Slashdot Mirror
MIT To Release Next-Generation OS "Cesium"
snowphoton writes: "Slant-Six magazine has an article about Cesium, a fascinating (and soon public) operating system from the MIT Laboratory for Computer Science. A virtual machine, an object-oriented database-driven filesystem, and a 3D GUI mean that this isn't your father's operating system." This article doesn't address licensing, except to say that it "is due to be released by the end of the year for free," so it will be interesting to see just what "free" means here. Update: Yep, it's a hoax. Fun! Tricks are neat!
The immediate problems that I see are
- The UI will scare people. Unfamiliar == Scary.
- Hardware support. Hopefully drivers will get out there fast.
- Very new programming model. People can't build off of years of UNIX experience as with linux.
- People like what they've got, be it windows, linux, mac, or both. Change is scary.
- Is the VM slow? Probably not, but we won't know until it's out.
That said, I really hope these hurdles can be overcome. I'm betting ideas from this will be incorporated in to the more traditional OS's in the long run, but who knows? I'm certaintly going to check it out and maybe give it a spin on a spare partition (if that's even possible yet). A 3d GUI! I can't wait."I may not have morals, but I have standards."
The guys at MIT have done an admirable job in merging several of these cutting-edge technologies into a seamless system. Their OS looks fascinating and I'd definitely like to take it out for a spin. But it's worth noting that most of what they offer is already available under Linux. As far as I'm concerned, they're re-inventing the wheel.
-CT
Well, from the details in the article, it certainly sounds promising. I'm glad someone's finally going "very public" with ideas to replace the underpowered filing systems we're using now, for a start. It'll be interesting to see how their OODBMS system works out.
That does raise an obvious question, though. Given that we're all used to filing systems where you have a lump of data in some form and you give it a name, and beyond that, it's just a hierarchical arrangement, how does that translate to/from their world? Converting from PC <-> Mac <-> Linux is no big deal (although even then we have details that may get "lost in translation") but they all follow a similar paradigm. How do you map from such a system onto an OO set-up? Given that they mention supporting an office suite, clearly a necessity for any mainstream OS today, they must surely have considered this issue in some detail.
If you disagree, post your argument. (-1, Overrated) isn't your personal censorship tool for views you don't like.
How come there are no pictures at the slant-six story, and no mention of this project on the website at http://www.lcs.mit.edu/ ? How come a search on google reveals...nothing? Do you really think MIT could develop the most fantastical OS ever for and keep it a giant secret? They said version 1.0 came out in 1993...so where is that?
Also, why would a university bother to write an office suite for this project? How exactly does that qualify as new research, worthy of publication leading to a master's or Ph.D.? I don't think any grad student actually interested in graduating would waste his time on such a thing.
hey, timothy, next time do some fact-checking before you post this stuff...
We don't have a state-run media we have a media-run state.
This sounds great and all, but like someone else pointed out, there's absolutely no mention of this anywhere on the MIT website, and a quick scan of the MIT website and directory for Harvey Dunkirk (the author, and supposedly an assistant to one of the lab directors) turned up absolutely no hits. I'd love to be wrong, this sounds cool, but this sounds really fishy.
-David Ziegler
-
Here's why.
--
Throwing out everything from the file/folder/desktop heirarchy
Is it just me? I really can't see a distinction between the folder/file relatoinship and the parent/child data relationship.
And how long ago would someone have said "I don't know about variable fonts and font sizes for all human readable text. And what good is a GUI?"
There are, of course, a fair number of people who *still* feel that any GUI is a bad GUI, but even the most hardened CLI zealot has to admit that the GUI revolutionized the computer industry. If it were all CLI, computers would be nowhere near as widely used as they are now.
Not to say that Cesium's 3d gui is going to do what those first gui's did (epsecially since I've not even seen so much as a screenshot), but maybe a 3d gui is what we need. Traditional point-and-click has reached something of a standstill. Whether it's KDE, Gnome, Windows, BeOS, or Aqua, it's all still the same basic principles. I've yet to see a really useful 3d gui, but from the (admittedly very sketchy details) it looks like a lot of thought went into Cesium. Maybe someone's finally come up with something that actually adds value.
See Sig append. Append Sig, append. Good Sig.
Fact is, for people like me who read the score 4/5 comments, /. remains a pretty reliable news source, if you take those comments into account. This, as I've said, probably contributes to editor slacking, since they feel that fact-checking isn't necessary. However, there is a significant portion (majority?) of /.'s audience that does not read the comments. You can say, "This is their own fault," all you want, but some people just choose not to spend more time than is necessary to get a brief glimpse of the news. For these people, /. is spreading uncorrected misinformation, and it is doing a disservice to its readership.
I think that, if the editors don't want to spend the time to fact-check, they should at least post immediate update/corrections to the article, visible on the front page, when a glaring error has been made as in this case. That way, the people who don't read comments can at least be informed of mistakes.
That means nothing, I never bothered to get an Athena account when I was at MIT except for a brief period when I had a student who I could not give a lab account to. LCS and AI accounts frequently have bizare names.
However an assistant to the lab director should be able to name him, there are only two labs in the building after all and both labs have only one director. The position of 'assistant' to the director is secretarial.
There was no such project when I was at the lab. Some of the project bears a passing resemblance to Oxygen, however that has long since been public.
There have been several O/S developed in the building. None was developed in secret. A secret project could not get students to work on it.
There have been non-public projects going on in the building. However those are effectively long term consulting for various (military) agencies.
An O/S is simply too large a project to be performed as a secret project. It is a packaging job, there is absolutely no value in an O/S developed in absolute secrecy.
There is no-one on the faculty who comes to mind as a specialist in O/S research. There are plenty who have made significant contributions to specific areas of O/S design. ITS, Multics, Genera were all open collaborative efforts.
Another point that leads me to doubt the story is the absence of any mention of security which has been the principle O/S related interest for the building in recent years.
Finally, LCS is not in the habit of announcing its projects in obscurantist web journals.
The whole thing smells bad. It is just not the way the lab has ever worked. The idea is to get your ideas 'stolen', that is another word for 'used'. If you don't want to share your ideas then join the NSA
Looking for an Information Security student project suggestion?
Try http://dotcrimeManifesto.com/
Step One: :P)
:P)
We wet out pants over it (I know I did
Step Two:
We realised it was a hoax
Step Three:
Are we going to do something about it?
This sounds very cool, the OODBMS for starters is cool, it saves so much wasted space; the practicalities of the 3dGUI are a bit suspect but
all the other details are good.
Personally, I think it would be good to wipe the slate clean and start from scratch every so often. It is about time that somebody created a new OS FROM SCRATCH. It would be a mammoth undertaking but in the end we would have something beautiful, the next generation of OS.
(If I don't stop in a minute I'll cream my pants again