Moore's Law Is Microsoft's Latest Enemy 395
Glyn Moody writes "Until now, the received wisdom has been that GNU/Linux will never take off with general users because it's too complicated. One of the achievements of the popular new Asus Eee PC is that it has come up with a tab-based front end that hides the complexity. But maybe its real significance is that it has pushed down the price to the point where the extra cost of using Microsoft Windows over free software is so significant that ordinary users notice. As Moore's Law drives flash memory prices even lower, can ultraportables running Microsoft Windows compete?"
And advertising/capitalism is Linux's enemy (Score:5, Interesting)
Re:Moore's law has nothing to do with price (Score:4, Interesting)
Flexibility Not Price (Score:4, Interesting)
The inability (well, ok, extreme difficulty in) to skin/specialize the user interface is going to hurt them. Microsoft appears to be mentally permanently stuck in one-size-fits-all land. And to be fair, it would be really hard to let people customize as deeply as they need to without letting them muck with the deep details of your OS.
If need be, they'll give windows away (Score:5, Interesting)
Microsoft can make money on windows without charging for it; they can charge $15/copy for the minicomputer version. Microsoft has an endless number of strategies, which they will employ to keep market dominance for as long as they can.
There will be a whole *series* of retrenchments. Microsoft is in a very powerful, very profitable place, so they will fight each retrenchment as hard as they can - but they're not stupid, they've got contingency plans to stay in the market and, frankly, to stay extremely profitable whatever happens. Put another way: they can compete with free, maybe not on a level playing field, but on the playing field that exists, and they intend to do so.
Forcing them to compete, even on a biased field, is good for the rest of us, so I'm all for it. But driving MS out of any market segment is going to be extremely difficult.
Re:Eee PC vs. REAL UMPCs (Score:5, Interesting)
Everex has now come out with the Cloudbook (Linux) at WalMart so, now it is being exposed to the masses. The revolution is starting!
MS strikes back (Score:5, Interesting)
http://www.blogeee.net/2008/03/06/le-eeepc-900-uniquement-avec-windows-xp-dapres-asus-france/ [blogeee.net]
The good news is that the French customer is very well protected and forcing a software with a PC down their throat is illegal. So essentially, what will happen is thousands of geeks demanding reimbursement of the XP licenses. That oughta hit Asus really hard, and teach them a good lesson.
I read that Asus Germany announced a similar "forced sale", but can't seem to find the article.
Clear for a long time (Score:5, Interesting)
If a P3 500Mhz system was coded with the efficiency and elegance that prevailed on the Commodore 64, your OS and every application running would be so blazingly fast as to seem instantaneous, and with 1GB RAM you would not require a harddrive for anything except storing large image/music/video files. Instead, my early-generation P4 2ghz machine at work with 2GB of RAM chugs and sputters and stutters along and I can't wait to get home and use my 'powerful' personal machine that operates much faster. It's absolutely ridiculous.
Why don't you actually read the Wikipedia article? (Score:5, Interesting)
Also, if you'd bothered to look at the article, you'd find that the quote provides a citation, and that citation points to a PDF file of the article in which Moore made the statement in question:
ftp://download.intel.com/museum/Moores_Law/Articles-Press_Releases/Gordon_Moore_1965_Article.pdf [intel.com]
In short, you lose on both style and substance.
Windows XP will soon go out of print (Score:5, Interesting)
received "wisdom" is wrong (Score:3, Interesting)
I think you meant "perceived" wisdom. But in fact, I've installed Linux on several friend's PCs who had never used a computer before (Mandriva 8 IIRC). None of them have had any trouble whatever using it. In fact, I get fewer "how do I" phone calls from them with Linux/KDE than I did when their new machines were running Windows.
Gnu/Linux/KDE (and most likely Gnome as well, although since I haven't used it I can't say) is easier to use than Windows for a variety of reasons, the first being that stuff is put in logical places (at least with Suse and Mandriva) as opposed to Microsoft's way of putting stuff any old place. At least that's what it seems like; I can't see the logic of where Windows' stuff goes at all.
So please stop spreading this this FUD. It's simply not true. Windows is NOT easier to use than Linux.
This is 100% true. (Score:4, Interesting)
Maybe in ten years that won't be true. After all, I didn't really expect Word to overtake WordPerfect and other alternatives in the market the way it did back in the 90's... but even in that case, it's because something has happened to Office, not because of Moore's Law.
