Post a Comment
The idea of using prizes to spur creativity and research is a technique that is known to be very effective. Witness the recent DARPA Grand Prize and the X-Prize tournaments. While the OS-Newsproposed contest is partially meant to be a promotion for OS-News (nothing wrong with that), it will also be beneficial in publicizing alternative OSes. Question is, does the world really need another me-too operating system? I don't think so. This contest runs the risk of being just another appeal to a relative small community of alternative OS fanatics who are in it mostly for the fun of it. What would be nice is a contest that offers a substantial prize ($millions a la X-Prize) for the development of a new OS that solves a particularly pressing problem in the industry. I suggest that the most pressing problem faced by the computer industry is not power management or processing speed. The biggest problem in the industry is software reliability. How about a $10-million prize for the first comprehensive OS that is guaranteed 100% bug-free? Not an easy thing to do and the rules would have to be worked out so as to be unambiguous on the meaning of bug-free. Let me suggest that OSNews get together with the X-Prize commitee (or DARPA, or whoever) and convince them of the potential bebefits of such a prize.
I'll be frank (shameless plug), I am proposing this contest because I believe I could put together a team to win it. It has to do with my faith in non-algorithmic, signal-based, synchronous software. I don't believe that unreliability has to be an essential charateristic of complex software systems.
Louis Savain
Edited 2006-06-16 18:47
RE: Great idea but does not go far enough
100% bug free? That's a tall order! Who's going to do all the testing to determine whether it's bug free. That's a fun idea, though. However, we already have lots of highly reliable OSes. The problem is, people want OSes with more features, and they'll generally choose them over more reliable ones, even for mission-critical stuff.
" The idea of using prizes to spur creativity and research is a technique that is known to be very effective. Witness the recent DARPA Grand Prize and the X-Prize tournaments. While the OS-Newsproposed contest is partially meant to be a promotion for OS-News (nothing wrong with that), it will also be beneficial in publicizing alternative OSes. Question is, does the world really need another me-too operating system? "
Does the world need more than one kind of car? Does the world need more than one color of car? Model airplane kit? Yo-Yo? Cable-knit sweater? Dishwasher design? Depends on end-user demand and idle time for development. If yet another "me too" OS improves the breed, I'd say so. Remember, all software is an incrementally improving effort, just exactly as all hardware is. It's easier to reach a pinnacle in hammer design, but software is a tad iffy-er.
" I don't think so. This contest runs the risk of being just another appeal to a relative small community of alternative OS fanatics who are in it mostly for the fun of it. "
Pray tell, who else other than a fairly lunatic fringe would even start to care? Look at us, Man! We're reading OSNews.com! Can't you see our pain?
" What would be nice is a contest that offers a substantial prize ($millions a la X-Prize) for the development of a new OS that solves a particularly pressing problem in the industry. "
First pressing problem: getting $millions!!
" I suggest that the most pressing problem faced by the computer industry is not power management or processing speed. "
I still kinda think it's getting the $millions.
" The biggest problem in the industry is software reliability. How about a $10-million prize "
Would you believe $145 and a free ride in the "Geek Squad" Volkswagen Beetle?
" for the first comprehensive OS that is guaranteed 100% bug-free? "
Define comprehensive. Define 100% bug-free. Would an ugly brown interface be considered a bug? Inability to read my thoughts? Being packaged in a box that is ridiculously huge for its contents?
" Not an easy thing to do "
Plausibly impossible to define, yet.
" and the rules would have to be worked out so as to be unambiguous on the meaning of bug-free."
Define unambiguous, and be precise as possible.
" Let me suggest that OSNews get together with the X-Prize commitee (or DARPA, or whoM, whom whom whom whom!!
(fist "whom" fix is free)
ever) and convince them of the potential bebefits of such a prize. "
Yes, like essentially re-inventing the wheel. You have me convinced, now get on the horn with DARPA, you sexy genius you! Oh, and I'm letting "bebefits" slide because I think you just sneezed while typing and it's VERY fun to say.
" I'll be frank (shameless plug), "
You're shamelessly plugging Frank? I sure hope he works at DARPA.
" I am proposing this contest because I believe I could put together a team to win it. It has to do with my faith "
My favorite quote on faith is: "Now faith is being sure of what we hope for and certain of what we do not see."
