Gates: Hardware, Not Software, Will Be Free 993
orthogonal writes "That's small-'f', not capital-'F' free:
according to Bill Gates, "Ten years out, in terms of actual hardware costs you can almost think of hardware as being free -- I'm not saying it will be absolutely free --...." Gates expects this almost free hardware to support two of the longest awaited breakthroughs in computing: real speech and handwriting recognition. He further predicts -- ugh! -- that software will not be written but visually designed."
I hope not (Score:3, Interesting)
Ugh? (Score:3, Interesting)
It's no different than using scripting languages, really; it'll have its own set of trade-offs.
visually designed software (Score:3, Interesting)
The reasons were because its is easier to CVS/grep/replace...
Interesting.. (Score:2, Interesting)
'Pentium 4' Processor,so it MUST be good >_> Bleh,if hardwares free,how will advancements ever be made in it -_- You can't make new hardware without money,and the good hardware will most likely rise in price...really not helping the poor geeks out here
Actually ... (Score:3, Interesting)
Microsoft DOES sell X-Box AND Human Interface Devices. They're certainly not giving THOSE away. Though if Microsoft could get enough royalties of games, I could see them giving X-Box away.
In the future, my desktop will cost $20 and my Intellimouse will cost $200. Go figure
He's right, but it doesn't mean it's a good thing (Score:2, Interesting)
And, yes, there are plenty of languages where you program visually. But when you want to change something? Ugh. Instead of being able to insert a line of code, you have to move the next 5000 symbolic representations manually. Heaven forbid you want to add more. And then you have to fix all your arrows so you can tell what's going on again. Worse yet if there was some novice to ever touch it, then it looks like someone used silly string in your IDE.
NOT a good thing for large projects, easy as it is to think abstractly with that tool. Shorter learning curve, though, for small scripts and mini-apps.
Re:Visual design (Score:5, Interesting)
First person shooter.
This reminds me of a cool hack that uses Doom as a "process manager" [unm.edu]. Killing a Doom baddie basically "kill -9"s the process.
Re:Please Bill.. (Score:4, Interesting)
Yeah, I also thought this.
But before the Linux-era, Billy was actually correct: At DOS-times, computers cost about 5000$, while DOS itself was less than 100$ (full version) IIRC. Today computers typically cost less than 1000$ but Windows XP (full version, crippled) costs 200$ or (full version, uncrippled) 300$.
On Windows-servers the ratio of the total system price which is going to Microsoft is even higher.
Also, Microsoft is doing much more against piracy these days (WPA, BSA-audits, etc.) than 20 years ago, which de-facto translates into yet another price increase.
Even though Bill Gates seems to have the delusion that this can go on like nothing happened, he is wrong: On servers, Microsoft already feels the heat from Linux and the desktop domination already shows some cracks.
Re:Visual design (Score:2, Interesting)
A nice animated penguin jumping all around and giving stupid advices!
My former SIG was "WYSIWYG, but what you see may not be what you want"
Visual tools are nice and helpful but the plain text is still by far the most 5, informative data carrier. Especially highlighted text...
MS Labview? (Score:3, Interesting)
I've used labview just for writing programs to link to IEEE hardware, and it certainly is much easier to
deak with a large number of modules when they're visually represented, and very fast to kludge together a fast fix.
The only thing is that the debugging and maintainence is a nightmare because unlike a normal C/Fortran, not all of the program is visible at once (it's in a thousand tiny blocks), and so looking at several related bits of code is very time consuming. So much so that we recently rewrote some labview code in c, just to improve the clarity and maintainability.
Visual designing is a reality now! (Score:4, Interesting)
Designer will create the proper UML diagram to represent the structure and some dynamic aspect of the application in a platform independent model. Then we apply some code generation templates to generate the code for a targetted architecture.
If we go a little further with the code generation, we could actually implements most of the business logic structure based on sequence diagrams.
For the front end, while it would be hard to generated a really nice interface, we can generated what need to by put in a screen. Then it's a matter of applying a CSS or using a visual editor to reposition the component in the screen.
