I guess you didn't notice the the ORIGINAL MESSAGE i was commenting on used the term "*nix variant". Don't tell me that "*nix variant" is more accurate than "unix variant" and anyway "*nix" doesn't and will never match linux.
I did notice. I did read it. I understood it, which is more than you apparently. *nix also censors the word unix. Your homework kid, is to tell us why.
Your second assignment is to come up with something either relevant or intelligent to add to the thread. It is not acceptable to do this assignment in crayon.
VirtualLarry
No Lifer
- Aug 25, 2001
- 56,587
- 10,225
- 126
Ah, but you see, *nix users understand wildcards and "globbing". Anyone who complains about the incorrectness of such, obviously does not. (Ok, ok, technically I suppose '*n?x' would be better, but that starts to look really retarded.)
Hehe. That too. Never can tell when some trademark lawyer will pop out of the woodwork, can you?Originally posted by: n0cmonkey
*nix also censors the word unix. Your homework kid, is to tell us why.
Originally posted by: VirtualLarry
Ah, but you see, *nix users understand wildcards and "globbing". Anyone who complains about the incorrectness of such, obviously does not. (Ok, ok, technically I suppose '*n?x' would be better, but that starts to look really retarded.)
Hehe. That too. Never can tell when some trademark lawyer will pop out of the woodwork, can you?Originally posted by: n0cmonkey
*nix also censors the word unix. Your homework kid, is to tell us why.
They're sneaky bastards alright.
Originally posted by: VirtualLarry
Ah, but you see, *nix users understand wildcards and "globbing". Anyone who complains about the incorrectness of such, obviously does not. (Ok, ok, technically I suppose '*n?x' would be better, but that starts to look really retarded.)
Hehe. That too. Never can tell when some trademark lawyer will pop out of the woodwork, can you?Originally posted by: n0cmonkey
*nix also censors the word unix. Your homework kid, is to tell us why.
i'm glad you agree that this thread is retarded
Originally posted by: user1234
Originally posted by: VirtualLarry
Ah, but you see, *nix users understand wildcards and "globbing". Anyone who complains about the incorrectness of such, obviously does not. (Ok, ok, technically I suppose '*n?x' would be better, but that starts to look really retarded.)
Hehe. That too. Never can tell when some trademark lawyer will pop out of the woodwork, can you?Originally posted by: n0cmonkey
*nix also censors the word unix. Your homework kid, is to tell us why.
i'm glad you agree that this thread is retarded
Larry mentioned nothing about the thread, just the tangent you decided you had to bring it down. Now, STFU or contribute. Thanks.
For someone that complained about someone else's spelling, that is horrid. :Q
I am definitely a Microsoft fanboy. I :heart: Microsoft. Really.
VirtualLarry
No Lifer
- Aug 25, 2001
- 56,587
- 10,225
- 126
ROTFL!!!Originally posted by: n0cmonkey
For someone that complained about someone else's spelling, that is horrid. :Q
I am definitely a Microsoft fanboy. I :heart: Microsoft. Really.
Btw, in a vain attempt to get this thread back on-track, I'm really curious, how will SkyFS's user-defined metadata be similar or different to what MS is promising with WinFS, and how similar is that to what BeOS had? (Having used BeOS, but not having played with the advanced filesystem features myself, though.)
It'd be great to see an in depth review of features between the three (hint hint AT article writers
).Here's a little bit of something stolen off of a post on OSnews:
I guess for one thing, SkyFS will actually exist, whereas the WinFS, which was supposed to be in Windows NT, and then Windows 2000, and then Windows Longhorn (and now....not even Longhorn) does not currently exist.
As far as compared to BeOS, probably the biggest difference is we have fused permissions and multi-user support into the filesystem, which did not exist in BeOS/BFS. Aside from that, really, it's quite similar to the BFS (since, in reality, that is what the base of it was made from).
VirtualLarry
No Lifer
- Aug 25, 2001
- 56,587
- 10,225
- 126
Originally posted by: n0cmonkey
It'd be great to see an in depth review of features between the three (hint hint AT article writers
Here's a little bit of something stolen off of a post on OSnews:
That's beautiful. I actually was working on a design for much the same sort of thing, for my "X-OS" project that I started in '96, after I left Turbine. I'll give 'em credit if they can actually manage to pull it off. They've had how many (announced) attempts thus far at pulling this off before WinFS? Five? And it's slipped so much, that it won't debut in the Longhorn client OS release either, like originally planned.
That's one reason why mine stayed in the design phase, to get everything "right" is actually quite difficult. Plus, my own design was even more far-reaching than MS's (globally-distributed, for one - think of combining P2P nets with a filesystem that powerful). And a few other things. I stopped working on it for a while, then picked it up again, and then found out about ReiserFS4 and WinFS, and kind of dropped it again. (Wish I had the money and will to patent some of those things back in '96, maybe BillG would be *my* b*tch nowadays. Too late for that though.) MS's problem is that they just can't seem to "think outside the box" enough - which is why Apple is still the primarily innovator in many spaces, and MS copies them six months later. (Ok, and "fringe" OSes like BeOS and SkyOS too - not to cut you folks out of the picture entirely either. Startup OSes have an advantage that way, they don't have to continue conforming to any particular "legacy" OS operation or design principles, things that have held back both Windows and *nix for years.)
It would also be interesting, to see how much of this that MS impliments piecemeal, and how much they embrace the "one singular namespace" philosophy that Reiser's paper discusses. The sad thing is, is that at the core of NT, there exists the Object Manager, and a true singular namespace like that, except that for whatever reason, MS has kept that hidden away from most programmers, by not directly exposing it as part of the Win32 API (that I am aware of, other than allow using of those path strings in OpenFile calls). That was one of the really nice things about VMS, the extensive "system logical" support. I think the Amiga supported something similar, although not as extensive.
