Slashdot Mirror
"Good Enough" Computers Are the Future
An anonymous reader writes "Over on the PC World blog, Keir Thomas engages in some speculative thinking. Pretending to be writing from the year 2025, he describes a world of 'Good Enough computing,' wherein ultra-cheap PCs and notebooks (created to help end-users weather the 'Great Recession' of the early 21st century) are coupled to open source operating systems. This is possible because even the cheapest chips have all the power most people need nowadays. In what is effectively the present situation with netbooks writ large, he sees a future where Microsoft is priced out of the entire desktop operating system market and can't compete. It's a fun read that raises some interesting points."
...and doggone, people like them!
Being saying since the Pentium II days. This "always-be-upgrading-the-latest-spec" is fine for hardcore users, but for everybody else, "good enough" happened quite a few hardware generations ago. The sad part is that we're only now having this conversation.
Linux, you magnificent bastard, I read the fucking manual!
Apple prides itself on some very quality products (both hardware and software) and makes quite a penny. Not to mention BMW, Dyson, etc... the list goes on about companies that spend a lot on design and reap the rewards.
Make sure everyone's vote counts: Verified Voting
The future operating systems will ensure that the average joe needs the latest and greatest so that the Geek Squad will support their PC.
This kinda of reminds of the '640KB should be enough for everyone' theory. If everyone is just content surfing the web and writing e-mails, then sure the 'good enough' solution sounds fair, but if 'good enough' also means dealing with a Windows ME experience then no thanks. At the same time what is considered 'good enough' will evolve over time and new solutions are created and user expectations evolve.
Will my 'good enough' computer handle my photo library, my 32MP entry level camera, recognise the faces in my photo collection. This sound like far fetched stuff today, but as these technologies peculate down from high end systems and people get used to the computer doing more of their mind-numbing repetitive tasks, user expectation will adapt and want them in their 'good enough' computers.
In many ways plenty of people are already using 'good enough' computers. Whether they are satisfied with them is a whole other question.
Jumpstart the tartan drive.
I'm all for cutting costs using an open source OS, but the problem with increasingly cheaper hardware is staying power. Yeah it might be all you need, but how long is it going to be around for. Of course the trade off is, is it cheaper to get short term cheap computers, or long term expensive computers. And, to top it all off, if we do switch to a disposable computing model will we having recycling programs in place to make sure we reuse the rare and valuable parts, and keep the really toxic parts out of landfills?
The musings of just another geek and his junk.
640K is more memory than anyone will ever need.
There is nothing particularly insightful about the article. Obviously the largest portion of the computer using population would never need cutting edge power, so effectively "good enough" has always been the paradigm. How many of us have super computers? This is just a piece with some wishful thinking hoping that people eventually see through Microsoft's coerced perpetual upgrade cycle.
I got a catholic block.
The article argues that people won't upgrade from XP - it expects that as MS tries to force them, people will migrate to Linux instead. I think as Microsoft discontinues support for XP, people will move to Windows 7 - sales of Windows based netbooks seem to be much higher than for Linux.
Whether the same will hold true when the time comes for MS to try to get people to upgrade from Windows 7 to whatever comes next, it's too early to tell. Hopefully by then Linux will have managed to gain enough market share that most people have heard of it and/or know someone running it and the barrier to a non-MS OS will be much lower
Struggling to find a day everyone can make? WhenShallWe.com
I don't know how less good enough computers are going to become over the next ten years. It might be a an issue of power, but I think what happened is that we realized that computers became over powered for the average user. This is not an issue of good enough, but of not expending resources on things that pretty much have no value.
"She's a scientist and a lesbian. She's not going to let it slide." Orphan Black
The corporation narrowly fought off an antitrust judgment under the (first) Clinton and Bush administrations
I thought Obama was going to serve 5 terms as president, after the constitution was amended of course. Mrs. Clinton never gets her shot as President! Obama, will after all, turn the world into a utopia, and "good enough" will apply to all aspects of society, and we'll all smile and nod our heads in peaceful bliss.
Hell, that was 10 years ago.
If we hadn't let the programmers run amok and force them to write efficient code, what we had back then was 'good enough' for most people. ( not all, but most )
And to prove my point, i'm still running a 10 year old desktop with a 900mhz PIII running Freebsd on a daily basis.
---- Booth was a patriot ----
Not willing to spend a lot of money on something that will lose its value faster than... well... anything, really, I have adopted the "good enough computing" doctrine years ago: I find computers that are sufficiently powerful for my use as cheaply as possible - nowadays they're usually free. I have gotten several perfectly good computers by saying "I can take that off your hands if you want.
So far all my software needs have been covered with Linux and other open source software.
I do have two Macs, but they follow the same philosophy: the combination of hardware+software is good enough for the purpose, and keeps its value better than a PC. [source: local sales of secondhand computers]
Usage: km/h for speed (kilometers per hour); kph for very slow impulses (kilopond hours).
Is predicting the rise of "good enough" really all that bold? Although we don't think of it this way, the rise of "good enough" has already happened at least once.
Remember all those $10,000+ Real Serious Workstations, running Real Serious OSes that real computer users did real work on, back when the kiddies were twiddling bits on the Z80 box they built in their garage? All of them are dead. Almost all computers now in use are the direct descendants of the low end crap of the past.
Further, even within the category of boring x86s, almost all of us are already running something much closer to "good enough" than to "good". Some enormous proportion of PCs are in the sub-$1000 category, which still entails a bunch of tradeoffs(not nearly as many as it used to; but still).
