Slashdot Mirror
Operating Systems Are Irrelevant
zincks writes "David Gelernter (Yale Professor of Computer Science, and Unabomber target) has a
story in the NY Times which states, (1) Operating systems are relics of the past, (2) We should be able to access data anytime/anywhere, by (3) seeing a stream of 3D documents(?), so (4) he's written such
software, and (5) that's all you should care about so it doesn't matter that it runs under windows.
This is a fantastic (definition: based on fantasy : not real (?)) vision of the future by a premier technologist."
I run a large particle physics laboratory. As you can imagine, we have a lot of computers. Some of them are traditional desktop PeeCees for checking email and viewing pr0n. But a lot of them are data gathering, collating and even simulation machines. These boxin' are Big Iron but there's no need to waste cycles on an Operating System when all that power could be directed towards running software.
Similarly, the more successful PDAs and cellphones don't have room for a lot of overhead, so the Operating System is dispensed with. There's no hard drive anyway, so what would you need it for?
If I'm misreading the article, someone please enlighten me, but it sounds like what he's really talking about is a modified file system and new searching methods based on that file system.
If this is the case, then an OS still needs to run off of that file system, so the OS is clearly not dead.
This is what longhorn's filesystem is supposed to do: It's SQL and metadata-based. I don't see how that's making the OS irrelevant. I think the author could have chosen his words a little better.
Well, he is pretty much thinking along the lines of Microsoft, which is gradually shifting to an OS that really does not exist but gives you the functionality you need based upon the services you require.
.Net wallet has been charged, thank you.
For example, if you need to write a Word Document (yeah yeah XDocs in Office11), you would boot up your computer which basically would make a call to a Web Service that will show you what you call the desktop (i.e. presentation layer) of your OS today e.g WinXP, Win2k, etc.
You need to write a Word Doc? Do you subscribe to the Word Web Service? If so the menu item in the program group will be there (Start-Programs-blah blah), you consumed it when your WebServiceOS came up, because you subscribe to it so you can go ahead and make a word doc. Thus, whatever data you need will be accessible when you want it, for a certain price that is.
Theoretically, this may seem like a great idea, software as a service, revenue for MS, you get only what you want i.e no bundled overpriced office products, but then again...oh nevermind.
And oh yeah, you can get your documents anywhere in the world since your profile will be associated with your ".NET my Services" account, so as long as a computer is using this next OS, which will probably come after longhorn, you have what you need everywhere..all you have to do is Consume and Subscribe! Theoretically although the vendor is Microsoft, is XML over HTTP really Microsoft Windows? No! Lets just call it MSWSVOS (Microsoft Web Service Virtual Operating System)...your
Solid!
The "memory enhancement" experts going all the way back to ancient Greece have long known that spatial cues are a powerful way of organizing human memory (especially in males). So any information organization metaphor that is spatial or geometrical would exploit this principle. That would mean the "desktop" or "office" isnt that bad, though there might be better ones. Perhaps rooms in a house or streets of a city (geocities@yahoo) might be more tangible than a desktop, but probably not worth the change involved.
...and now everything looks like a nail. And the things that can't be made to look like nails he sees as unimportant.
It's a common enough malady among geniuses that have been too long surrounded by people telling them how smart they are.
--
CPAN rules. - Guido van Rossum
The MS trial was about business and politics. From a technology standpoint it was, indeed, pointless. One of the key points was whether it's ok to bundle an internet browser in the OS. If that is the thing to do logically from a technology standpoint, somebody should do it and move on. Even the distinction between the two is a pointless one to make, technologically speaking. It ground you in the concepts of "this is an OS, these are the functions of an OS...this is an app, these are the functions of an app...." when in reality, technology should be free to stand all that on its head if it makes sense.
And how is the OS irrelevant? Maybe to him it is, and to the home user
I think that was his point -- that it *should* be irrelevant to the user, but isn't.
www.HearMySoulSpeak.com
I knew this sounded familiar, so I did a search for his name on Slashdot. Yup.
