Slashdot Mirror
Linus Looks at His Crystal Ball
Magorak writes "There's an interesting article I found here about Linus Torvalds' take on the future of the software and hardware industry. There's tidbits about Linux's future itself, the competition with Microsoft, and about customized software becoming more important. It also gives us an interesting view on how Linus sees the future of computers and technology. "
Take a look at eCos, I think it's made by Cygnus. It's an embedded OS under the GPL.
Wayne
Here is some info on cross-platform integration if you are interested.
This message posted via NetScape 4.61 running on Mandrake 6.0 with a KDE window manager being ran via a Hummingbird X Client on a Windows 98 machine. Integration is fun! :)
I predict writing gigantic, all-purpose software in C++ for income is a thing of the past. To earn a living programming you'll need to consult and do small, specific, internal company projects in a scripted language. All-purpose software in legacy languages, to make grown women cry will be the domain of dorm hackers in their sophomore years.
I doubt purpose built devices will benefit anyone but the manufacturers. I do everything on one computer because it's cheaper than having a dedicated TV for TV, a dedicated VCR for VCR, and a dedicated stereo for stereo, but Transmeta is of course in the embedded system business so his opinion is reasonable.
If you think QNX (cost-wise) is unreal, notice what happened to OS/9 when they got their first set-top box contract. BOOM, all support for small (20,000units/year) customers went away, and the prices went up for small unit volumes. We ended up writing a an OS from scratch that cost less over a small number of units than an OS/9 solution, back in the days before Linux was ready for prime time. There will be a rather large shake-out in the embedded market for the majority of applications where "hard" real-time is not required once Linux is sufficiently well-documented to be used in commercial development. The tools are almost there, and all it will take is for someone to release an embedded development kit with porting tools that support the CPUs normally used for embedded development. Then you will see a shake-out in the embedded market.
Dammit, I knew I should have previewed...
:-)
I sit here at my desk at home and what do I have... A desktop, a notebook to my left, a calculator to the right of the mouse and paper and pencil. In front of the desktop keyboard is a musical keyboard. Why?
Sometimes it's just easier to flip over and use the notebook rather than flip windows around or windowshade a few apps. The calculator is *FAR* faster to throw numbers around in than pulling up a calc app. Finally, the pad of paper and pencil (not pen!) are there because it's a thousand times faster to play with ideas on paper than it is to try to pull it up in a sketch app.
Computers are very fast and do many wonderful things. However they will NEVER replace a calculator, nor will they ever replace paper. When I know what I want to do, I turn to the computer. But if I'm just playing with numbers or trying to shape something, it's paper or calculator... something lower tech and far easier to maniuplate than a keyboard/mouse.
The MIDI keyboard? Well I like to play by ear when I'm listening to MP3s.
I think there will be a steady need for new features as broadband reaches more and more people. Just as the Internet has caused many applications to become "net aware" and spawned many new features (such as your CD player to fetch track names from CDDB) the next level of access will do even more. I'm sure we'll all be disappointed someday when our favorite app doesn't yet support voice recognition, or application sharing over the internet, or doesn't support the newest lossless graphic standard.
Not to mention that we havne't run out of ideas in the hardware arena yet either. Take a look sometime at the discussion over on Macintouch about Connectix not taking advantage of the Altivec instructions in the next version of Virtual PC. These people really are clamoring for every speed increase they can get, and the company does not want to bother to re-code it for free of course. As long as there is new hardware, there will be new software written to take advantage of it.
On the flip side, of course I haven't really found anything new that I use in Microsoft Office since Office95 came out, unless you count Microsoft Frontpage as being part of the office suite. But, like I mentioned above, voice recognition, video-on-demand, and the ability to use VERY large amounts of data in your work without worrying about the amount of bandwidth of your audience will change a lot of things, I'm sure.
When in danger or in doubt, run in circles, scream and shout. --Robert A. Heinlein
To use office equipment as an example, a combo fax/printer/scanner/copier/etc device makes sense for the small office/home office (SOHO) segment of the office equipment market. Single-function copiers will make sense to the "big business" segment of the office equipment market. For either combo units or function-specific units to dominate would mean that one segment or the other is not being properly served (i.e., the SOHO segment will be faced with spending a lot for unneeded functions, like collation options, various paper size trays, etc., or the big business segment will be faced with equipment that doesn't meet functional needs).
