80s or early 90s you would not recognize the system at all. It was basically the OS that Cutler wanted to implement inside DIGITAL. And that is where the lawsuit originates.
The lawsuit was not pertaining the similarities between OpenVMS and NT, but rather between MICA and NT. MICA was an internal research OS that Cutler had been developed inside DEC in the mid 80s.
In many senses NT is the anti-VMS. Other than both have DIR as a command in their shells there is really not that much in common between the two systems.
LOL … VMS Fanboyism
If you’re old enough to know about VMS, you should be old enough to avoid “my OS is better than your OS” debates.
I’m glad to see though that even fanboyism can be brought to VMS threads.
It is mostly an opinion, but when compared to VMS it’s a lot friendlier, trust me.
Once when I connected to a VMS machine in the HP Testdrive program, I couldn’t even execute a ‘dir’-like command.
It somewhat requires you to script even to list a directory.
What an absurdity, a shell is made for ease of use and to access resource not for scripting. I’d rather use a separate program for scripting if I really wanted it.
And to think that I was able to use HP-UX and Tru64 almost without an itch, by only using my previous experience with Linux, *BSD, a bit of Solaris and yes even a bit of that evilness called ‘SCO’ OpenServer.
And even if you don’t know the *nix basic commandline and you only know DOS one if you get a list of commands it’s quite easy to figure out.
To use the VMS shell, maybe you need to learn a whole new scripting language and that’s way unfriendly.
You don’t need to write DCL scripts to do stuff in VMS any more than you need to write shell scripts to use *nix. I ran VMS inside a VAX emulator and while the command line lacked some features (most notably any sort of auto-completion features) it was pretty good. BTW, if you couldn’t get a simple file list, you obviously don’t remember DOS… “dir” does just what you would expect.
Well it all depends on your background. You can get a person who cut his or her teeth on VMS and drop them into an unix shell, and they will experience the same bewilderment. A la “what is this ls crap?” or “This file system structure makes no sense.”
Familiarity should not be mistaken with user friendliness.
That being said. It was a running joke among VMS bigots about how user unfriendly unix was when compared to their precious DCL. More astroturfing than anything, IMHO. Both systems require a fairly steep learning curve. And with products like GNV, that create an unix-like environment, the VMS fellas seem to have relaxed a little bit their holier than thou attitude.
I have always found *nixen far more developer friendly. Up to recently |D|I|G|I|T|A|L| pretty much demanded your first born child in order to obtain a half assed C development environment. And to this day, developing/hacking on VMS can be in some ways an exercise in reviving the late 80s :-).
However, once you get a system fine tuned and doing what you need it to do, it just keeps on going. VMS is not a sexy system, it stays out of the way in the background. Which I guess it is a good thing…
PS. I can’t believe no one has brought the whole meme of how Windows NT is the evil step child of VMS 😉
Edited 2007-11-02 15:32
unix is user friendly, it’s just really picky about who it wants to be friends with
VMS shell commands are more similar to those of DOS and Windows than Unix. So you should use ‘dir’ command instead of ‘ls’.
“FreeVMS” can be obtained from here –
http://www.systella.fr/~bertrand/FreeVMS/indexGB.html
I haven’t tried it myself, but it seems to show some promise…..
Well Windows descended from MS-DOS which in turn descended from CP/M. Some people say CP/M was a nearly clone of an operating system running on PDP-15.
VMS descended from RSX-11 and RT-11.
Most similarities in command shell come from that epoch. DOS line of operating systems borrowed a number of commands like DIR, COPY, RENAME, ASSIGN, CLS, DELETE, TYPE, FORMAT, device names were separated by “:” sign (just like in RSX/RT-11) and command switches were separated by “/” unlike UNIX where “-” is used. CP/M even had PIP command which deepens the similarity. “8.3” format of filenames was directly borrowed from DOS-15, a DEC operating system running on PDP-15 (RSX and RT used “6.3” filenames).
Blackwizard,
Actually, if you examine the chronology, CP/M borrowed its user interface style from DOS-11, which predates it by quite a few years. RT-11 was a direct descendant of DOS-11.
– Bob Gezelter, http://www.rlgsc.com
With the 30th Anniversary of the initial release of VAX/VMS, now OpenVMS, there were comments made that IMHO are undeserved.
In July 2006, OSNews published an overview, admittedly authored by me, as part of their OS Contest, “OpenVMS — A System of Structure”, it it available online at: http://www.osnews.com/story.php/15222/OpenVMS–a-System-of-Structur… . It was written presuming a readership with limited or not exposure to OpenVMS, and was well received by the OSNews community.
OpenVMS, is as I implied in that article, heavily based on explicit architectural concepts that provide a large element of leverage to the developer. There are an infinite ways in which this leverage can be employed, both technical and managerially. In addition to my technical paper cited in the OSNews article, I offer a paper which appeared more recently in the OpenVMS Technical Journal, “Strategies for Migrating from Alpha and VAX Systems to HP Integrity Servers on OpenVMS” (available online via http://www.rlgsc.com/publications/vmstechjournal/migrationstrategie… ).
On the question of usability, many of the earlier comments have been along the lines of “it is not *IX”. No, it is not. That is not a bad thing.
Not to get into the Tower of Babel debate over tools, command languages, and utilities, this is a matter of taste. The same observation can be made about human linguistics, English is not French, nor is it Italian, and it is even at a far more distant connection to Mandarin, Farsi, Hebrew, Arabic, Swahili, Hindi, … . A comparative discussion of the usability of those languages would be equally pointless.
The utilities do exist, generally in both ported *IX tools, and the native OpenVMS tools, for development tasks. In some respects they do differ in underlying philosophy, but that is unsurprising.
OpenVMS is an excellent platform, for both development and production. As a stable base that can be relied on as a solid foundation for years and decades, it has proven its worth.
– Bob Gezelter, http://www.rlgsc.com
Edited 2007-11-08 12:30
In one of the earlier comments to this article, a comment was made to the effect that it costs US$ 90/year to be an Encompass member.
This information is erroneous. As noted on the Encompass US www site at https://www.encompassus.org/membership/benefits.cfm , Associate Membership is available for no fee by simply registration.
The Hobbyist site provides licenses for OpenVMS and many layered products for non-commercial use at no charge to those who have an Encompass membership number.
– Bob Gezelter, http://www.rlgsc.com