I can see that in 10 years, most of the business code will be written that way... Note that one of the premisse of this happening is that proper analysis and design must be done. For that we must change the mindset of a lot of people.
As for people fearing for their current developer status... These people will have to grow up and start doing real developement instead of using the use the force approach. And for really good developers/architect, there will always be a need for someone to define an architecture and create/maintain the templates required to translate the visual design into real code. And there will also be a need for good developer to write code to implements the complex algorithms required by some applications.
Anyway, writting most business related code is boring and repetitive, so why not generate it!
Re:Please Bill.. (Score:2, Interesting)
Software has a cost.
I never said it didn't. (I"ve worked writing software for many years) However when Gates is saying that hardware will be virtually free while his software costs hundreds (or thousands) of dollars, money that will be added to MS' existing billions, his arguments come across as completely disingenuous.
Re:Please Bill.. (Score:2, Interesting)
Thats what happens in a free market with competition.
software has gone up
I wonder why? One would almost think there was a monopoly market at work creating an artifical scarcity of software!
Re:Please Bill.. (Score:4, Interesting)
Handwriting? (Score:2, Interesting)
Why hardware won't become free, or even close (Score:5, Interesting)
- Chip makers will contine to create advancements and will want their R&D dollars back, just like Mr. Gates. This is why software is expensive; it is cheap to to burn a CD but time consuming to develop.
- Two words
.. advertising costs.
- Chip makers delay the release of new chip sets if they have significant inventory of other models. This keeps the prices of current chips artifically high until the manufacturers feel they can't milk any more out of consumers. Chip makers will be sure to not release new products until demand is there and they recover R&D costs for older chips.
- CPU and memory chips account for less than half the cost of a PC; disk drives, monitors, DVD/CD drives, cases and motherboards make up the rest. These items have too many mechanical/structural parts to realize significant savings from improved chip manufacturing techniques. Even if the memory and CPU were free, systems will still cost a few hundred dollars.
- Some people will always want/need advanced features, and computer systems and chip makers will always charge a premium for those items.
- Chips contain software (on-board video, BIOS,etc.). I doubt if the makers of those software components will start giving it away. But, if open-source alternatives became available, those items would realize additional savings. I would not be surprised if more software wound it's way into hardware as the cost of updating firmware becomes cheaper. Hardware video drives can be a lot more effective than OS video drivers.
Until chip manufacturers stop releasing new products every few months (reduces R&D), stop advertising, and create an entire system on a chip, including structerual components, external interfaces (wireless??), storage, and displays, computer systems will never be 'almost free'.Re:Visual design (Score:5, Interesting)
Just try to visualize (pun intended) a fairly simple event driven program with lines connecting all events, triggers, functions, data and UI components and you get the idea.
Re:Ugh? (Score:3, Interesting)
Let me explain... Firstly, I should say that visual techniques have enormous potential and should not be ignored. UML is responsible for a snails-pace revolution in object oriented design and flow charts and schematics (a historic mainstay of computing) are becoming ever more advanced with automated techniques and tools. That said, there is still no way to arrive at a program without writing it. The best one can hope is to find a more appropriate syntax in which to write programs. If you have an effective visual language then there must exist one-to-one mappings from programs in that language to similar programs in any sufficiently rich textual programming language. Furthermore, gurus of theoretical computer science will be quick to tell you that it is possible to re-write any program written in one computationally complete language in another. (In case you wondered - practically useful languages are almost always computationally complete!)
The real challenge of visual languages is to effect a notation which is more convenient to that offered by a conventional textual form. With the exception of a few specific circumstances (e.g. WSYWIG word processing in place of programmatic typesetting; visual form design and video sequencing - for example) every visual language I've seen hyped for 20 odd years has been vapourware. Thinking back to the early 80s there has always been some well meaning salesman or other telling us that generic visual programming is just around the corner... yet I am still to see a single convincing example where, for example, a classic algorithm can be more easily or more clearly accurately specified in a graphical format than in a conventional textual language. I won't say this will never happen - I just retain a strong sense that I'll believe it only if I see it. I seriously doubt they will make programming any less demanding a task.
