HP has released a roadmap outlining future development of OpenVMS, the operating system that commercialized many features that are now considered standard requirements for any high-end server operating system. (Such as Integrated networking, Symmetrical, asymmetrical, and NUMA multiprocessing, including clustering, distributed file system (Files-11),
Integrated database features, support for multiple computer programming languages, hardware partitioning of multiprocessors, etc). With over 30 years of development, OpenVMS has stood the test of time and has continued to evolve as one of the most secure and trusted mission critical OS’s of our time.
OpenVMS 8.4 Roadmap Announced
28 Comments
-
2008-10-12 2:55 pmsbergman27
[q]And if you run Windows, some might argue that you’re running a logical descendant of it:[q]
Others would tell you to wash your mouth out with soap.
-
2008-10-12 4:06 pmmbpark
Sbergman,
Which is why I posted the archive.org link to Mark Russinovich’s article that compares the internals . He’s the one Windows expert that could write that article and make the argument as convincing as he did.
His current employer would probably not allow such commentary.
-
2008-10-12 5:21 pmronaldst
Thanks for the link to a very good article by Russinovich.
http://www.win2000mag.com/Articles/Print.cfm?ArticleID=4494“ rel=”nofollow”>http://web.archive.org/web/20020503172231/
-
2008-10-12 6:32 pmsbergman27
Ronald, if you are going to troll, at least try to make it less obvious.
And I believe that this article is about OpenVMS?
-
2008-10-13 8:03 amSoulbender
I would say it’s too bad MS own hack programmers fscked up the good foundation Cutler’s team created.
nice OS, I would like to see a port on x86 (pleaseeeeeee)
Although that would be great, I’d say that over half the performance and stability is due to the combination of the hardware and software rather than simply the software alone. I remember writing an assignment on OpenVMS, and the history of DEC’s hardware. From the early days to VAX then through to Alpha. The only let down DEC had was that it was a company run by engineers – resulting in great products but marketed so poorly.
With that being said, they could do a x86-64 port but it would require very very narrow parameters, and it would only run on a very small range of hardware – then at the end the question could be asked, would it make businesses sense? I guess there have been questions raised like this in HP but the business boffins have number crunched and decided it wasn’t feasible. With that being said, it would be interesting to see once Intel moves to the single motherboard platform where Xeon and Itanium can be swapped – it’ll mean that Itanium processors will become accessible through retail channels; be it they be very specialised.
Edited 2008-10-11 23:17 UTC
-
2008-10-12 7:14 pmPlatformAgnostic
Itanium is a cool and funky architecture, but average people aren’t going to want to run them. For one, they produce an appreciable and surprising amount of heat (my relatively early generation HP Itanium2 machine produces more heat than 4 other machines put together).
-
2008-10-13 8:52 pmdekernel
Sorry, but I think people in a professional enviroment will run an Itanium with no questions.
-
2008-10-12 8:14 pmsergio
I’d say that over half the performance and stability is due to the combination of the hardware and software rather than simply the software alone.
Sysadmins have to tatoo that quote on their chests. It’s the key of any reliable system. That’s why OpenVMS, zSeries and UNIX boxes rule.
Wintel and Lintel boxes lack that deep hard+soft integration. And please, don’t tell me that Linux boxes are stable and great. They suck and suck really hard. Sorry.
PD: I love GNU/Linux, but the truth must be said.
-
2008-10-13 2:44 amkaiwai
I’d say that over half the performance and stability is due to the combination of the hardware and software rather than simply the software alone.
Sysadmins have to tatoo that quote on their chests. It’s the key of any reliable system. That’s why OpenVMS, zSeries and UNIX boxes rule.
Wintel and Lintel boxes lack that deep hard+soft integration. And please, don’t tell me that Linux boxes are stable and great. They suck and suck really hard. Sorry.
PD: I love GNU/Linux, but the truth must be said.
I remember when Linux first came out – most people here ignore who used it first, and the reason for it – thats not to say that there is anything wrong with Linux, it is just the fact that once you get to a certain level of reliability as part of your system requirements; it seperates the boys (Linux and Windows) from the men (OpenVMS, Solaris, AIX, HPUX etc.).
Yes, there are some vendors who are producing very reliable hardware for the Windows and Linux world – but the cost of that hardware is so high, the difference between that an a UNIX RISC system is so small, one might as well just go for the UNIX system and be done with it.
-
2008-10-13 7:43 amVanders
And please, don’t tell me that Linux boxes are stable and great. They suck and suck really hard. Sorry.
PD: I love GNU/Linux, but the truth must be said.
Take a look at all those supercomputer clusters out there, with anything between 1000 to 4000 nodes in them, all runing Linux. These systems don’t have a habit of failing on a regular basis.