It will, indeed, be interesting if Microsoft hits the chopping block during the next round of "good enough"ing(or, more realistically, gets shoved to rather more cost insensitive business sectors that like backwards compatibility, the same way IBM was); but "good enough" is already all around us.
Especially since the advent of "Slop-Ware" and Windows versions that need exponentially more power and capacity than the last version.
Tsukasa: All I really want, is to be left alone...
Even now, the low end, cheaper netbooks [often with no CD drive or even hard drive] are very popular.
A lot of people like to use them as a smaller, less costly replacement or addition to a full blown laptop.
If each mistake being made is a new one, then progress is being made.
we have to remain careful of competition - being cheaper doesn't help if someone is selling hardware or software under market price in order to maintain market share.
nobody can deny that Microsoft is basically giving Windows XP away for free on netbooks. While they are totally able to do this, Linux can't make up for this loss by stashing vast amounts of money from other overpriced software.
what we need to do is beat microsoft on usability on every aspect, not just price. Including marketability, liability and everything you can imagine.
an OK Computer?
The year of the Linux desktop!
Microsoft knew this a long time ago. That's why they are where they are today... everywhere. You don't need something that's perfect and awesome, you just need something good enough so people can get by. The cost savings you get by not putting tons of effort into perfection can be passed on to consumers, who almost always buy on price alone.
By selling Windows XP you can bundle in a lot of trial versions of programs like Microsoft Office, virus scan etc. New computers are stuffed up with adware these days.
This means the effective price of Windows XP is actually negative. Something Linux can not compete with. Who wants to pay to bundle a trial of an office package with Linux that comes with Open Office preinstalled?
To continue the usual car analogy, this isn't what has happened with technology such as cars. Cars were "Good enough" long ago, but these days most cars still have an excess of performance and are far from "Good enough". Ok, I'm not entirely serious - I think we'll reach a point with computers where the performance gain becomes negligible (Either that or the current trend of bloat and crap increasing and everything being just as slow will continue). As there has been a recent surge in more environmental/efficient cars, similar things seem to be happening to computers - there are a decent number of advances in saving power and things these days in technology.
-- All your booze are belong to us.
Sounds just like a man whose never played Crysis on the 'Superduper Extreme Fantastic' Graphic setting.
On an unrelated note, can I borrow some money?
Keep in mind that we cannot predict the future. There may come along some amazing, got-to-have killer app that netbooks and five year old Dell desktops can't run. Maybe high-def Twitter video feeds. Who knows? But let's not get complacent and assume that just because hardware is good enough for now, that it will always be.
> To keep costs down, the manufacturers shied away from pre-installing expensive Microsoft products and instead distributed Linux (and later various renditions of BSD and OpenSolaris too).
I believe that Microsoft pays manufacturers to put Windows on their machines. I don't think most of their profit has ever been from their OS, but rather from their add-ons like MS Word and such. Things people don't think they can live without.
I could be wrong, but I believe there was an article about that with regard to netbooks a while back.
I've been in the small/medium sized business support for a while and I'm here to tell you that "Good Enough Computers" are the standard. You'll have a few engineers and designers (along with a boss or two that is a wannabe nerd) that have the latest and greatest but the vast majority of users in those businesses have had good enough computers for a long time. Sally Dataentryspecialist has a computer that she can type up Word documents on. Jimmy Executive has a laptop that's just good enough to browse porn and play DVDs. This includes home computers. They never ask about some brand new state of the art system (see exceptions above), it's always about the eMachine or Gateway that their dear grandmother left them when she died, and the only use it saw before they had it was traveling to church websites on Sunday.
This is especially true in small town America.
The reality is that computers today "live longer" than they used to. Having a 9-10 year old computer was once unthinkable; it's now almost normal for just about any old Pentium 4 to still be in use today, and the Pentium 4 was apparently released in late 2000.
I put a new (but cheap!) AGP video card into an older P4 desktop computer (hint: PC-133 RAM!) that my son now uses to play Spore - one of the newer, hotter games around - it plays just great.
It's a trend - computers are "doing" for longer than they used to. They are in use for longer, and people hang on to them longer. They are less willing to buy the top-end because there's no reason to.
I have no problem with your religion until you decide it's reason to deprive others of the truth.
But "good enough" computing won't suffice for gamers. They're usually the ones who drive the cycle of upgrading usually anyways. Most gamers' systems are ridiculously overpowered (mine included), and will continue to be so, well after games have reached the point to be indistiguishable from reality. They're always going to want to push that just one FPS more, that extra level of AA, etc. PC gaming enthusiasts won't go away, and as more generations grow up with computers, they'll become more adept at using them, meaning they're going to be doing more, pushing systems harder. Frankly, most PCs from the past might very well have been good enough - if they'd had the RAM available to run a Web Client, Email Client, IM client, video player, casual game, and random other widgets and programs in the background without slowing to a crawl. Think that's excessive? I've seen it, multiple times. So imagine kids wanting to do even more than that all at once. Ignore the Windows vs. Linux argument. The core of it is, even if people got to the point where they all flipped over to Linux, the "good enough" computers of today just wouldn't be acceptable for the kids of tomorrow. However, it will also be like the seniors of today using their large cabinet TVs that are 20+ years old. Once someone today finds a computer that will let them do all their stuff that they need, they'll stick with it, or something similar for as long as they can. TFA is just an author-wank.
Canada: The US's more awesome sibling.
If you are still running everything locally then 'good enough' is a moving target. I've found that my desktop requirements are dropping as I move my storage to a NAS appliance, my applications to server class hardware, etc. In business it is very much the same. Doesn't have to be the cloud.