Re:I hope not (Score:4, Interesting)
I don't think this was ever true.
Computer Science on the other hand, is a mathematical discipline which involves working out how to do things better, faster, and with less energy. It's about algorithm design, and ways in which to make a computer most efficiently process mathematical representations.
Certainly, certainly, but how is this different from programming? Programmers work out how to do things better, faster, and with less energy. Programmers design algorithms. Programmers design ways to make a computer most efficiently process mathematical representations. And not just mathematical representations, either. All kinds of representations, in fact.
I won't dispute your central point. I think it's vital to make a distinction between hard programming and soft programming. But the gap between the theory and practice just isn't as clear cut with computer science as with other disciplines. There is a big difference between designing an engine and building one. The difference is much less pronounced in software, because at some point the design or description becomes a program in its own right.
Re:Precisely - we can't even get WYSIWYG HTML righ (Score:5, Interesting)
Then CSS/Layers became totally (mostly) supported. Now WYSIWYG editors work QUITE well... (Even some non editors generate perfect code. Photoshop's image ready generates some very nice code)
Anyway, point being, when something is designed to be designed visually it can be visually designed much easier. *grin*
Visual Programming != GUI (Score:3, Interesting)
Re:Visual design (Score:3, Interesting)
The only difference is that Bill forsees getting rid of the intermediate step of writing code to represent the visual system.
In Univ, we did a couple of exercises. The task was to write a simple SIP client (just session establishment, nothing transferred) and server. We did it first in C using some standard libraries. Then we did the same thing with Telelogic's [telelogic.com] SDL Suite [telelogic.com]. We basically drew the state machines in a flowchart (only the application layer). We then hit "generate" and it created a bunch of C code that went through gcc.
With the SDL, I could practically convert RFC to a working protocol stack in a few hours. Of course, there was no transport layer or anything - I guess they supply a set of standard protocols like TCP/IP-stack, but we never got around to check it out. The application-layer endpoints were directly connected.
Oh, and I don't consider myself a coder. I know C++ and can write some shell scripts. I basically want the computer to DO some things, and not spend time telling it how to do it. Back in the 80's on my C-64 I told the computer "10 PRINT "Hello!":GOTO 10...we have not gotten too far from those days yet.
If someone invents a programming language that includes a way to tell computer "do what I meant it to do and stop complaining about irrelevant crap", I might consider programming as a way to make living
Re:Free (Score:5, Interesting)
The price of the average "IBM" PC sold has dropped by roughly 400% since I first bought one in 1989. At the same time, processor speed on these average machines has increased by 50,000%. If this trend continues, and I see no reason for it not to, the average computer in 15 years will have a 10 THz processor and cost $125.
Now, the while the cost of hardware continues to go down, the cost of software continues to go up. The number of people who are needed to build the massive applications to make use of 10 THz will be huge. Somebody's got to pay the damn programmers, right? So the price of software will continue to go up. Even if OSS succeeds and the operating system and incidental programs are free, the CUSTOM programs will be expensive.
Therefore, it makes sense to give the hardware as an added bonus with the software. The same way you have cell phones given away with calling plans today. This isn't a Microsoft thing...this could easily be an IBM thing or an Adobe thing, etc.
Re:Enough (Score:2, Interesting)
Same was with laptops. Right now I'd rather have a laptop for my home computing than a desktop. It's powerful enough, the components are affordable. It has a nice screen. It's QUIET!!!! And the keyboard is more comfortable for touch typing.
Every time I turn on my desktop I shudder. There HAS to be an alternative. And a laptop is really it. But the laptop also has faults. It's big, it's heavy (relative to other stuff I like carrying around) and I need to sit down to use it.
The wearcomp is the future. Some people out there are working on these babies right now, and let me tell you. It is going to be hot! (I'm not talking about Xybernauts. Check out Steve Mann's work on eyetap.org [eyetap.org]. The man is perhaps not as clear as some of the marketing pros out there, but you only need to look at his work to see that he has the right idea. He's on the fringe now, because computers aren't fast enough for his algorithms yet. But when they are, you will see the real paradigm shift (dammit that's such an ugly term, but it applies here) in personal computing. Until then, you will just see a gradual merging of the cell phone with the PC(tablet/laptop whatever).
