Submitted via IRC for TheMightyBuzzard
Or 2018 if you're brave. For now, we have a boot screen!
Story's a bit dated but being as they're in no rush, I don't see any need for us to be either. So, you lot think we'll ever actually get to play with a VMS box on cheap hardware or is this going to be another DNF situation?
Source: https://www.theregister.co.uk/2016/10/13/openvms_moves_slowly_towards_x86/
Previous coverage:
OpenVMS Not Yet Dead.
(Score: -1, Flamebait) by Anonymous Coward on Thursday April 20 2017, @11:10AM
What a flamefest this is going to be...
(Score: 3, Interesting) by NCommander on Thursday April 20 2017, @11:15AM (17 children)
OpenVMS is one of those platforms I actually admire for being rock solid and having a lot of nifty things that *nix doesn't, but mainframes do like versioned filesystems. I actually have an Itanium system, but the pain of getting hobbyist PAKs for it has left me in the case most of my VMS investigations is mostly reading articles about it, and occasionally using one.
Unfortunately, OVMS on x86 is about six years too late to be relevant. When they got stuck on the Itanium tarpit, a lot of places decided to jump ship due to cost, and with it, most of the ISVs that supported it. Without that install base, I don't think VMS has enough traction to be relevant because it had a good niche between "Linux stability" and "NonStop/Stratus VOS's never fail" for corporations that wanted a very sane and stable platform, but needed more features than NonStop could provide. There are a few cases where I considered OpenVMS for a client's needs, but the cost of hardware + lack of support ultimately had them go to AIX or a Linux distribution.
Still always moving
(Score: 2, Insightful) by Anonymous Coward on Thursday April 20 2017, @11:57AM (3 children)
If they really wanted to compete, the way to draw vendors back would be to make it an open(ish) platform, get it on as much different hardware as possible, then tout the features that make it rock solid and thus competitive against linux, unix, and windows.
But instead they still keep up the stupid facade that proprietary software that has long since fallen out of relevance is financially valuable as IP.)
(Score: 0) by Anonymous Coward on Thursday April 20 2017, @04:17PM (2 children)
They renamed VMS to OpenVMS.
(Score: 3, Informative) by tangomargarine on Thursday April 20 2017, @06:37PM (1 child)
It says "open" on the tin but it's not really open source. If you believe what random people on the Internet say:
https://community.hpe.com/t5/General/what-does-it-mean-OpenVMS/td-p/3265043 [hpe.com]
Not quite. POSIX had already been around for several years/releases. The "Open" prefix was added in an effort to state that although "proprietary", VMS was "Open" in the sense it compiled with virtually all recognised standards and would talk to just about anything. POSIX was part of the story, but certainly not the whole reason.
OpenVMS also means you can get the source listings easily for one or two $K. Try doing that with the AS400 and some other OSes; if you as a system admin can get them at all, it won't be easy.
"OpenVMS" means, "come look at me and see how I work." For a proprietary operating system, that is pretty amazing.
"Is that really true?" "I just spent the last hour telling you to think for yourself! Didn't you hear anything I said?"
(Score: 2) by kaszz on Friday April 21 2017, @02:27AM
Easily for one or two $K also mean few people will just download it to explore on a Saturday. And because of that the next step won't come to existence.
(Score: 5, Funny) by TheRaven on Thursday April 20 2017, @01:05PM (7 children)
sudo mod me up
(Score: 4, Interesting) by NCommander on Thursday April 20 2017, @01:18PM (6 children)
You're thinking segmenting that died with x86_64. Rings survived (and are still used in some cases)
Modern x86 has five or six rings depending on how you look at it.
Ring -2 (SMM)
Ring -1 (vmx/vmd)
Ring 0 (kernel space)
Ring 1-2 (drivers, theoretically)
Ring 3 (userspace)
Still always moving
(Score: 3, Informative) by TheRaven on Thursday April 20 2017, @01:33PM (5 children)
sudo mod me up
(Score: 2) by NCommander on Thursday April 20 2017, @10:09PM (4 children)
Can you cite sources for ring 1-2 not being usable on x86_64? I googled for it but found nothing.