J2EE Java Enterprise Edition
WSDL Web Services Description Language
EJB Enterprise Java Beans
JSP Java Server Pages
JSTL JavaServer Pages Standard Tag Library
JMS Java Message Service
JTA Java Transaction API
JAF Java Activation Framework
JAXP Java API for XML Processing
JAX-RPC Java API for XML-based RPC
SAAJ SOAP with Attachments API for Java
JAXR Java API for XML Registries
DOM Document Object Model
SAX Simple API for XML
JNDI Java Naming and Directory Interface
JAAS Authentication and Authorization Service
Please, help expand
This is the reason why I felt compelled to upgrade from dual-core to quad-core for the processor, or go from 4GB to 8GB for the system RAM. None of the software that I use will benefit from the increases. Now if the price was right, then I might upgrade because I'm a cheap bastard. I'm still waiting to get a quad-core processor for $50 USD or less.
Parkinson's law is "Work expands to fill all available time". It applies to processing power too. What's "good enough" today won't be "good enough" tommorrow, because someone will invent some CPU-sucking memory-hogging disk-flogging killer app that everybody will want to have.
I don't know what it will be. But then again, who predicted grandmothers would be editing home movies of their grandkids on their computers? Try that on a machine which is just "good enough" for email and the web.
I spoke to a colleague earlier this month who takes mounds of corporate data with her on the road.
Her office doesn't allow remote access yet.
When it does, all she'll need is a terminal-appliance notebook that has wireless access and a printer. She'll be able to trade her full-featured notebook for a cheaper, less fragile, more theft-resistant, full-featured desktop.
We will, within ten to twenty years, find ourselves in a world where, in theory, everyone has access to the same technology. The fact is Microsoft, Oracle, IBM and all of the others will realize that the Internet is like Aerosmith, the poor man's Rolling Stones. All data will be centralized and coordinated by governments and the wealthy to be parlayed into election results and financial gain.
Those unlikely enough to be on the outside of this new class-based proprietary world will be lulled into believing that they are happy. The end result will be a pseudo-matrix that allows the haves to control the have nots without anyone being the wiser.
Or perhaps not.
William David Howell Sr.
The obvious going mainstream seems to be the stimulus for it ceasing to be true. Extrapolating from the popularity of sensor equipped devices, like the wii & iphone, it seems likely that computers that monitor and respond to your gestures, voice and attention will be arriving soon.
IF people where to migrate to OSS, it would have happened already: see Windows ME and Windows Vista.
Microsoft survived the release of Windows Vista and will survive the release of Windows 7.
And of course, I've been reading wishful histories for the last 10 years about how OSS would rise and destroy Microsoft. And still, OSS didn't get any close to that...
For most people software written a decade or more ago was "Good Enough" and they don't need modern technology.
It is called Retocomputing when you use old computers and old software. You can buy them cheap at Auctions and Garage Sales and eBay.
Remember, Slashdot does not have a -1 disagree moderation, and no, troll, flamebait, and overrated are not substitutes.
Ok it's only been 8yrs not 10, but yes, XP is pretty old now and most people still view it as "good enough".
It is patently and obviously ridiculous. A Pentium II PC, especially on a Pentium II-compatible motherboard with its memory and other characteristics, would not be an acceptable platform for the average user. It would be very slow and would immediately have memory issues. Current graphics hardware would probably not be compatible, and even if it was the 3D software like OpenGL or the MS equivalent would have unacceptably bad performance. Contemporary games would be dreary experiences indeed.
Lots of multimedia authoring software can use as many cores and as much RAM as you can afford. 3D gaming environments with ever more active objects, each with some amount of basic AI and moving parts, will also keep pushing the envelope even further. "Tab creep" in your web browser, where you end up accumulating open tabs, each with graphics, javascript, and maybe audio or video give memory footprints well into the hundreds of MB.
Maybe deaf and blind little old ladies with severe arthritis can get by with a Pentium II, but not too many others. In 2025 the things that will pass for personal computer desktops (something like them will still exist in spite of the cyclical "The PC is Dead" hype), will have a dozen or more CPU cores or perhaps hundreds of smaller cores of various kinds to distribute different types of processing. Cache memory will be much larger than today as will be system RAM and storage. Software will be similar to today's except for far greater detail and granularity of content, and multiple new ways to interact with the data. That will demand a lot of compute power.
No doubt people will continue to say things like "an exaflop and a zettabyte ought to be enough for anyone," and people like me will continue to deride and mock them.
Will the concept of "good enough" really apply to future technologies is the real question. What about surface computing? Augmented reality? 3D gaming? Incredibly fast processing (be it encoding, copying, etc.) of high-definition content (the future holds far better than 1080p quality video). What about all the data that needs to be indexed, and indexed quickly. Computers will eventually hit a point, long in the future, where there is so much daily data being processed and indexed that some huge leaps in CPU, GPU, Memory, Storage, etc. will not only be demanded, but required. The entire computing process is going to change and evolve heavily in the next decade, much as it has throughout it's current existence. I can see no reason why that might stop, especially in the long run.
You know, short of nuclear fallout or global warming......
In 1995, listening to an MP3 file required most if not all of your CPU power.
In 2005, listening to an MP3 file required 1% of your CPU power, if even that.
What is "good enough" depends on the tasks to be done. Opera has proven that good code allows the Nintendo DSi to make a "good enough" browser on a very limited platform (and to those who have tried Opera on the DS/DS Lite - there's a world of difference between Opera DS and Opera DSi).
In the end, your hardware is only as "good enough" as your software can make it.