I'm personally most interested in DragonflyBSD's ideas for extending FFS. The journalling would be nice, but live backups and clustering options would be pretty cool too.
I think comparing bfs, skyfs, winfs, and reiser4 would be a great topic for a paper. It's more of an ars thing than an AT thing though...
VirtualLarry
No Lifer
- Aug 25, 2001
- 56,587
- 10,225
- 126
Btw, the most difficult problem seemed to be this - how do you ensure logical consistency, in a globally-distributed, decentralized system, without a "central authority" - which would otherwise be a performance/scalability bottleneck, and potentially a single point of failure. The was the part that I struggled with, and I don't think that I ever came up with a "perfect" answer, but it does bring up the image of an analogy with human thought and conciousness, and how our brains, in a localized fashion, either agree, on the whole, or don't, at least generally-speaking.
Think of it this way - if, in this system, I'm allowed to simply "pick my own address" (assume IP address to make the comparison simple enough) - if so, then how can I: 1) ensure that I don't accidentally pick a duplicate address, or 2) even if I do, then ensure that it doesn't cause a problem, and related to that 3) if I know of another node, say a friend, with a specific IP address, and I want to connect to that person, how can I do so, given the IP address, if duplicates exist in the system? Should I automatically connect, at the same time, to all of the nodes in the system with the same IP address, and then perform some sort of secondary authentication checking to validate that I am connecting to the right node? Under who's authority is that validation performed? Could that also be spoofed?
Or should there be a pre-defined manner in which IDs are assigned? Assuming that a centralized authority is ruled out, due to the aformentioned issues - then how can IDs be assigned, and thus guaranteed not to be duplicates, and likewise, any "rogue" intentionally-duplicated/spoofed/incorrect IDs could be flushed out or blocked out of the system, such that the uniqueness of ID could be guaranteed within the system?
Those are some interesting and potentially-difficult questions to answer, and largely the biggest holdup to completing at least the design part of things. The possibilities that the solution to those problems points to.. a world in which there is no real difference between "local" and "remote" contexts, at least with regards to file/object-stores.. where everything is just part of this one, singular, huge, "sea of data", but where things like privacy and access-rights would be controlled by encryption, etc.
I'm kind of surprised that most P2P software is still oriented around "illegal file-sharing", instead of some global, unified, distributed-object store, that could serve as a basis for other, bigger things. The FastTrack v3 network, as used/implemented by Skype, starts to point in that direction too. OpenCola Folders was another project that looked promising, at the time. (Guess it's dead now?)
Think of it this way - if, in this system, I'm allowed to simply "pick my own address" (assume IP address to make the comparison simple enough) - if so, then how can I: 1) ensure that I don't accidentally pick a duplicate address, or 2) even if I do, then ensure that it doesn't cause a problem, and related to that 3) if I know of another node, say a friend, with a specific IP address, and I want to connect to that person, how can I do so, given the IP address, if duplicates exist in the system? Should I automatically connect, at the same time, to all of the nodes in the system with the same IP address, and then perform some sort of secondary authentication checking to validate that I am connecting to the right node? Under who's authority is that validation performed? Could that also be spoofed?
Or should there be a pre-defined manner in which IDs are assigned? Assuming that a centralized authority is ruled out, due to the aformentioned issues - then how can IDs be assigned, and thus guaranteed not to be duplicates, and likewise, any "rogue" intentionally-duplicated/spoofed/incorrect IDs could be flushed out or blocked out of the system, such that the uniqueness of ID could be guaranteed within the system?
Those are some interesting and potentially-difficult questions to answer, and largely the biggest holdup to completing at least the design part of things. The possibilities that the solution to those problems points to.. a world in which there is no real difference between "local" and "remote" contexts, at least with regards to file/object-stores.. where everything is just part of this one, singular, huge, "sea of data", but where things like privacy and access-rights would be controlled by encryption, etc.
I'm kind of surprised that most P2P software is still oriented around "illegal file-sharing", instead of some global, unified, distributed-object store, that could serve as a basis for other, bigger things. The FastTrack v3 network, as used/implemented by Skype, starts to point in that direction too. OpenCola Folders was another project that looked promising, at the time. (Guess it's dead now?)
VirtualLarry
No Lifer
- Aug 25, 2001
- 56,587
- 10,225
- 126
Actually, no. Link? The problem is not necessarily just generating a unique one, but a tamper-proof, provably unique one.
You guys know... the equivilent to BeOS R6 is out there. Check out this link. I'm not going to pay ~$120 to play with it... but it's out there.
Joe
Joe
The screen shots look good. However, what kind of hardware & software support does it have?
I remembered my old shool purchased a wide 24 pin colour printer for their Amiga that cost over 3K, and the driver wouldn't work (Amiga never fix or update the driver). My bestfriend & I played around with the driver and manage to get it to print 1/2 page (couldn't get it to print full page or multiple pages).
What good would it serve if it doesn't work with your equipment?
let's try to be less superficial then the "screenshot looks good" comments. after all, anyone can create a great screenshot even without a good OS, and vice versa.
I have an ongoing love affair with BeOS... at least with the whole history and concept of it is/was just fantastic and WAY ahead of its time. Ahh... what could have been but isn't.
Joe
Joe
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 22K
-
News NVIDIA and Intel to Develop AI Infrastructure and Personal Computing Products- Started by poke01
- Replies: 388
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
Facebook
Twitter