Yes, I have worked in the Mann's lab in my last year of Skule. And even though I did not really contribute to his work (hey, I don't have the gift), I saw the future in his lab. Working there, you could see the science fiction coming true. I was in constant awe and I had adrenaline in my veins when we unleashed some of the algorithms on the poor raw data we collected. I am not, nor ever will be a computer genius, but I envy the folks who work with Steve every day, because that lab is what passion for computing is all about.
[blink,blink] what am I doing here?
I should go now...
Re:Visual design (Score:5, Interesting)
Software is written because software is a set of instructions. Software is a set of scripts that respond to events. If software were spatial and totally right-brain (and analogous to engineering or construction), AI would work, and software would probably rely on the immutable laws of physics and chemistry, rather than homespun rules. When I write software, it is frequently because I am taking a "break" from other totally creative pursuits.
The only visual constructions relating to software engineering (SE) that I consider appropriate, are those that relate a large system in terms of its data, logic, and interfaces. This is not necessarily the Rational Unified Process with UML -- indeed, I tend to think people take that too far (eXtreme Programming seems to take a nice perspective on SE in this regard). People also like to relate Classes to real-world objects, usually real-world objects that relate to "parts" of the project. This is tempting but is, I feel, usually inappropriate! A good compromise is a balance between the format of the data (with appropriate, thin, "agnostic bridges"/Classes) and an easy access point for real logic (the Model, of the MVC pattern). I would also recommend a sort of laid-back attitude when developing software: don't live your life by a paradigm or methodology, especially in an immature field (SE) that has a lifetime of problems to solve. You know what problems need to be solved. You also know that not once did you wish you could draw a picture instead of write code. I mean, what the hell? Someone take Johnny Mnemonic away from Gates.
If the software you write, however, is modular enough that you can arrange the pieces/modules/methods like components in a circuit, then go for it. However, this level of widespread code reuse is frankly fantasy; reuse will remain, I believe, as it has: generic libraries used in a custom fashion, i.e., not suitable to be "visually" "dropped-in." Code generation is nice, but it's only appropriate for certain large-scale applications (like large database-driven applications).
If one is to believe Gates on this issue, one is also compelled to believe that Microsoft's research and development department has created software practices at the forefront of software engineering (and indeed computer science. Remember computer science?). I do not believe this to be the case, and I'd make the indictment that this "release" by Gates is purely worldfair in nature, and is for the hoi polloi.
Perpetual Gift Computer System (Hardware GPL) (Score:2, Interesting)
I just started a discussion on the DebianWiki about the legal and technical implementation of making a computer system into a perpetual gift. That would be a more personal and specific way of making hardware free for the user, although not for the first owner of the system.
IANAL so I need help on wording the legal contract. For you lawyers, paralegals and armchair philosophers out there; If you feel like doing some constructive legal work for the Debian project I welcome your advice. I think this idea has a lot of potential not only for a gift between friends but as a way of donating computer systems to charity and ensuring that they will remain gifts after they are no longer useful to the recipient organization.
This brings to mind the vision of an admin staying up late and GPG signing the contracts for a one kilobox donation. *shudder*
PerpetualGift [debian.net]
Re:Too Cheap To Meter (Score:3, Interesting)
It actually would be slightly more expensive than my current usage. I considered signing up for it and offering to run computation farms for my research lab, but I suspect the fine print must have something about not exceeding my average usage over some number of months. I didn't really read the fine print because I assume its in there.
Re:Visual design (Score:3, Interesting)
Or does.....
He further predicts -- ugh! -- that software will not be written but visually designed.
mean....
We have found a small startup that has geated visual design software and are, at this moment, lining up a licencing contract that will allow us to build it into Longhorn before enhancing it by adding our own proprietry extensions and destroying said company with a 5 year long bank draining/bankrupting law suit using our lawyers of mass destruction.
Re:Ah, visual design in VB (Score:3, Interesting)
So, I spent a couple hours and got something working -- in fact it was kind of cool because the ball would spawn new ones as time went on, getting progressively harder and faster, which was cool.