The market wants cheap Java coders mass produced from India and Nehalems, not expensive Linux programmers and P3s.
Gamers will always demand up-to-date powerful systems. Gamers drive PC sales. You lose this argument.
You know, I understand why everyone loves to bash Microsoft and their business practices.
But I can't help but wonder about the effects of the downfall of one of the biggest tech employers in the United States.
A work that expires before its copyright never enters the public domain and thus enjoys eternal copyright protection.
Some of the most fundamental materials used in creating ICs and LCDs are beginning to run out in their natural form (some expected to be depleted as early as 2015), hence recycling old parts and R&D into using other materials may set, as far as manufacturing is concerned, some technologies back to the early 90s.
How can any rational person assume future computing technologies will be "Ultra Cheap"? Case in point, we've been making cars for nearly 100 years, the "ultra cheap" cars are the rust buckets from 2 decades ago. In most of the 1st world a new small hatchback like vehicle is roughly 25% the average annual salary, and this has been the case for the last 40+ years regardless of all the manufacturing and design advances that have come about and increase in demand for such simple models.
Arash Partow's Philosophy: Be a person who knows what they don't know, and not a person who doesn't know.
Useless article. The standard has always been decided by what's available. Unless people suddenly lose their human nature, there will always be the need for better then what's the current standard, and therefore there will be automatically bad good better and best. If there is nog need for 'best', 'better' will be 'best' etc.....
Well then, get on writing efficient code that'll decode HD video on a 900mhz processor. Don't tell me that is something "normal users" don't want, video on PCs is exploding and people are all about a higher res better looking picture. Don't forget the 5.1 audio that goes with that, and HRTF calculations for those that want to wear headphones but get surround. Oh it can't handle that? Well there you go then.
I get real tired of this whining about "Programmers aren't efficient," thing, as though the be-all, end-all of coding should be the smallest program possible. No, it shouldn't, computers are getting more powerful, we should use that power. There are a number of reasons for programs to get bigger and require more power:
1) Features. I don't want computers to be stuck and never get any better. I want more features in my software. This goes for all software, not just power user type apps. For example one thing I really value in Office 2003 (and 2007) is their in line spell checker. It is very good at figuring out what I mean when I mistype, and learns from the kind of mistakes I make to autocorrect and make more accurate guesses in the future. Well guess what? That kind of feature takes memory and CPU. You don't get that for free. No big deal, my computer has lots of both. But it isn't "bloat" that it has features like that, rather than being a very simple text editor.
2) Manageability of code. Generating really optimized code often means generating code that is difficult to work with. I mean in the extreme, you go for assembly language. You get the smallest programs doing that, and if you are good at it the fastest. Ok great, but maintaining an assembly program is a bitch, and it is easy for errors including security issues like buffer overflows to sneak in. Now compare that to doing the same thing in a fully managed language like Java or C#. Code will be WAY bigger, especially if you take the runtimes in to account. However it'll be much cleaner and easier to maintain. No it won't be as efficient, but does it matter? For many tasks there's plenty of power so that's fine.
3) New technologies. HD video is an example that is out now, true speech understanding (as in you can command the computer using natural language) would be one that we haven't reached yet. These are things that are only possible because of increased processor power and memory/storage capacity. Look at video on the computer. For a long time it was non existent, then when it started it was little postage stamp sized things that weren't useful, to now where you have full screen HD that looks really smooth. It wasn't as though peopel haven't always wanted better video, it was that computers back in the day couldn't handle it. Only recently have drive become large enough to hold it, and CPUs fast enough to decode it in realtime.
4) Faster response. Computers have gotten MUCH faster at user response. The goal is that users should never have to wait on their system, ever, for anything. The computer should be waiting on the human, not the other way around. We keep getting closer and closer. If you don't try new systems it is hard to appreciate, but it has been massive strides. As a simple example I remember back in high school when I went to print a paper for school, I'd issue the print command and wander to the kitchen. Printing a 5 page paper was a lengthy process. The computer had to use all it's resources for some time to render the text and formatting in to what the printer can handle. Now, I submit a 50 page print job with graphics and all and it is spooled nearly immediately. The printer has the entire job seconds later, since these days the printer has it's own processor and RAM. It is printing before I can walk over to it. Things that I used to have to wait on, are now fast.
5) Better multitasking. People like to be able to have their computers do more than one thing at a time, and not bog down. It can be simple things like listen to music, download a file, and surf the web but not that long ago it wasn't possibl
So this is /. and it just wouldn't be right if we didn't have a car analogy in this thread, so here's mine:
Computer speed and capabilities I expect will be like cars, which initially went through a growth phase where new models often produced substantial advantage over the previous models but then new advantages levelled off sometime in the 80s (e.g. a Ford Model T gets the same fuel economy as a Ford Explorer). Nowadays, I could go shopping for a Formula 1 racer or a Porsche, but why would I bother to pay that much for something that goes a lot faster than I need. What I need is a vehicle that gets me from point A to point B. Sure, if I'm into racing then I'll get a faster car, but if I want to go to the grocery store the Porsche won't expedite my trip sufficiently to justify the added cost of the Porsche.
With computers, we're at the point where if I want to run a word processor and some nice-ish graphics, I can buy a >2 GHz machine that does what I need and I don't need a dual quad-core Xeon. Sure, if I'm a gamer (=car racer) I need a suped up machine, but for more mundane tasks like e-mail and web surfing it doesn't really justify the additional cost to have a top of the line equipment. With cars, we're limited by the physical infrstructure and population density (I can't do 120 mph to go to the grocery store), whereas with computers, we're reaching the limits of what you need to have a good experience with a basic computing tasks, the physical limit of having a 2-D monitor and rendering images and sounds on it.