Also, I used to develop firmware :)
Still always moving
(Score: 0) by Anonymous Coward on Friday April 21 2017, @09:07AM (1 child)
I think parent is mistaken:
https://www.virtualbox.org/manual/ch10.html#idm9820 [virtualbox.org]
For guest code in ring 0, VirtualBox employs a nasty trick: it actually reconfigures the guest so that its ring-0 code is run in ring 1 instead (which is normally not used in x86 operating systems). As a result, when guest ring-0 code (actually running in ring 1) such as a guest device driver attempts to write to an I/O register or execute a privileged instruction, the VirtualBox hypervisor in "real" ring 0 can take over.
(Score: 2) by TheRaven on Friday April 21 2017, @01:53PM
sudo mod me up
(Score: 2) by TheRaven on Friday April 21 2017, @02:02PM (1 child)
sudo mod me up
(Score: 2) by NCommander on Friday April 21 2017, @11:59PM
Huh. I actually have found contradictory information when I went looking . Xen's security post suggests the IRET mechanism is there and the Priv spot is still present in the GDT selector. [xenproject.org]
INT/IRET on 64-bit use the 64-bit IDT format which has a spot for the code selector in the GDT you can land in, and you can still setup multiple GDT entries in place. Won't having a specific entry in the GDT at Priv 1-2, and setting the sector via IDT give you an interrupt call which would pop you into the Ring 1-2?
I may have to break out the assembler to experiment.
Still always moving
(Score: 1) by khallow on Thursday April 20 2017, @01:39PM (3 children)
VAX's ability to do detailed security allowances by file/command was interesting and powerful, but boy, can it be abused by the clueless!
(Score: 2) by kaszz on Friday April 21 2017, @02:44AM (2 children)
Systems run by clueless people will always be a disadvantage to whoever has to endure them. The problem is clueless people, not software. Which in the modern era is solved by that even a handheld phone has more computing power than a early 1990s supercomputer (8*10^9 op/s).
(Score: 1) by khallow on Friday April 21 2017, @02:53PM (1 child)
(Score: 2) by kaszz on Friday April 21 2017, @03:16PM
So true, no default micromanagement capability. Thwarting incompetence since dawn of mankind.. ;)
(Score: 2, Informative) by mechanicjay on Thursday April 20 2017, @11:12PM
but the pain of getting hobbyist PAKs for it has left me in the case most of my VMS investigations is mostly reading articles about it, and occasionally using one.
Dude, it takes like 10 minutes to get hobbyist PAKs.
Once you get that, a nice gentleman from HP will even offer you an FTP download of install media.
Note: the hobbyist situation for VMS on amd64 is still an unknown.
--Jason
My VMS box beat up your Windows box.
(Score: 2) by EvilSS on Thursday April 20 2017, @11:33AM (2 children)
(Score: 2) by NCommander on Thursday April 20 2017, @11:40AM (1 child)
At least DECnet was relatively easy to setup.
Still always moving
(Score: 0) by Anonymous Coward on Thursday April 20 2017, @04:31PM
And then DECnet V happened.
(Score: 4, Interesting) by turgid on Thursday April 20 2017, @12:13PM (2 children)
Didn't Microsoft already hire Dave Cutler to do it?
I refuse to engage in a battle of wits with an unarmed opponent [wikipedia.org].
(Score: 4, Informative) by Nuke on Thursday April 20 2017, @01:48PM
Didn't Microsoft already hire Dave Cutler to do it?
That's what I first thought of. In 1988 the DEC VMS team was poached by Microsoft to write NT, and they took the VMS ideas, if not some actual code, with them. DEC were thinking of suing, but MS settled with them out of court. Windows NT (and hence XP etc) is what VMS became.
Microsoft never really wrote an OS from scratch. They bought DOS (and hired its developer, Tim Paterson) from Seattle computers, and hired the DEC team (headed by Dave Cutler) to write NT. Windows pre-NT was started as a GUI (inspired by Apple's Lisa and also VisiOn which Gates saw at a trade show) duck-taped onto DOS.