I'd written it in C, on my math-coprocessor enabled 486, and it ran just fine. I was amused, so I gave it to a friend, who had this *amazing* tiny HP laptop ( circa, I believe, 1995 or so ) with a nothing 286 or so processor ( on the othe rhand, it was less than a pound and ran off little batteries ). The program, needless to say, barely ran.
Now, I was just a kid ( teen ), and I didn't really understand, since I was still pretty new to code optimization. I'd had *good* highschool classes that taught me six ways to sunday how to optimize an algorithm, but nothing about actual hardware stuff. What I hadn't realized at the time was my code was using floating point math to position the "balls" and that required at the time a math coprocessor ( I guess ). Even though my blitting code was all fast integerial VGA framebuffer kind of stuff, the 20 lines of positioning code was enough to bring the HP to its knees since it had to emulate floating point math.
This was one of my first lessons in writing fast code -- I rewrote it using long integers to do bit shifted floating point arithmetic. Suffice to say, it hauled ass on the HP when I was done.
The sad thing is that this all worked just fine on minimal hardware back in the 70's. I learned great respect for low level programmers, then.
I still believe in algorithmic optimization above all, but now and then, when I profile my code ( Apple's Shark is your friend ) I'll find some boneheaded ( I blame only myself ) use of stl's array index operator that's eating up 75% of cpu cycles in some inner loop.
Re:Visual design (Score:2, Interesting)
in most operating systems it's usually suggested to just install a fresh copy, it will reduce errors later on down the road
Try visual programming in Java right now! (Score:2, Interesting)
http://tomatoide.sourceforge.net/ [sourceforge.net]
It works. Almost. I've largely lost interest in it. The reasons why it isn't a practical way to program are legion, and BlueJ [bluej.org] is probably a better implementation of the same idea, anyway.
Best feature: You can modify a running program while it's still running.
Re:Yeah, right (not with bloatware) (Score:1, Interesting)
The big LCD screens eat as much power as anything and spinning the harddrive/cd/dvd isn't free either. (I won't even mention the trend of powerful GPUs in notebooks. Oops too late!) Popping a 0.1W processor in there isn't going to get you a PDA-like lifetime out of a notebook. The Pentium M and Mobile Athlon already cut back on the power consumption drastically when asked too.
In the end, I think most people find it's easier to find a plug or carry a spare battery.
This *is* evil (Score:3, Interesting)
Microsoft can only make money, with free hardware, if the box is locked down to prevent it running software that isn't authorized by Microsoft. It's not hard to see that this is precisely MS's strategy. The Xbox is just the tip of the iceberg. Fortunately MS are not too hot with security and for at least a few years we can expect PCs to still be hackable even when supposedly locked down. But eventually MS will learn, as they always do. And then we'll all be screwed.
Re:Free (Score:2, Interesting)
If visual design tools and other automation improvment can be realized, development costs in human hours, and better reusability, will see software reliability improve dramatically while costs drop substantially. In which case, software should become really cheap, just like hardware. And the raw materials of software are technically free (or at any rate, not in limited supply) unless you count patent fees and the costs of managing the complexity of choosing the right tool for the job, which still requires human brain power.
The improvement of tools is absolutely and utterly the main roadblock to a mind-boggling improvement in software quality and functionality.
So, if Bill gets what he wants, he won't get what he expects.
Re:Precisely - we can't even get WYSIWYG HTML righ (Score:3, Interesting)
Does the Lego Mindstorms "program editor" count?
* Take bright yellow "for...next" piece, and place on page
* Attach it to end of program
* Attach bright red "toggle variable x" pievce, and attach it to the middle of the for loop
I can only assume the lego people have 8000x4000 pixel screens or something...
Re:Yeah, right (Score:3, Interesting)
It's not humanoid, and it only does one thing, of course, but robots are here.
Re:Yeah, right (not with bloatware) (Score:3, Interesting)
I would say that right now we have officially reached the point where 2 year old hardware is more than adequate to do everything excluding gaming/video encoding etc.