The same applies to computing products (hardware and software). What makes sense to one segment of the overall consumer population (say, for example, the all-night hacker type) will not serve another segment (e.g., the "how do I print my letter to Granny" type).
To put it much more succinctly, it's about using the right tool for the job -- and recognizing that the simply-described job (e.g., "copy a document") may mean radically different things in various market segments.
Some tools to consider...
No Laughing Allowed!
According to Linus, Transmeta "makes stuff". So, I assume, Linus was hired to help them make stuff. Linus will be bringing his expertise in stuff design and probably begin expanding into actual stuff manufacturing. With the team they have, I imagine they'll make quite good stuff. Maybe useful stuff. Definitely stuff of some sort or other.
/dev/stuff file or maybe even /proc/stuff, a new stuff driver and an expanded stuff interface). I hear O'Reilly will be partnering with Transmeta to release "Stuff in a Nutshell" and the new "Stuff Administrator's Handbook".
Perhaps the stuff will use Linux, but Linus was not hired to work on Linux (although this is not precluded by his contract). I wouldn't be terribly surprised to see some of Linus' work go into Linux (like a
Citizens Against Plate Tectonics
I believe that devices will become even more multi-purpose. In the future the vcr, stereo, tv, computer, news paper
ayottesoftware.com
>> . There will ALWAYS be a need for constant upgrades while the proprietary software model reigns king
Ah, but by the same token, the proprietary oftware business if pretty reliant on the need to upgrade. That's why I think Linus is right on the mark.
As Microsoft continues to add features to Office, each new feature will appeal to a smaller percentage of users. And as PC prices drop, it's going to be harder and harder to justify a $600 word processor, much less a $300 upgrade every 24 months.
I've watched PC software evolve since the beginning, and there has always been a sort of promised land on the horizon. We were always just a hair short of where we wanted to be. GUIs were unreliable at first, they were too slow. Upgrades seemed necessary, just to make it from Z80-CP/M to a stable 32-bit GUI platform with decent multitasking. It took time, and we bought upgrades along the way.
Now, GUIs and multitasking just work, and we don't think much of it. Windows works pretty well most of the time, Gnome/KDE just need a year or two of polish. And then we're done. The PC has finally evolved to where you turn on a new computer, and no matter how cheap it is, you'll get a pretty snappy GUI interface. And making it twice as "snappy" is no longer a big deal. Most "new" features in current software upgrades are either bells, whistles, or advertisements.
This situation will probably make it pretty difficult to charge a lot of money for the latest version of software in upcoming years. John Dvorak's article yesterday showed rare (for him) insight, when he observed that Linux's desktop challenge will eventually come in the form of a $199 PC that comes with a complete office suite.
I happen to be working with Photoshop every day for a few hours. I tried Gimp as well. Photoshop is the only windows program I'm using. So why am I not switching? Simple. Gimp can't cut it for professional graphics manipulation - far from it. I don't think any Photoshop users (except for maybe warez kids) switched to Gimp - those who payed $600 for Photoshop, need it's high end capabilities. I haven't been following Gimp development effort, but I *hope* in a year or so I will be able to switch. I plan to try the next big version when it comes out. It's a little annoying to see people knock Photoshop though - that pretty much means that you are not a graphics professional and have no business judging Photoshop to start with.
-- ATTENTION: do not read this sig. It doesn't say much.
yeah, i have one too.
it was an accident when i was a kid, i had one of those bmx bikes and i landed bad, and did a full nutcrusher on the highbar. the left one is okay, but the right one is gone, gone.
it was kinda embarrassing at my junior high, because they had those damn hardwood chairs, and if i wore thin shorts you could hear the "clink, clink, clink" as i settled in my chair. i was late to class a lot, so you can imagine what it was like.
i guess they're using silicon now, but i don't wanna get in one of those fucking class action lawsuits if my hair and teeth start falling out or something.
my crystal ball is doing fine, thx. it doesn't talk to me the way linus' does. maybe thats what the new transmeta appliances are, rf networked prosthetic crystal testicles or something.
I don't really care much about for-profit or not-for-profit motives, and I don't entirely subscribe to the FSF's moral obligation arguments....but I think the useless upgrade path created by MS is a great reason to get out and code....
So I'll cut this comment short...i've got code to write :)
Werd.