Gentlemen! You can't fight in here, this is the war room!
In one sense I despite bloatware, and graphical sugary like Vista's aero skin (despite how cool it looks), if it comes at the cost of unresponsiveness.
On the other hand, it's actually a really good thing, since the market then forces CPU speeds to increase. Extra CPU speed = good for scientific research, day to day tasks and of course games etc.
The great thing is that previously small, efficient, fast apps can now be lightning fast. Everything steps up a gear. Realtime raytracing becomes possible. Writing code can become simpler and more modular etc.
It's all good.
Why OpalCalc is the best Windows calc
Microsoft has already been undercut on the desktop but recoups the losses on the server side of things. Their licenses arent at all that cheap once you really start to add it all together. They have understood what the Linux vendors cant get into their thick skull, that the path to the server market goes through the Desktop and thats why they are more than willing to continue even if they loose money on every desktop license sold.
Until some other OS vendor really steps up to the plate and really commit to the desktop Microsoft will keep their hold on the market for desktops and make serious inroads on the servers.
HTTP/1.1 400
Then they started creating apps and applets in not-so-lean languages like Java. I don't anyone stopping, either.
I agree with this article whole-heartedly and thank the author for it. I have recently begun purchasing (and selling) reconditioned systems from reseller companies on Ebay and they run Acrobat 8 and Office 2007 just fine. What more could be needed? Let the prosperous folk have their overkill and I'll be just fine touting the speed of a Pentium 4 2.8 to my clients. Based on some of the other comments though, I do think a Pentium II and III are super ancient and further precious time shouldn't be wasted on them. It is great that XP has finally hit it's stride just like the days of Win98b. Now if I could only change the case on a Dell GX270 (small form factor), I'd be a happier camper!
Sorry. You lost me at "Over on the PC World blog, Keir Thomas engages in some speculative thinking."
Just news, please.
"When you see a unixer brainwashed beyond saving, kick him out of the door." - Xah Lee
Still camera and video camera resolutions are on the increase, not decreasing. Cheap chips are not going to be able to handle well consumer focused applications making management of these large volumes of data easy and practical.
I would totally agree with your premise (Good Enough is what people mostly do) but not with your conclusion (that underpowered systems will really be that useful for a large number of people).
You tout Spore as one of the "newer, hotter games" - but let's be realistic about how taxing it is, nowhere near the level of a modern photo management/manipulation app.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
Computers (specifically x86 based PCs) have been "sufficient" for the bulk of computing tasks for close to ten years now. The power race has given way to the low power consumption race. Historically electronic devices were built for a specific purpose and programmed to do just one task. The PC was envisioned and designed to be general purpose because the average household could not afford an electronic word processor AND game system AND spreadsheet AND etc. Now that x86 hardware is a commodity the switch to special purpose devices for the consumer is progressing. I suspect we will be seeing more and more devices with an x86 core that have nothing to do with being a PC.
"A person is smart. People are dumb, panicky dangerous animals and you know it." - K
With the debut of Windows 3.1 'good enough' became the accepted norm in computing.
You could pay more for a NeXT workstation, a Sun workstation, or even a Mac. However Windows 3.1 was 'good enough'. Most people didn't need networking support built in, or the compilers or software that was available for the other platforms.
You could have gone all multimedia with a fancy Amiga that did incredible sound and graphics, but 16 colors and trading files via floppy was 'good enough' for the majority of people. You could add hardware and software to Windows 3.1 computers if you really had a need to network them. The computers Windows ran on were capable of displaying better graphics (games that booted to DOS showed this), but Windows 3.1 was 'good enough'.
Windows 3.1 really did make computers easier to use. Macs, Amigas and NeXT did a 'better' job of making computers easier for people, but Windows 3.1 did a 'good enough' job at making things easier. At about US$2,400.00, a mid range computer with Win 3.1 on it was a lot cheaper than the competition. It was 'good enough' and cheaper.
The history of economics shows that 'good enough' and cheap wins.
Think of the 'best' hamburger that you ever ate...
Did you think of a plain old McDonald's hamburger? Probably not. In any scale of human measure (taste, smell, satisfaction) McDonald's hamburgers rarely rank as 'best'. But measured in market share the McDonald's hamburger is the best.
Ford's Model T was not as fast or as fancy or as comfortable or as good in quality as the hand crafted automobiles it competed with. But thanks to mass production and economies of scale it was cheaper and it was 'good enough'. Ford and other mass produced vehicles dominated the market. There are still purpose built vehicles, but they are a small specialty segment of the market.
'Good enough' and cheap is always the 'best' when you consider things from a market dominance point of view. What a human thinks is 'best' and what the market thinks is 'best' are not the same thing.
Nobody knows what UI fads, ideas, and abilities will be available in the future. Maybe voice recognization will replace cell-phone menus, for example. These may require fat processing to do well. Thus, the "good enough" CPU claim is suspicious. Or maybe a combo of voice recog. and finger gestures. Nobody knows, and if they claim they do, phooey!
After all, the ideal interface would be like an e-secretary that knows your preferences and habits and it not overly literal like today's computers. That obviously is going to require a lot of processing power to even approach it.
Table-ized A.I.
What we will have are personal computing cores that are capable of connecting with a wide variety of peripheral devices. Such devices might include displays, input devices and combo devices that people might mistake for laptops. PCCs will contain the processing power and the primary storage that people use for their day-to-day activities. And PCCs will be incredibly tiny devices and will go anywhere.