Re:Yes? Is this a question? (Score:5, Interesting)
Simplicity does not mean usability (Score:5, Interesting)
The main problem Linux faces is not that it's too complex, but that it's designed with a philosophy that tends to value "technologically correct" above all else. There are times when being less precise, less technically oriented, less detailed or less optioned is better for the human user, even if it is not as "true" to the computer itself. Apple seems to explicitly understand this, Microsoft seems to sort of intuit this without understanding it (so they don't make the right choices, but they realize such choices need to be made, which is better than nothing), while on Linux, this seems to be poorly understand, and often seen as a negative.
With most cases of usability efforts on Linux, it's often just trying to copy (and improve upon) some existing system (GIMP vs Photoshop, KDE vs Windows, GNOME vs Mac OS (classic), etc.), it's an attempt to be more usable for admin-types (dselect, aptitude, etc.), or--and this is where Linux truly falls flat on its face--when someone attempts to make a truly usable Linux, they don't think, "let's make a Linux that works the way people work," they think, "let's make an interface that is so simple, even an idiot can use it." Instead of respecting the humanity of their target audience, they insult them.
That is a problem Moore's Law can't do anything about.
Linux won't truly take off until they stop insulting the normal person, and start respecting them. Ubuntu is close, but it's still too technically-oriented. The thing is, though, I'm not sure this is a bad thing. It might be, as it does keep Linux from being a mainstream OS, but on the other hand, it *is* an excellent OS for the people who are more technically-minded, and prefer absolute control, who value technology over aesthetics and the humanity of the interface. If Linux truly evolved to become a user-oriented OS, it would leave a void for the technical user. I suppose there'd still be the DIY Linux distros, plus there's always BSD or Plan 9, or some new OS yet to be created. Still, I'm not sure that if a User-Oriented Linux became a major OS player, that the more bare-bones technically-oriented Linuxes wouldn't find themselves losing significant attention by both users and developers alike.
Re:Slashdot (Score:3, Interesting)
I use to be be a diehard Free Software avocate... Over time Free software has disapointed me, with dealling with with people who have many different motives Bragging rights, Software Purity, Freedome, etc... vs. dealing with a company who has one motive... To make money.
In a hundred years from now do you think any software stands a chance in hell? Computers and software as you know it today will be so antiquated You Newest and Fanciest Computer Today in a hundred years will be like those Machanical Adding Machienes 100 years ago. If not true AI at least good enough to get and alter data when you need it will probably just a thought away, stored in a spec imbedded in your earwax.
All this fuss about priority and open source software may be just a bit footnote in some history book
Not a revolution (Score:4, Interesting)
I am running a 289 dollar "piece of crap" desktop. I have been 4 four years. It plays WoW and does general work just fine... stupid computer, I promised i wouldn't by another one until it broke. I gave it a year.grrr.
Maybe I should install Vista, that would break it.
Re:Slashdot (Score:3, Interesting)
Sure. Why would you think proprietary software would "go away"?
It just is not sustainable to have every business, school, and government paying 1 provider of software for an operating system.
Agreed. But these two points within your paragraph are non-sequitur. ("Proprietary Sotware" != "Operating System)
What's happening in software is the same thing that happens to any marketplace that gets commoditized... the base price of the commodity (EG: Operating System) drops to a very low level based on the cost of production and distribution. But value added can increase that price sharply.
A tomato is quite cheap. Tomatoes made into salads and served on attractive plates by sexy waitresses in fancy restaurants are not cheap. The value added is in the air of the restaurant, the clean plate, the sexy waitress, and the tasty salad.
Grocery store tomato is analogous to OSS software.
Restaurant is analogous to proprietary software.
It's been happening in marketplaces for a long, long, LONG time.
Re:Clear for a long time (Score:3, Interesting)
That's the argument, but I don't buy it. As you even acknowledge, in many cases those choices are just fluff anyway, so why bother? What exactly do you do with your computer that didn't 8 years ago? What capabilities has all that additional cruft enabled?
Personally, I think the time has come for an old idea to return. We need to see the resurgence of low power, fixed (or mostly fixed) spec machines ala the Commodore 64 and the Amiga.
Force Development to return to "the bad old days" of using lower level, incredibly more efficient languages. Object Orientation has not just cost us in speed and memory usage, its nigh impossible to multi-thread as well. Turns out we need to turn back and dust off Procedural techniques to make use of new hardware. A return to ISO C, or possibly a new derivative with more advanced support for multithreading (but which would fundamentally work the same way). Object orientation is high level cruft to be discarded, and that means Python, C++, C#, Java, and a host of others get tossed out the window. Good Riddance.