Doesn't anyone else here think it's kind of odd that all of a sudden many stories are popping up about Linux on the newswires, when, if you look at it, there really wasn't a news event that predicated the need for an article? Now, I'm going to theorize on a few possibilities:
;) is that this is a plot by the OSS movement to Stick it to the Man and to create good vibes towards OSS and Linux in general.
1) Linux makes news. Good news makes money. Money makes people happy.
2) The increasing amount of tidbits related to Transmeta (yet another reference!) and Linus' relationship to it are nothing but fodder to increase awareness about the obvious branding of Trasmeta's product without said product actually being even known. This awareness and publicity are precisely the kind of calculated moves I would take if I were trying to hype the Next Big Thing (tm). I might note that, for the most part, the actual news item, "Software Price Crashes" in this case, are really of secondary importance. If you'll notice, the discussion of Linus, precedes the implication of Redhat's IPO with precedes the blurb about Transmeta, which then precedes the meat and potatoes of the article. Granted, this is an often used trick in the news industry to make sure people have a background for the material to follow, but isn't this a bit much?
3) (The least plausible
I admit that all of this is supposition and is baseless. I have no proof to back any of this up. IANAL. I'm just an avid slashdotter. AT the very least, it's good to not only read the articles presented, but to read past them and between the lines to derive the intent behind the posts in the first place.
If you believe what Sun, microsoft, IBM etc... are saying by the time the upgrade crash occurs most of the infrastructure for web-based office suites/desktops will be in place for a massive shift away from desktop computing by the average user. (uh, if you're reading this you'r probably *not* an average user)
So you won't need to upgrade, just pay your monthly fee. Your provider will upgrade automatically (hardware, not just software) for you becuase he wants your business. Otherwise you may go skipping off to a competitor.
So in my opinion the upgrades will still occur. There will just be a shift in who is actually paying for them. The user's cost will stay constant year to year. The ASP would take the big $$ hit for upgrades.
Assuming the rumors about Star Portal being closed off by Sun actually come to fruition... All we need now is the "Apache" of web application servers. Is this in the works? is it already supported by Gnome or KOffice?
And how will this proposed upgrade crash affect something like the game industry? I'm much more willing to drop $30 on a new game or expansion set than for my word processor.
Sig:
Barbeque is a noun. Not a verb.
Did anyone notice how they talk about embedded stuff, customization, and many devices per home (appliances in buzzspeak, which he avoided of course
Now, the only thing to hope that the three years he is talking about is not the announcement of their product (remember, on Comdex we will know, what it is or we will know when we will know :), but the final takeover worldwide, where there's not anything traditional left, but customized Linux devices running transmeta parts :)
(transmeta reveleation mode off)
"Ten years from now, they could do it in a few seconds." -- The Racketeer of the Hellfire Club, 1993, Phrack 42
"but Torvalds said he does not believe consumers want all-in-one systems."
I fear Torvalds may have underestimated the powerful attraction between Americans and their chairs. I can think of several million consumers who want nothing less than the world delivered to their couch. Or car. Or CRT. Or PDA. Or forearm. Or brainstem.
Then again, I certainly agree - in the near future, at least, I don't foresee all-in-one devices becoming any more usable.
-=Best Viewed Using [INLINE]=-
The article didn't make this very clear, but I think what Linus meant was that the common practice of paying for bugfixes (oh, I mean upgrades!) would soon end.
It's about time, too. I've seen plenty of bugs in Office 97 which prevented people here from doing their work. Microsoft acknowledged the behavior as a bug and recommended purchasing Office 2000 (which would come out seven months later) as a fix.
To make an annoying story short, they ended up using the beta version of Office 2000 just to get work done. I won't miss that sort of thing.
Compare that to the Linux kernel development model. How long had the 2.2 series been out before 2.0.38 appeared? And how much did the kernel maintainers charge for that?
It's not the upgrades that are the issue. It's being "forced" to pay to upgrade software which should have worked correctly in the first place that is the problem.
--
QDMerge 0.21!
how to invest, a novice's guide
The main reason developers can buy computers cheaply right now is that they're the same computer everyone has to buy.