"I think there is a world market for about five computers"
http://en.wikipedia.org/wiki/Thomas_J._Watson
They were good enough then. Since then, the market has expanded a little.
I think you're missing the point, which I take to be this: We've reached the point with regards to hardware, that we *already* have "staying power", except in all but the highest-end applications.
Even now, what you'd most likely deem "cheap" hardware is more than capable of running the most common applications well, and the OS' themselves are sufficiently reliable that one of the compelling reasons to upgrade, better reliability from a new OS (as Microsoft has always promised, but failed to deliver [1]), has passed.
Well, if it's all you need, then by definition it will around until that is no longer the case. IMHO, *that* is as it should be: Use it until it 1) fails or 2) No longer does what you need/wish it to do [2]. And that's the point, I think, and one of the reasons why Netbooks are so popular now: They do what they do, regardless of hardware and OS, well enough for those that use them, at an affordable price.
But, that's no longer really the case, you see. To coin a phrase: We've reached the age of "utility computing", where a computer's usefulness is no longer measured so much by it's raw specs and OS revision, but rather, how suited it is to the task(s) for which its user needs it.
This isn't a bad thing, in my estimation, and in the long run, addresses your final point:
While computers may well be "disposable", in some sense, they are also longer-lived in general now, which offsets that to a large degree [3], and recycling programs are already in place for them once they are no longer useful to their owners: There's sometimes a family member that can use it, or charities to which one can donate such, and other recycling programs at local and national levels (at least here in the US: After not being able to find a home for the various old computers, monitors and peripherals I've accumulated over the years, I save them and take them to my county's local drop-off point - they advertise such at least once per year now).
In addition, all of the companies for which I've done service this past decade or so also have recycling programs in place now. Old hardware is replaced on a planned basis: They amortize it from an accounting perspective, then replace it once it is out of warranty - typically 3 years. After that, some systems become "beater"/test platforms for their hardware/software engineers (which saves them the money to have to purchase such, and also allows them to test on older hardware), or is donated to local charities (giving them a tax write-off), or is taken away by a recycling company (which gives them PR value as being environmentally conscious). And, this isn't stated cynically, mind you.
I, for one, welcome such, and hope that it will (re)create an era of software efficiency and reliability: Computers are/should be tools, after all - maybe "slowing down" ever-increasing hardware and software upgrades will bring that back into focus.
Regards,
dj
Notes:
[1] And, certainly, Microsoft is by no means alone in this. Apple has been known to release OS upgrades from time to time that make older Apple hardware "obsolete", and let's face it, the most popular of current Linux distributions also increasingly fall into this category as well. Certainly, one can say that this is "the price of progress" - but let's at least be honest about it, and what it enta
I'm probably going to be laughed at, but computers today are "good enough". Ever since getting a core2duo processor a few years ago, and just recently 4gb of ram, I rarely approach those limits. The bottleneck seems to be the harddrive, even with RAID0 (I guess this may be solved in the future with solid state disks).
My music is already FLAC and I can play 1900x1200 video with no problems, so unless we are moving to lossless 32bit 192 kHz 10.1 surround sound and holodeck virtual display technology, the main bits of my computer will be good enough for me likely for the next 5-10 years. I never said this in the 286/486/pentium 4 days.
No wait, make that 2000.
Don't jump to your guns and proclaim ESR's clairvoyant genius, though. In 2000, he said to LWN "I believe that will happen probably within five to six month from now." I hope for his sake he didn't sit and wait it out...
"computing hardware had evolved sufficiently to reach a level of performance that allowed for speedy execution of virtually all common computing tasks."
640Kb was once way more than enough for speedy execution of virtually all computing tasks. In fact show me a point in computing history where current commodity hardware wasn't adequate for all common computing tasks? TFAs premise, despite only being a bit of fun, is a fallacy unfortunately. Who's to say there won't be compelling new computing tasks demanding ludicrous performance in 2025?
I think the uptake of low-power low-cost computing in emerging markets and in ultra-portables, is being misconstrued as a shift in trend and not what it really is - a niche being filled with new consumers.
After logging in slashdot still does not take you back to the page you were on. It's been that way for 20 years.
I do not think it means what you think it means, Mr. Article Writer.
The geek is always willing to bet that Microsoft [and Intel] can't compete at the low-end of the market.
It's a bit sad really.
Rather like watching the loser who can't resist feeding another coin into the slots at Vegas.
and gmail is leaving beta
Or maybe now you're older, less susceptible to marketing, and play fewer games that crush the exponential edge of hardware. In other words, perhaps now you have more perspective than you did during the time when you had to have the latest and greatest.
There will always be higher res movies to view and process
But not much higher. At some point, you run up against the limits of the cornea, lens, and retina. Snellen's eye chart is based on a resolution of 60 pixels per degree,[1] which is close to the resolution of a 1920x1080 pixel monitor at 30 degrees across.
[1] The glyphs on the 20/20 or 6/6 line are 5 arc minutes (5/60 degree) tall when viewed from 20 ft or 6 m. And an E is obviously 5 pixels tall.
let's throw in one more pretty mainstream application: television with timeshifting. Mom doesn't have to be "hardcore" to want a PVR. Now you're encoding/decoding video.
On an ASIC. The CPU in a TiVo DVR is fairly slow, or at least it was in the early Series. It uses a dedicated MPEG chip for the heavy lifting, which is more power efficient than trying to do the same thing on a conventional CPU.
WinCE isn't Windows.
But it does have applications.