Here's a similar article from December 2001.
And another from July 2000.
And I predict another one will be posted in October 2003.
Thank you.
"And like that
Maybe he is being very idealistic. Maybe he is being unrealistic. Maybe this thing will flop and fall into the bunghole of history.
/.) Microsoft squashed a concept for almost a decade. Maybe the hardware wasn't advanced enough yet, maybe it would have been a Newton. But maybe it would have been a Palm. Now we'll never know. What other innovations are we missing until Microsoft deems it 'time'?
But it's still good that he's doing it.
Someone has to question how things can be made better. Perhaps the worst thing about Microsoft is that the Windows desktop has pretty much stopped that questioning. This works in two ways, by Microsoft deliberately squashing competition and by people getting too comfortable inside the MS box. (including GNOME and KDE)
Nor is it an adequate argument that the Windows interface (even as embodied by GNOME and KDE alternatives) is "good enough" just like the steering wheels and clutch/brake/gas pedals of a car.
Back in the early-mid 90's there was a company trying to introduce Pen Computing - flat screens operated by a stylus. (I think the company may have been Go, but I'm not sure.) They were put under by a piece of vaporware called, "Pen for Windows" that never materialized, at least not until that Microsoft Innovation in the past month of Tablet-XP. (or whatever it was reported as on
The living have better things to do than to continue hating the dead.
Actually this is a highly accurate explanation of the problem with his ideas.
Far be it from me to say someone so much more educated than I is a complete idiot that doesn't understand no one WANTS to stop using operating systems because they are what PAYS for the technology.
Yes, even Linux is supported because people are willing to pay for the server space and bandwidth. These people, and universities, are getting something out of their association with Linux.
Never confuse volume with power.
Once people started buying these 386's and what not en masse, demand for more powerful software increased, which demands more powerful hardware, etc etc etc.
So no, MS's R&D department didn't figure out how to clock chips up to 3GHz, but they did a whole lot to create the demand for that kind of hardware.
Doesn't it?
I tend to agree with your assessment. My college professors (way back when) seemed to have this impression of PC O/S's in their infantcy. It is quite ironic now that many of them still hold their die hard beliefs that Windows are just bells and whistles. For example, the professor who teaches Intro to Object Oriented Programming doesn't believe in Objects. Another professor teaching Assmebly needed to write IO libraries for the assmebler in pascal, because doing that in assembly was "below" him. This guy went on to make us implement advanced data structures like graphs and trees in assmebly instead of teaching what assmebly was really about...doing low level crap that higher level languages can't do, or don't do efficiently.
From my experience with college professors, I have a hard time believing that this guy has actually written real software. At best he's got a couple of whacked out ideas, and is making graduate students implement them. None of my professors ever wrote anything that could be considered software. There's a lot more to writing software than programming an algorithm that calculates the day of the week for any given date.
Back in '96, Apple introduced a similar concept. They called it Project X, then renamed HotSauce. I remember it being a browser plugin that allowed you to fly through space, seeing documents and websites as 3d objects. Here's a google search that turns up more info.
Someone, please shake me from this wide-awake nightmare.
Windows is the marketplace victor and has now won a decisive legal imprimatur. There is no technical reason for us to move to Linux; why should we switch? Why should our customers?
Oh, because Microsoft has a Monopoly and we should just accept that because:
Windows is a reliable, solid, reasonably priced, nearly universal platform - and for the software future, "universal" is nonnegotiable. We need to run the system on as many computers as possible and manage the maximum range of electronic documents.
Ah, gotcha, they already have a monopoly, and we all kinda need one anyways because we all need to run the same software so we should all just stop this pointless flame war complaining about lack of choices because choices break apart our vision of a unified digital playground of knowledge. So everyone run Windows because our new visionary software only runs on Windows because:
Windows is a reliable, solid, reasonably priced, nearly universal platform
Although we already said that, but we thought we should say it again because its really important. So is everyone ready to stop wanting choices and merely accept the new hand crafted future built just for us so we can stop working on our own visions because this one is the very bestest and is the one true software we all need because we all need the same software for this to work... blah, blah, blah.
