Windows has ACLs and they are an important part of Windows administration, and used extensively for managing file permissions.
Windows has supported ACLs on NTFS since Windows NT & NTFS were released in 1993 (possibly partly influenced by AIX ACLs in the late 80s).
ACLs were not introduced to standard POSIX until c.1998, and NFS and Linux filesystems didn’t get them until 2003. In fact, the design of the NFSv4 ACL standard was heavily influenced by the design of NTFS/Windows ACL model – a specific decision by the designers to model it more like NTFS rather than AIX/POSIX.
Technically, at the filesystem level, exFAT also provides support for ACLs, but I am not sure if any implementation actually makes use of this feature (not even Windows AFAIK, certainly not any desktop version).
Yep, you’re right. I was thinking of an ACL evolution/chain of influence of VMS -> AIX -> NT, but it seems VMS -> NT and VMS -> AIX as two separate histories is much more accurate. Thanks for the correction – I’ve updated my comment accordingly.
Yeah, I’m familiar with VMS, and Cutler bringing a lot of the internal design to W/NT. (I’m told in particular a lot of the data structures for system calls in NT look like VMS.) My AIX experience has consisted entirely of “This is weird. This isn’t normal for Unix.” Ha ha. (I had a 1st gen RS/6000 at home briefly in the late 90s.)
Damn, giving me flashbacks of slowly moving through ACLs then hitting domain groups, domain local groups, global groups, then eventually universal groups as AD moved forward in complex situations.
Sorry, but this is completely wrong.
Windows has ACLs and they are an important part of Windows administration, and used extensively for managing file permissions.
Windows has supported ACLs on NTFS since Windows NT & NTFS were released in 1993 (possibly partly influenced by AIX ACLs in the late 80s).
ACLs were not introduced to standard POSIX until c.1998, and NFS and Linux filesystems didn’t get them until 2003. In fact, the design of the NFSv4 ACL standard was heavily influenced by the design of NTFS/Windows ACL model – a specific decision by the designers to model it more like NTFS rather than AIX/POSIX.
Technically, at the filesystem level, exFAT also provides support for ACLs, but I am not sure if any implementation actually makes use of this feature (not even Windows AFAIK, certainly not any desktop version).
Windows NT ACLs come from VMS.
The Unix world has traditionally not liked ACLs because Multics had them, and Unix was an ultra-minimalist response to Multics.
Yep, you’re right. I was thinking of an ACL evolution/chain of influence of VMS -> AIX -> NT, but it seems VMS -> NT and VMS -> AIX as two separate histories is much more accurate. Thanks for the correction – I’ve updated my comment accordingly.
My AIX experience is very limited. What was the VMS connection?
VMS implemented ACLs in the early 80s. It’s design influenced the design of ACLs in both AIX and Windows NT.
Yeah, I’m familiar with VMS, and Cutler bringing a lot of the internal design to W/NT. (I’m told in particular a lot of the data structures for system calls in NT look like VMS.) My AIX experience has consisted entirely of “This is weird. This isn’t normal for Unix.” Ha ha. (I had a 1st gen RS/6000 at home briefly in the late 90s.)
And I do have a “grey wall” in my library:
Damn, giving me flashbacks of slowly moving through ACLs then hitting domain groups, domain local groups, global groups, then eventually universal groups as AD moved forward in complex situations.
Got to admit it worked well though.