My computer is 3 years old and still works fast enough for everything...I dont think the newest version of windows has ever ran on such old technology. And since hardware speed doubles every about year...this trend will be exponential. So in 10 years, the newest OS will run on maybe 8 or 9 year old computers is my best guess...
Although predicting the future is like russian roulette...if you win its fun, but most people end up losers
Re:Visual design (Score:2, Interesting)
Re:Visual design (Score:3, Interesting)
This reminds me of a cool hack that uses Doom as a "process manager". Killing a Doom baddie basically "kill -9"s the process.
Cool idea. Have you read Disclosure by Michael Crichton? There's one imaginative sequence about a VR-based file explorer and UI. The concept is that you walk around hallways and chambers, opening drawers and reading files (which float in midair). Other logged-in users are visible too, in the appropriate locations (directories/files/work areas). If they are using a VR interface, they look lifelike. If they use a conventional CLI or GUI they look like cartoons on stick figures (apparently the system pulls an image of their face from an employee database and shows it on top of a stick figure). Was a pretty imaginative concept. I don't know if anyone has implemented anything similar yet.
Re:Cost of hardware =0? (Score:5, Interesting)
It's not that insane, and remember that Gates is a billionaire in part because his company has been abusing its monopoly. To an extent, if he wants something to happen, he can make it happen. What he means by "free" is that users will "subscribe to" software and in doing so, receive a machine on which to run that software, effectively for nothing. This is what Microsoft wanted to accomplish by bullying retailers not to bundle other operating systems. My guess is that they will attempt to use "Trusted Computing" (or some technology just like it) to make their intention into a reality; if you want to run Microsoft's software you will have to run it on computers which only run Microsoft's software or software written by Microsoft's partners (in other words, companies which have bought the right to have their software run on MS's hardware). So they can make the cost of hardware approach zero, so long as they can be sure the hardware is only usable for purposes for which they can make some money. Of course, all of this depend on governments around the world letting MS get away from it.
Re:Why visual design will help, but is no panacea (Score:1, Interesting)
Generally, yes this is true. However, simple components can be PROVEN to be correct. You can tell exactly what all the possible states of a proven component are. More importantly, you can tell what states are NOT possible with a proven component. If you build up from proven components, the number of states is drastically reduced. Now all we need is for somebody to start mathematically proving some libraries...
Re:Precisely - we can't even get WYSIWYG HTML righ (Score:3, Interesting)
You know...I've tried and tried to use the 'visual' SQL generators...and but for all the simplest models...I cannot work it right. I can much more easily and quickly write SQL by hand. Dunno, I'm usually very visually oriented in many cases, but, for complex queries...I can do it with a text editor better than I can the ones like you described...
Re:Visual design (Score:3, Interesting)
Film at 11 (Score:3, Interesting)
Of course bill would say these things (Score:3, Interesting)
Competition (Score:3, Interesting)
Conversely, with Microsoft's OS and office software monopoly firmly in place, prices have been going up and innovation has been stagnant. Can you point out any feature added to Word since 1997 that you actually need?
If you assume, as Gates obviously does, that Microsoft's monopoly will still be around in ten years, then his prediction that software will not be commoditized is correct. On the other hand, if OSS breaks Microsoft's monopoly on the desktop, mature software will be free (as in beer) and service will cost money.
Healthy would be if Microsoft were to be relegated to having to actually innovate to earn money while markets that have been around for a while open up to competition and get commoditized. If a software component is so mature that a handful of college students can replicate the functionality in their spare time, professional software makers should have to move on.
We see a little bit of that in the server market where Microsoft is having trouble leveraging its monopoly in order to kill the open source competition. Poor reliability and lack of embraced and extended standards that create lock-in have successfully thwarted all attempts by Microsoft to corner this market. Result: Choice, higher quality and lower prices.
Hopefully, Novell will be able to aggregate and focus the community's effort to dislodge Microsoft from the desktop monopoly sooner rather than later. Also hopefully, the increased visibility of Linux by way of the laughable SCO lawsuit and recent endorsements by HP, IBM and other fortune 500 giants will enable Linux to gain critical mass in this market too.