Phew. Ok, breathe... and exhale. Good. We now return you to your regularly scheduled reality.
A steaming cup of soykaf would be real wiz right now.
You hit the nail on the head, it's Microsofts protecting it's monopoly that's preventing any new model from rising. To protect the Windows OS monopoly and therefore protect their application( MS Office ) monopoly, the OS must be relevant.
Just like C++ and object oriented frameworks threatened Microsoft/MS Windows in the early 1990's by abstracting the OS API's, OpenDoc threatened them in the mid 1990's by abstracting the OS AND greatly reduced the barrier to entry into the application space. The full force of the FUD machine and purchasing power put the cork on those two ideas. Granted, OOP made a decent comeback with Java and Troltech is making a living at a C++ framework but we still require huge applications with redundant features to read/write documents.
I've seen and used OpenDoc and the concept of data-centric computing is smart and far easier to use. The problem persists as to what to do about Microsoft's continued slowing of progress?
BTW, I've helped a few small business's in the early 1990's in streamlining how they used computers( PC's ) and it was the OO desktop that saved the day. Where OS/2 could be installed, it was or else it was HP's NewWave OO desktop manager. In both cases, I implemented data-centric templates of folders and data objects/icons so the use concentrated on the DATA for the task and not what application needed to be started and where that file needed to be saved to, etc. The Data-Centric method worked and worked well.
The browser is kinda like the OpenDoc container but without the Bento filesystem to hold all the different apps(Parts) data. The Inet becomes the replacement for the Bento filesystem. I don't think plug-ins can use other plug-ins in a standard way so that for instance, one spell-checker plug-in could be used in the text/html editor AND email plug-ins...
LoB
"Anyone who stands out in the middle of a road looks like roadkill to me." --Linus
Quoting reallocate: As I take it, his basic point seems to be that both Windows and Linux are based on OS concepts developed at least 30 years ago
When talking about technology we need to be careful about looking at when concepts were developed. Things like the wheel and agriculture are thousands of years old in concept and yet we are still finding new ways to improve on that original concept. While computer technology has seemingly improved at a very fast rate, I would submit that it is still in its infancy and that we ain't seen nothing yet.
Given the perceived speed of technology advancement and the age of the basic operating system paradigm it can be tempting, I think, to view OSes as dinosaurs. I'm all for people promoting new ideas but think that we should all be careful about knocking technology because of its age. Whatever you're knocking could be the new wheel, somthing that'll still be around long after you are dead. By the same token we should be careful also about knocking new ideas that may be the technology of the future that no else was able to see.
Never disturb your enemy while he is busy making a mistake.
One of the largest barriers to the adoption of OpenDoc was the business model for software providers. The commercial distribution of software at the time couldn't support a large number of lightweight apps, plus it wasn't clear who to call in case of a problem. The large application developers also cared about the branding of their products, something lost in the OpenDoc world.
His product is called Scopeware.
I am a beta tester.
It is kinda cool. Basically it turns your personal computer, or all the computers in an enterprise into a searchable internet. It indexes everything -- documents, powerpoints, email, mp3, jpg, etc etc.
You can search once, and it'll bring up all the results in order of time created, or relevance. So, you can see email that are related to documents and powerpoints -- and they are related by the search term.
HOWEVER, the index file takes up to 1/3 of the original filesize!!!
To index my 300 Gig home network could take up to 100 Gig.
Wheeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee
geeks are cats who dig a certain kind of cool
This notion that a file is "located" in one place only, and therefore you need two files to locate the file in two places is already false. You may have heard of a little-known family of operating systems called UNIX. Their filesystems typically have a thing called linking. If you "put" a file in a directory you aren't *really* putting it there. You are just putting one *reference to* the file there. And that one reference to the file doesn't have any precedence over another one you may make later somewhere else. When you "remove" a file you don't really remove it. You just drop the reference to it that was in that directory. Only after the file becomes fully orphaned (all the references go away so you could never find it again anyway) does it physically get deleted. It's like languages with garbage collection. The concept already exists and people don't use it much. I think that speaks volumes. The notion that a thing can exist in multiple locations at the same time is counter-intuitive. It's incredebly useful, but not the sort of thing a lot of people are going to "get".