Porting Windows to ARM would be as pointless as their previous ports to PPC, MIPS, Alpha, Itanium, etc. If the ISV community doesn't port the whole effort is wasted
Windows NT 5 (Windows 2000, Windows XP) on i386 could run apps for MS-DOS through NTVDM. So why can't Windows 7 on ARM run apps for Windows CE through a similar technique?
Especially since I don't think the Windows toolchain is setup to cross compile
On which toolchain does one normally develop Windows CE apps?
Linux is already far more prevalent than many of us could have dreamed a decade ago. Linux ships on cell phones, set top boxes, routers, laptops, desktops, even TVs.
But how many of those cell phones, set top boxes, and TVs use code signing to shut out applications developed by third parties? Or would you argue that "the unwashed masses" do not need applications developed by third parties?
Call me back when my computer can infect my brain with a virus by projecting a bitmap on my retina. Until then, it's not good enough.
Photo albums? On the web. Document processing? On the web.
Yeah, when 4G wireless comes out. Until then, if you happen to live in an area without cable or DSL service and are not in a position to move, good luck with your 0.05 Mbps connection.
Not really. The megapixel wars are basically over.
Even if that were true (and if you've been paying attention to recent cameras, it is not) you discount other improvements beyond pure megapixels - cameras delivering more HD video on the fly, or voice annotations.
But let's not discount megapixels just yet, because the marketing departments will be cramming them in for a while yet.
And of course you've also discounted all the processing power required as I said to run complex photo management and editing apps. It takes a lot of heavy lifting to make things like face recognition easy for users, or video effects not drag you down in hours of rendering cycles.
Video is the same. There is really no need to ever capture more than 1080p for home use, the human eye simply can't perceive any quality improvements on screen sizes that are realistic for the home.
Yes but systems now have trouble doing a lot with 1080p, and consumer gear is only starting to reach that point (still a lot of 720p around). And that means again a future for powerful equipment built to work with very large data streams.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
I recently tried Ubuntu after leaving Linux as my primary OS in 2003. You're wrong. The GUIs are only fine if you're willing to stick with their narrow limitations. I think it's because they're constantly being rewritten instead of incrementally improved.
Examples:
When I hooked up a second display and clicked "detect displays", it did nothing. No error message, no effect. I see no way to fix this without editing config files manually.
My sound doesn't work at all. It's listed properly in all the config screens, but nothing comes out of the speakers. Now what do I do? I see no easy way to try different driver or other things without delving into a kernel module mess. Hello, terminal.
How do I disable that wretched shutdown beep with a GUI? The mute control has no effect on it, nor does disabling the system beep in the sound preferences.
This is basic stuff that's been an issue for 10 years.
Sorry, but desktop Linux in 2009 gave me the same experience as desktop Linux in 2003. I.e. 3 days of googling and sludging through manuals to get things working. The process is a tad smoother now, but it's still only good for two groups: Grandma who'll leave it the way it is, and experts who live Linux. Almost everyone I've ever met falls somewhere in between. It's hard to be just savvy in Linux. It's all or nothing.
Pretty skins are just that, skin deep.
Don't give me that paid support crap. I've never called MS support. I've never called Apple support. I can figure out how to maintain their systems by using them. If I'm going to have pay someone to help me with how to do basic things in Linux then I might as well just buy one of the other two.
boldly going forward, 'cause we can't find reverse
Photoshop Elements 7 for XP and Vista at $70 currently ranks #4 in software sales at Amazon.com.
The geek never quite grasps the notion that publishers like Adobe compete for the mass market as well as the pro.
The alternative to the GIMP isn't Photoshop at $700 - it is Paint Shop Pro at $40 after the mail-in rebate.
... the next killer app that requires more power will move "just good enough" up a notch, IMHO we already have a lot of pressure. As soon as computers can do more complex tasks you can bet the "just good enough" computer will no longer make the cut.
I always find these articles redundant, the same thing is said every decade for some given technology.
Listening to an MP3 requires 1% of your CPU power today mostly because of the proliferation of SIMD hardware (MMX, SSE, etc.) that can speed up signal processing.
You may be right about 1995. There were quite a few Pentium I machines produced without MMX extensions. So listening to an MP3 required quite a bit of CPU.
But in 1992, MPEG decoding hardware was available, and was even commonplace, on many desktop PCs. On the right PC, listening to an MP3 then required scarcely more processing power than it does now.
Just nitpicking...
"I assumed blithely that there were no elves out there in the darkness"
The apps that need the most power are games. With gaming shifting to consoles, partly because of value for money they represent to buyers (when not taking into account game costs) and partly because publishers and developers prefer a platform with less piracy, that's a main "killer app" moving away from the PC, making it even more "good enough".
Moving games out of PC space is likely to hurt Windows as a dominant OS. Maybe some years in the future we'll see Microsoft as a hardware maker instead of an OS maker.
I'd agree for better 'default' themes (if and when we can define what better is for the mainstream), OTOH to me it's pretty clear that a picker during install is DEFINITELY 'click the wrong button and you're fucked'.
Sorry, but to me allowing to change theme during install is just the opposite of easier. It's geeky. You are a geek ;-)
Herve S.
Before Adblock... you had iCab, for ten years, on Macintosh. ;-)
And still going on
iCab indeed invented ad-filtering.
They are mac-only, they are closed, but boy are they an alternate browser.
In this present world where I can't but cry in front of Apple's obvious shift towards locked/signed systems (on iPhone/iPod), and this moreover in a manner that brilliantly works for everybody from devs to users, the presence of iCab is one of those things that keeps me on Macintoshes.