-
2008-10-13 7:49 amPlatformAgnostic
I’m not so sure… if you have a well-designed cluster a few nodes falling over for whatever reason should have no effect on the overall system.
-
2008-10-13 9:12 amVanders
Depends. If your head node(s) fall over you’re in a lot of trouble. The fact still remains that failures for nodes running Linux are not significantly high.
-
2008-10-16 2:51 pmabraxas
Sysadmins have to tatoo that quote on their chests. It’s the key of any reliable system. That’s why OpenVMS, zSeries and UNIX boxes rule.
Wintel and Lintel boxes lack that deep hard+soft integration. And please, don’t tell me that Linux boxes are stable and great. They suck and suck really hard. Sorry.
The difference here is that Linux is slowly taking over that space because the hardware companies can actually submit changes to integrate their hardware better with Linux. Even if the code is not accepted, vendors can maintain patchsets for their hardware. This is much more difficult for Apple or Microsoft operating systems becaues you have to rely on the OS vendor to make these changes. Over the past 10 years a lot of enterprise features have been added directly from integrated UNIX systems like Solaris and AIX.
Really, this turkey is cooked, but I know it’s got like 30 installed sites.
It could be positioned as the AmigaOS Server platform.
-
2008-10-13 2:16 am
-
2008-10-14 9:59 am
I hope HP follows through with VMS and don’t lose the ball like they did with Alpha and Digital UNIX (TRU64).
If I could run my business on VMS I would do so immidiately, it is in many ways superior to Win/Lin/Mac, but . . . . .VMS is a Server-OS and no Desktop-OS.
The combination that Linux offers, desktop+server for a very small compagny is just not the thing VMS was made for. I loved VMS for what I could do with DCL and the sys-internals but that was in a 400-user environment.
If HP lets me run a “free” copy (even virtual) I would do so, until then I will stick to Linux. What would happen when HP would turn VMS open source?
-
2008-10-13 6:37 pmsmashIt
If HP lets me run a “free” copy (even virtual) I would do so, until then I will stick to Linux. What would happen when HP would turn VMS open source?
hp had (don’t know if the still have) some kind of hobbyist-programm where you can get a media-set for a few ~Ac^A'^Anot
all you need is a vms-capable box and a free membership in some vms-usergroup
-
2008-10-13 7:38 pmpoundsmack
http://www.openvmshobbyist.com/licenses.php
thats the one. (goes back to tinkering with his OpenVMS machine)…
Faced with porting your operating system from its old minority CPU platform do you (a) Port to the platform with the largest ongoing investment that runs the worlds largest collection of software natively OR (b) select another minority low volume CPU that almost no one understands – mainly because you invented it.
This should also be the first version that can run virtualized (using HP Integrity Virtual Machines), so you can run Linux, Windows, HP-UX and VMS on the same box at the same time.
-
2008-10-14 9:35 amopenbsd-user
csynt==”nice OS, I would like to see a port on x86 (pleaseeeeeee)”==
Virtual OpenVMS hardware for x86 intel/amd processors:
Charon AXP http://www.charon-axp.com/ http://www.emulatorsinternational.com/en/personalalpha.htm
or try the ES40 emulator.
ES40, emulating the Alpha AXP EV68CB processor and devices http://www.es40.org/Homepage
Desktop OpenVMS runs X clients too!, on every great Xserver.
It was a fine platform for developing. Excellent, optimizing compilers, easy to build projects (MMS, CMS), easy to create and link to shared images without fuss…
I learned all the tricks of the trade on VMS. Apparently it IS standing up to the test of time. If HP is bothering to maintain it then it must still be making them a lot of money.
Some would consider me a Mac fanboy and I do love the Mac but I’d take OpenVMS as the underpinnings of Mac OS X any day over Unix and that’s not to take anything away from Unix.
I’m am a current sysadmin for 3 OpenVMS servers, nearly 60 Windows servers and just a huge personal fan of the Mac but OpenVMS is awesome. The BASIC compiler for VMS kicked serious butt! The only thing it lacked in my opinion was pointers.
I did all of my programming on OpenVMS using BASIC, Macro (aka assembly language) and yes good old DCL. You haven’t lived until you’ve modified keyed binary files in DCL (grin).
Rock solid, runs forever, never breaks, secure and no headaches… now that’s OpenVMS in a nutshell.
nice OS, I would like to see a port on x86 (pleaseeeeeee)
There’s FreeVMS, available at:
http://www.systella.fr/~bertrand/FreeVMS/indexGB.html
And if you run Windows, some might argue that you’re running a logical descendant of it:
http://www.win2000mag.com/Articles/Print.cfm?ArticleID=4494“ rel=”nofollow”>http://web.archive.org/web/20020503172231/