I just wish hardlinking had some means of following the reference bidirectionally, so that given a file the system could efficiently tell me all the linked names it has. Right now the only way to do that is to scan the whole filesystem for other filenames pointing at the same inode number, and THAT is horribly inefficient.
Don't label something "offtopic" unless you know the topic well enough to tell what's on topic.
A phone is for talking to someone else. Simple concept, simple operation (with additional features). What does a computer do? You cannot answer that with a single answer. I avoided the computer/phone analogy for that reason.
Simplicity comes when you find opportunity to say "this is the way it works. Always." People can understand that. We geeks have this Utopian vision in mind where everything at our fingertips is infinitely customizable, but where did we get that? I don't have 12 different ways of playing media on my television, but I'm not whining about it. My car's transmission comes in automatic or manual. I'm thankful for that choice, I'm not encouraging people to come up with more choices for me. Sometimes you just settle with what you're given. If it's really *that* bad, another choice will almost always surface. But if the existing choices work for most users in most cases, why keep adding new choices unless you're demonstrating that they're better?
Again, televisions are single-purpose devices. Of course the basic interface is the same. On/off , channel up/down, volume up/down. Again, what is the one thing a computer is used for? Internet appliances failed. Email stations aren't in high use. Typewriters were superceded by word processors, which were superceded by word processing software. Things are getting more and more complex, so you will not be able to come up with a simpler interface. Basic phone service is simple, but you have voicemail, call-waiting, caller-id, etc. I remember when we only had rotary dial phones. Now everything is tone dialing. So the basic operation changed a little. The use of mobile phones and messaging is HUGE it countries outside the US. We are left behind because of our stupid greedy companies who fear change.
Cars used to be manual. Then there was manual and automatic. Now there is manual, automatic, and "steptronic" style, which is a combination of the two. More complex. Records, 8-tracks, cassettes, CDs, Digital. It is getting more complex. But complex is not bad! The learning curve of our populace is what is holding things back. I try to keep up on it, but I know people who cannot. It just isn't in them to learn new things. My point is that computers will never be stable enough for long enough to come up with a single universal interface. Nothing else has, and they have been simple devices. Why would computers, when they are more complex?
My beliefs do not require that you agree with them.
The Registry is a horrid implementation of an *awful* idea.
You want to make a centralized database *cache* a la the MacOS desktop file, go for it. I don't like the idea of having a single, nontransferable crucial chunk of data that's a single point of failure. It's idiotic that you can't simply copy an installed application on Windows to another computer.
May we never see th
I call big BS on that one. The limitations of PC hardware and the Windows operating systems ARE relevant, and as long as I can't write to a floppy drive and listen to an MP3 at the same time, or the machine's default network sharing and mail systems leave me ripe for butt rape by script kiddies, I certainly won't be running this vision of the future.
Computers have been in use for more than fifty years. At fifty years cars were pretty much mature. Iron ships follow about the same timeline. Liquid Rockets, same thing.
jet plane. Skyscraper.
Spinning reels.
Not much has changed has it?
love is just extroverted narcissism
Enh. It's actually a nice argument, and one made previously by Raskin et al.; the desktop "OS" as we know it today is largely a glorified file manager ("Disk Operating System" mean anything to you?). [Of course, the OS is kernel + UI, and this is outside of kernelland.]
It's been said over and over; for the average desktop worker (NOT the average hacker), it might be better to just dedicate the entire system to managing workflow, rather than managing "files" and "applications" and other things the poor moronic user shouldn't be messing with.
Of course, just going to some goofy relational structure like Be had for a while might be the best compromise.
It's not an "OS" issue, it's a "UI" issue.