Hebrews 11:1 (NIV), and buddy, you have it in spades. You honestly believe in a comprehensive computational system that is bug-free. Of course, you also seem to believe that for the world to deserve this system, they must first give you a poke at a few million dollars, as opposed to the lovely folks at itty-bittyOS who do it for free.
" in non-algorithmic, signal-based, synchronous software. "
Odd that you mention this. I have one of those lying around here somewhere. Methinks it's called "Leisure-Suit Larry"
" I don't believe that unreliability has to be an essential charateristic of complex software systems. "
Define reliability. Define essential. Define complex. Oh, yeah, this contest is off to a boiling start already. Tell you what, just MAKE the thing and you'll be a guarateed $millions-aire by my sister's third trimester. Then you can tool about in your Murcielago and laugh deep, throaty laughs at the peons still struggling along with their algorithmic, something-else-based, asynchronous software.
" Louis Savain "
And remember, folks. It's not software, it's Savain-ware!
(Hey, if you get rich and use the slogan, I'd like $million for thinking it up.)
RE[3]: Great idea but does not go far enough, INDEED!
RE[4]: Great idea but does not go far enough, INDEED!
Thus far, all you've got to your credibility is a fancy website with all your theories, but no real useful demonstration to show that your theories are actually practical.
Once you've got an OS using COSA and all your other theories that allows a user to run a COSA program (which you must also provide) that performs a perfect implementation of Windows Note Pad (a simple program to reproduce as proof-of-practicality!) then you will have some amount of credibility. After all, Note Pad is incredibly easy to implement (I'll even relax the details of having to handle different fonts with proportional spacing: only monospaced fonts required) and shouldn't take more than a day to implement in any programming language for someone that's not incompetent. Also, if your COSA stuff is so easy to work with, it should take even less time to decipher how to have everything react and work correctly.
So, I'm issuing you a challenge right here and right now to produce an COSA-based OS that boots and runs the Windows Note Pad clone and some other process at the same time, within the time constraints of this contest. The benefit is I'll mod you up everywhere regardless of how off-topic you are (like you have been in this thread, along with almost all others, spamming your concepts) if you accomplish it within that timeframe. The down side is I will mod you down everywhere you post with your spam, and encourage everyone else to do so, too, if you don't accomplish this within the timeframe of this contest.
After all, if the concept is so brilliant and practical, you'll easily have an OS and a Windows Note Pad clone within that timeframe that is completely error-free, and then you can write about it and submit the article to OS News to be published, because then you'll have something that people can actually *try* with their machines.
Frankly, I strongly suspect you can't accomplish this, because of all I've read on your site, and how little you seem to know about what else in the outside world resembles what you preach, and yet you fail to give any notice about that outside world. I'll give you a hint: what you're proposing is practical to do is equivalent to writing a word processor with the same methodology used to control your typical city's traffic light system.
It sure is easy to make claims when you know that you'll never have to substantiate them, because your preconditions will never be met. The more pressing question is, why do you require the attention of others so much? If you can't be famous, why not be infamous? Is that it?
I'll be frank (shameless plug), I am proposing this contest because I believe I could put together a team to win it.
A convincing testimonial of faith in ones own abilities. Until, that is, one reads your website and encounters some of the other things you believe - E.g., http://www.rebelscience.org/Seven/bible.html . I was especially amused by your claim that scientists have an "anti-religion prejudice" because they discard data which cannot be measured or observed. If anything, scientists have a prejudice against unverified and unverifiable assumptions; the funny part is that the claim clearly demonstrates your bias.
We need more stuff like this! It's boring to read about Microsoft, Linux and Apple everyday
By the way, I strongly encourage people to be a little technical in their review. It's cool to know how networking works, what kind of memory management is used, filesystems, etc.
Oh! I hope all the reviews will be posted on OSNEWS. Not just the chosen ones.
Great idea guys!
But as an OSNews poster already, what exactly is the value of free OSNews subscriptions? Is there a lot more content that people might pay for normally that I don't see as someone with this type of account?
Otherwise, hey, why not do such a contest? Surely it can't result in horrible things...
