Guest post by VSI (the men and women porting OpenVMS to x86 hardware) has released an update outlining some of the issues so far in porting this old battleship of an operating system to x86 and liberating it from IA64.
This update provides a high level view of our current efforts to port OpenVMS to the Intel x86 hardware platform. The report highlights topics including: Compilers, Objects & Images, Early Boot Path, Virtual Machines, Dump Kernel, Paravirtualization, and Condition Handling.
Still a long way to go, but it is exciting for VMS fans.
As the owner of a more than 30 year old VAX 4000, which I haven’t booted in 10 years, this is great news.
Underneath a big table are 2 huge cabinets, one for the disks and one for the actual computer. One terminal and keyboard make it possible to connect to the computer. I never boot it as it take too much effort and the noise is makes sounds like a vacuum cleaner.
This is probably the way to go if you want to preserve a very old computer: try to make it work of newer hardware. Though I think it is also possible to run VMS on a software emulator.
Edited 2017-01-17 07:34 UTC
And VMS is still the only OS with a versioning filesystem AFAIK (no, I don’t think MVS is as good or as sane).
Please, someone, make it happen for Linux (hi TUX3, how are you doing?).
ISO-9660 (The original CD filesystem) supports versioned files. Probably specifically so it would support versioned files on VMS, and of course it’s read-only, but it’s in there.
just curious, what’s the benefit from having native FS versioning ? I’m guessing FS with snapshots don’t count (ZFS)
To me it is the inherent “beauty” in the implementation – a snapshot is a subset of versioning (IMO).
And the practical side is not having to think about remembering to snapshot; the N previous versions will always be there anyway whatever the activity.
You should ofc also have pure snapshots available as a point-in-time reference, either for backup purposes or for archiving.
But there is a distinct difference.
A snapshot captures the entire file system at a moment in time. File versioning is just that — files, individually.
So, you can have:
file.txt;1 file.txt;2 file.txt;3 file2.txt;1 file2.txt;2
what you don’t know is how they relate to each other. Such as the relationships of the files might be:
file.txt;1 file2.txt;1
file.txt;2 file2.txt;1
file.txt;3 file2.txt;2
File versioning is much like working with old school RCS (Revision Control System). It, too, simply worked on individual files. It was, essentially, a primitive versioning operation. When I used it, back in the day, I had to write scripts to correlate my entire project directory as a whole with RCS, RCS alone wasn’t adequate (CVS rose out of the need above individual files).
So, simply, a versioning file system is “neat”, but no panacea. It certainly has value. And, also, modern Macs “do this” at the application level, to a point. But it’s not an OS FS first class facility.
I remember VMS file versioning. It worked pretty well. You just has to PURGE every once in a while…
HAMMER supports fine-grained history.
Thanks, I didn’t know that. Even an “undo” utility. For interested readers: see under the “File History” header at
https://www.bsdnow.tv/tutorials/hammer
SCO and Genera used versioning filesystems FWIW.
With all the file crypto attacks, especially in Windows, it would be nice to have such functionality on VMS-lite (AKA WinNT etc).
But how does the file versioning work? Is it versions at the file name level? Or the inode level? The difference is critical, as applications tend to write new files in their entirety to a new temporary file (new inode) when saving data, then renaming the temporary file over the original file, which allows an atomic save operation. If versioning is at the inode level, you’ll just end up with lots of files with a single version, and no benefit for most applications.
NTFS was developed by many of the same developers behind TOPS-10 (the VMS filesystem), and the “versioning” from TOPS-10 is present in NTFS as “alternate data streams,” using a slightly modified syntax (colon instead of semicolon, and ADS’s can be named, not just numbered).
I’ve never heard of this before. Intriguing. Got any link to this in more detail, and does it actually work (and is anyone using it)?
This might be closer to what you want: https://github.com/artagnon/phoenixfs
There was (still is?) one hobby operating system which implemented file versioning into its native file system. It is O3One (The Object Oriented OS) http://o3one.org/ .
While the site stills exists on the web, it does not appear to have been updated since April 2004!. Interesting project although the “The” projects a bit of arrogance.