Yes I got a linux MSI Wind for travels. And back home I rush to the mac with iCab and Vienna...
(Yes someday linux will also copy an efficient RSS reader too, probably)
Herve S.
"Good enough computers" maybe a term that small/medium companies are using even today. But, what about the average joe? I have a 26" monitor that I adore, and I had to upgrade my graphics card to be able to use it in all it's 1920x glory. I have an old AMD 4000+ and I run XP because with Vista, it just crawls. I am inclined to upgrade the rest of the system as from a few months ago, some HD movies are just unwatchably slow on my machine. When I bought it almost 4 years ago, I thought it would be enough for a loooong time: but it was a self deceiving thought I always knew was wrong... I also do some video editing from my family's birthday parties and x-mas, and I have to admit that the newest software is demanding more power. The capabilities of software is always connected to the capabilities of the hardware, and I can't imagine this any other way... Not in such a short period of time as by 2025.
15 years ago (ok, give or take a couple of years) windows 95 was available, as was linux. I think we could try to extrapolate what will happen in 15 years time... I'm sure that, by that time, Windows, linux and a lot of other OSs will still be around, we will need the extra power to do a lot of things we don't think about right now, but when they appear, we suddenly can't live without them (as usual)...
I run a Dell PIII 500 mhz, circa 1999 at home. It's my one and only home pc. I run Win2k with 512MB mem. I run AutoCad 2005 among other things and it is all I need. Yeah, it would be faster if I upgraded but I don't -need- to. Most people don't -need- to. I totally agree with this article and have been shaking my head at the housewife/metrosexual moron for years that forks out for the latest and greatest only to surf menshealth.com and use QuickBooks to balance his personal checking account. When and if I finally do upgrade, it'll be to one of those uber-fast 1.8Ghz beasts.
2025 will be the year of Desktop Linux!!
80% of users need something to browse the web, play media and run MSO. Anything else is asking too much of them. Give folks an OS that is a glorified web browser/video game console that resets when they shut down.
The 3 computers people need
1) Video Game
2) General use (MSO, browser) {build this system and you'll be a brazillionaire}
3) Specialized systems (engineering, number crunching, media editing)
Non-people need Servers.
Sys Admins aren't people.
Computers are hard and scary because they have too many features, too many options. Even iMacs are too flexible/powerful for most users.
Utilizing the synergization of benchmark e-solutions to pre-workaround action items!
I remember commenting on Morgan's Law about the curve that technology (chipsets) have, and
telling people (non believers) that it was all fine and dandy to keep pushing the envelope
with new PCs today, up to a point, that one being the point that the majority of PC users,
are now capable of sustaining themselves for more then the next 5 years with that dual core PC,
and that to get anymore changing to happen, would be like pulling teeth.
I know if I had all I needed, power , speed, performance, and you come up to me and say, you need this for an extra 2k, to be up to date, and I really don't see the need, I won't buy it, and so will many others to satisfied already with what they got.
Just how fast do you need to burn that dvd, or copy that mp3, or how much room on that usb key do you need (64gb is enough ,...trust me!)
I've been a high-end buyer for 20 years now. My experience has been that the cost of purchasing a high-end system has steadily declined over the years. I used to spend $2,500 on a pimped out machine. The next one was maybe $2,000. Then $1,800. Then $1,500. I recently spec-ed out a real nice Core i7 rig for about $1,200. I think this is the more important trend. A few more iterations like this - say another decade - and a high-end PC might run $500. That's the point at which commodity economics really start rolling. It's not so much about the low end being usable. It's about the high end becoming increasing affordable. When the high-end drops enough, that's when good-enough computing really arrives. The price difference between low-end and high-end becomes so small that there is no room for additional costs (e.g. expensive software).
iCab indeed invented ad-filtering.
Are you sure it wasn't in proxies before that?
(Yes someday linux will also copy an efficient RSS reader too, probably)
Akregator seems OK, but I haven't really bothered to look for one.
Don't thank God, thank a doctor!
The question will depend on how many people will actually use/need/want those features. I suspect a lot of applications like that never really get used. Probably because it takes time to do it and master it.
Thus the need for more computation in order to reduce the complexity to the point where normal people CAN (and thus will) use it.
Technical people always seem to underestimate the ability of the non-technical to assimilate and use technical features. What you are doing is the equivalent of dismissing texting because it's "takes time to master" entry of text on phones.
There are so many realms of improvement in what technology can do for photos and video in areas people will use when they work - automatic editing to make videos less boring, person (not just face) recognition (which we are seeing today in consumer apps), automatic integration of video and stillas base don geotagging...
There is a world of possibilities people will fawn over once it arrives. And all of it requires copious processor and memory.
There will always be a place for light devices but there's also always going to be a large market for heavy ones too.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
Aren't chip manufacturers ruled by economies of scale? When people only buy "ultra-cheap PCs", the chip factories will stop making high-end CPUs, etc.
People will also keep their computers longer, since a new PCs won't be any more powerful. There will be fewer chips made & semiconductor prices will go up.
TV Tuners use DSPs, not ASICs. They are not the same thing.
But a DSP does share several die-size-saving and power-saving characteristics that ASICs have over general-purpose CPUs: less need for cache, no MMU, less back-compat cruft in the instruction decoder, VLIW instead of a complicated decoder capable of out-of-order execution, SIMD to take the load off the decoder and put it on execute units, "saturating" arithmetic modes that make code less branchy, and better support for fixed-point math where it makes more sense than floating-point. Things like this are why the Cell Broadband Engine has one CPU and seven DSPs.