I wasn't even aware there was a subscription! I just looked and it's hidden down the side menu labelled "Membership" completley undistinctive.
Although there's not much value to subscription, it is for good karma. People should support the site if they use it regularly, I know that this is my second most read site next to Digg.
I'll definitely be looking into getting a subscription. Is there any distinction shown in the comments for subscribers? And will it fix the problem of me having had 0 votes for the last year despite being modded up often and e-mailing you regarding it?
Would it be OK if I wrote an article about how I came to write my own OS? It would be about why I did it, how I started, the pitfalls I encountered, the lessons I learned, where I got the know-how, and of course, how I actually did it. The OS is called Cefarix by the way, and the latest version is on SourceForge.
Edited 2006-06-16 20:34
It's still quite rudimentary. No graphics, networking, or user-land applications support yet. Just a minimal (and buggy) command line interface. It does have a namespace system, file system, HD support, multitasking, and the other essentials of a modern OS. Is it still acceptable? Or does this require an OS that is more usable?
since you state the source code must be available
I didn't state that, and if I did, I didn't mean it. This is what the article already says:
The work should cover an existing desktop OS available to the public for download or purchase
If just the binary is available, that's ok too. The point is, it shouldn't be some top secret project you've been working on in your basement but haven't shared with anyone yet.
and PowerOS
http://www.poweros.de/
Almost forgot:
http://www.vitanuova.com/inferno/index.html
http://www.eros-os.org/
http://www.cl.cam.ac.uk/Research/SRG/netos/old-projects/nemesis/ (gone?)
http://www.oberon.ethz.ch/
http://www.rebol.com/
http://www.opencroquet.org/
http://www.theos-software.com/ (gone?)
http://www.freedos.org/
http://www.rocklyte.com/athene/
I would love a review of this beast:
http://www.1060.org/
(or you wouldn't call it an OS?)
http://sourceforge.net/projects/glider-kernel
Prizes you say? I'll send in a review of Glider's design and current progress.
Nothing indicates that the news stream is dying out. However - quite a few things indicates that people are interested in operating systems, including the alternative OS'es. Alternative OS'es do not get the same amount of media coverage as mainstream OS'es, so that's probably why there is such a contest.
I agree with the previous statement. When I first was turned on to OSnews, there were a lot of articles posted for the smaller, more obscure, and hobby OSs, which made this site very interesting and unique. Lately, though, most of what we are seeing are posts about Vista, OSX, Linux and BSD. I would very much like to see a return back to what we used to see regarding other OSs. While the mainstream OSs have a place in the reporting, much of that info can also be gleaned from a myriad of other well known sources. So, if a contest like this is what is needed to return OSnews back to its "roots", then this is good.
I don't think Ubuntu qualifies as an "alternative OS", no matter the platform. Of course I could be wrong, but as I see it, Ubuntu is Ubuntu - no matter the kernel.
Think "small hobby OS'es - really small hobby OS'es", and combine that thought with "small commercial OS'es - really small commercial OS'es" - and I think you'll get quite close to an understanding of "alternative OS'es".
i think Nexenta rocks. Really, I do. However, when all is said and done, it's Solaris 11 build 40. And there's tons of iformation about Solaris internals. What we are trying to do is educate people about OSes that aren't already covered everywhere. So, unfortuntely, no, I don't think Nexenta qualifies.
A couple of weeks ago, I wrote a review of Rockbox OS, an open source operation system for Archos, Ipod and iRiver jukeboxes. I mailed to Thom, but he forgot about it (I nagged him as well). It's quite a pity really, I put in a lot of time in that review (it's over 2000 words long and don't forget all the time needed to mess around with the OS itself).
Hopefully, Adam will be a bit more respectful of my 'creation' than Thom. Though I don't think my review will qualify as it mostly focuses on the user experience as opposed to the internal of Rockbox. I didn't even mention what language the devs used to write the OS.
"Prizes!
We will be awarding up to 5 two years subscriptions to OSNews and up to 10 one year subscriptions to OSNews as prizes. In addition, we have other prizes, such as an mp3 player, books, and more. Details and quantity of prizes will depend on the number of qualified submissions.
Why is there a prize for a subsciption to OSNews? Isn't site registration free?