More refined programming practices (Score:3, Interesting)
I do not believe there is anything malicious that has caused this inefficiency to rise. The cost of developing software means that slow and bloaty is what we end up producing as software engineers. It just makes more sense economically.
I hope that in the future, with capped per-core CPU speeds, we will see a renaissance in tight programming. Perhaps new languages will spring up that offer the efficiency of C++, but with the coding efficiency of ECMAScript4 or even C#. D is one such language [digitalmars.com], and there may be many more to come.
We may also see much smarter compilers built on ideas like LLVM [slashdot.org] that will offer statically compiled languages some of the benefits of dynamically compiled VM code, just as taking advantage of specific architectures, and extensive inter-process analysis of code.
With many software problems becoming better understood, we could see much more extensive system libraries that offer the same features as say the .net environment. Cocoa and QT4 are already heading in this direction, with a really feature rich set of libraries, but also with the eye on cutting down memory usage and CPU cycles. In the future, we may see much more optimized shared library usage for system-level applications. This will lead to a snappier user experience. We won't need to have 100s of megs of shared libraries duplicating so much across so many apps.
I think massive parallelism in user applications will never happen without a complete rethink from the OS up, or a new application development paradigm. Furthermore, most developers simply aren't up to writing thread-safe code - it's very hard to get right, and often you don't gain that much for standard applications. That's because of the types of problems being solved in typical applications. We're having a hard enough time writing single threaded apps regardless.
I admire the managed code empire that M$ has built into vista, but ultimately we may want better performance than this heavy-weight approach can offer. Both OS X and KDE are staying away from the managed code "heaven" for the time being. I believe that apple will find a way to make future versions of Objective-C have most of the benefits that managed code can offer, but also with extremely tight machine code.
Re:Eee PC vs. REAL UMPCs (Score:3, Interesting)
Re:Failure of Moore's law is more of a threat to M (Score:-1, Interesting)
You make the mistake of thinking that MS has a monopoly because it has, or had, good software or was the best choice. This is quite wrong, MS software always has been only 'good enough'.
MS has its monopoly via two mechanisms: purchasing power and legal.
The legal mechanism is by tying partners and OEM with contracts. Such as OEM 'per box' contracts, or high discounts dependent on every box installing MS only. Such as non-compete clauses.
The purchasing power is used to buy competitors and either rebrand as MS or simply dump them. It is also used to give away a 'good enough' MS product, or bundle it, to cut off the competitors revenue.
The reason that Linux and FOSS has survived and grown is simply that MS cannot buy it out or compete with cheaper free alternatives.
It is natural selection at work. Given several business models that software vendors used a decade or more ago, MS's was the 'fittest' by being able to drive the others into ruin and take their revenue. MS is the T.Rex, the other dinosaurs have been eaten or driven into dark corner niche markets. FOSS is the mammals.
The problem with the MS business model is that it requires that revenue increase every year. It is running out of markets to take over, or is still losing money in the ones it has recently moved to take over. Its next move was to take over the OEMs by selling its own XPC based on XBox. However, it failed to get its new
With hardware prices falling it was becoming obvious to the OEMs that MS was raking in large amounts of revenue while the profit from the hardware falling. Vista was supposed to help by requiring massive increase in computing and graphics power to raise system prices and revenue and Vista increased in price to ensure that MS got the same or more percentage.
A few years ago MS would have paid the OEMs to bury competing software, or used contracts to ensure this (eg as they did for BeOS, DR-DOS, Netscape, Word Perfect). Today they are paying ASUS and OLPC to put XP onto those machines and giving away XP Basic for free in 3rd world ($3 is 'free').
Re:Slashdot (Score:3, Interesting)
Re:Slashdot (Score:-1, Interesting)
Outdated software? I don't know about anyone else, but I am running Windows XP Home Edition. The last time I updated it was last week. I wouldn't exactly call that outdated. But then again, I'm not an MS bashing troll.
**shakes head**
what cost? (Score:3, Interesting)
There is also the issue of people who have licenced windows in the past and thrown away those machines. I expect to see consumer issues if consumers can't transfer those lic. Esp., in Europe with the regulators having MSFT in their sights.
With only 1 Billion PCS in a world of nearly 6 Billion, I still feel the world needs a $25 computer.