(Score: 3, Informative) by Uncle_Al on Thursday April 20 2017, @03:56PM
This is what Dave was working on at DEC West
It wasn't VMS, or for a VAX
http://bitsavers.org/pdf/dec/prism/mica [bitsavers.org]
(Score: 2) by Nerdfest on Thursday April 20 2017, @01:10PM
I remember some perhaps apocryphal stories regarding Windows NT's design, relating to Windows NT -> WNT -> VMS, doing the whole HAL/IBM thing.
(Score: 3, Insightful) by jmorris on Thursday April 20 2017, @03:05PM (3 children)
Mostly dead closed source platform might successfully port to x86_64 as it realizes no new hardware is likely for Itanium, which it had been hoping was its future. If you are unfortunate enough to have a large investment in hard to replace VMS software you probably care that you might be able to replace hardware for a few more cycles. If you actually have buildable source and staff to do the port. Nobody else should care, certainly nobody should get marketed into wasting developer resources on new projects for a dead platform.
(Score: 3, Interesting) by requerdanos on Thursday April 20 2017, @03:30PM (1 child)
[Will x86 VMS be released] or is this going to be another DNF situation?
Well, DNF came in 13 years late* and infinitely over budget, brought to market by a rescue team** instead of its creators, but it did get released, to everyone's surprise. Ars Technica called it "Barely playable, not funny, rampantly offensive." [arstechnica.com]
Now here's VMS being ported to x86 years later by another rescue team; if the analogy holds, there will be a release. Thus it's not "will it be released OR will it be like DNF" because though it took many years (the case with both DNF and any current VMS), DNF was released.
TFA does mention that "the x86 rewrite is a slow process."
* DNF Announced for 1998; released 2011 [a-13.net].
** DNF Designed, incompetently half-produced, and smothered by Apogee/3d Realms [3drealms.com]; Salvaged and released by Gearbox. [dukenukemforever.com]
(Score: 0) by Anonymous Coward on Thursday April 20 2017, @04:25PM
Thanks, your post prevented me from googling for that mysterious DNF operating system.
(Score: 2) by kaszz on Friday April 21 2017, @02:55AM
To be of use to organizations with existing VMS software. Wouldn't it require the system to emulate the instruction set of VAX-11 etc?
Even if one has the source code. Recompiling might not be a straight affair.
(Score: 2) by mechanicjay on Thursday April 20 2017, @11:18PM
My VMS box beat up your Windows box.
(Score: 0) by Anonymous Coward on Friday April 21 2017, @02:38AM
...for SoylentNews to migrate to VMS!
(Score: 2) by kaszz on Friday April 21 2017, @03:00AM (2 children)
How does VMS, Unix and Hurd compare on a design idea basis? virtualization, swap, device abstraction, private filesystems, permission structure, administration etc?
(Score: 4, Interesting) by mechanicjay on Friday April 21 2017, @04:22AM (1 child)
Well, one of the things with VMS is it's really a mid-70's product, and a direct decendent of RSX-11 for the PDP-11 series of computers..and it shows. It was fairly well modernized through the 90's however, until DEC went tits-up. Then it became part of Compaq's portfolio, who, being a hardware company, had no idea what to do with it. Then quickly it became HP's product, where it competed against their own in-house proprietary OS's. So, it was basically left to languish by the year 2000. The port to Itanium never really went anywhere, as performance on Itainum, until the most recent iteration was basically barely on par with the old Alpha servers. Anyway, as to your design questions:
My VMS box beat up your Windows box.
(Score: 2) by kaszz on Friday April 21 2017, @06:55AM
How is memory management? ie where a memory section in your process memory may be fragmented to just about anywhere in physical memory or even disc blocks.
What makes it possible to migrate nodes? it must both transfer memory and instruction pointer, register setting, processor mode, device state etc and then there's multiprocessing. So it must more or less stop a processor, save all status to a communication line, verify and continue on another node. A lot of atomic operation issues and consistency. Though a modified scheduler on Unix should maybe be able to do the same?
(I'm thinking this has some similarities to the "Freezer" cartridges in the past)
Can a user mount their own filesystem, that only they can access? (like on Hurd) as for the rest it seems VMS more or less deals with the filesystem as a SQL database. Every storage object is a SQL row or similar paradigm.