Suppose that home consumers all decide purpose-specific appliances are the way to go, and business decides thin-clients make sense. You won't be able to develop software from home very well on either -- a powerful box with a big hard drive is just much better for that. One of the main reasons is being free to choose your own tools, rather than subscribing to someone's "development suite." Even if you were willing to suffer development on a thin client it's likely that the usage fees would be prohibitive for such an uncommon activity.
So if Linus and Sun are right, expect the price of the box you develop software on to skyrocket, as only enough for developers will have to be made. Note that corporations will develop happily on thin client solutions, internally served and administered. Only open source developers and kids trying to learn programming will really feel the crunch from this.
Under the GPL.
Those are big words. When you market a product that runs on QNX, you pay a smallish royalty on each item you sell.
When you licence it under GPL, you give your competitors the source code.
I don't think that the need for upgrades will disappear in the near future. There are two reasons for this:
Purpose built devices are great, but I think that the future of purpose-built devices will be soft upgradeable as well.
___ This sig is in boldface to emphasize its importance!
thats why its called "open source". BTW, eCos is also dual licensed and available if you pay for it without the GPL.
I've been thinking along these lines lately. I discovered mh and instantly loved the philosophy behind it. For those not in the know, mh isn't a single program. You're never "in" mh, you're just "in" bash (or whatever you prefer). When you want to read the next mail, you issue a command, "next". When you want to see a list of messages, you issue the "scan" command. I'd like to take this kind of thinking a little further. For a long time, the move has been toward making everything graphical, because it's supposedly simpler, easier, and more friendly. However, think about how you talk to the ultimate data processing machines, those that we want our computers to ultimitely be able to mimic: people. Do you use any icons? Click anything? No, you just talk to them. When we want to start talking to our computers, all the work that's been put into GUIs is doesn't help one bit. The command line is the starting point for that work and I think there's a lot to gain by working on a smarter commandline right now. That's not to say that GUIs will or should die out. They have their place. It's just that that place isn't "everywhere".
--
Fuck the system? Nah, you might catch something.
A lot of the features the GIMP lacks is due to nifty patents. Mind you, I'm not exactly "knocking" Photoshop, per se.. and I won't even begin to claim myself as a graphics professional, but aside from the really really serious stuff, the GIMP will handle most of your graphics needs (the other general alternative being Paint Shop Pro, which cute though it may be, lacks a lot of the more useful features of Photoshop, and so is rather pointless except for the low end.. certainly not worth the money for it - which is why I don't even bother to compare PSP with GIMP :).
Besides, I think the chances of the GIMP being "on par" by the time GNU/Linux storms the desktop are fairly good, which is why I like it as my example. :)
~ Kish
If that would be true, how the heck did Furbies ever come into existance?
In our western societies need is really at the bottom of the list when it comes to motivations for a purchase. (Which is a good thing, don't get me wrong)
Well, both the GIMP /and/ Photoshop are for more than "touching up images".. However, I thought it would conserve space if I didn't sound like I was trying to plug the hell out of both of them. :)
I haven't used MacOS in a while.. From what I remember, it's not nearly as bad as Windows, though I'd prefer something else myself. I just figured if Red Hat ran fine on a box with 16 megs of EDO, and a 166mhz Pentium (no, not II, III, Xeon, or even Celeron.. just Pentium.. don't be surprised, it's still using EDO for the love a.. :), then I can't really see the incentive to upgrade from 128 megs of SDRAM and a 500mhz Pentium III to anything higher unless you absolutely /must/ have Win2k.. =P
Therefore.. I'd bet you're right. Which was my point exactly, really. :)
~ Kish
"I'm ignorant on this subject (among lots of others :D). Can somebody explain to me why the need for constant upgrades is going to disappear and how that's supposed to translate into crashing software prices?"
I think what he is saying is that software will be distributed with a piece of hardware, and rarely upgraded, you'll rebuy the hardware and new software together when you upgrade. When was the last time you replaced the ROM firmware running your microwave oven or tv?
The per unit price on the software will be have to be extremely low, basically cost to develop divided by units sold, very little markup, because its critical to profitability of the consumer product. MS will find it very hard to compete with Linux in this regard.
The idea of a all in one device sounds good at first thought. I'd love not to have to carry a bunch of gadgets around with me, but then I got to thinking about it. Some things just don't work well when integrated. For example - a cell phone pda combo. If you're talking on the phone, you can't look at the calendar, or addresss book easily at all (at least not in any of the setups I've seen), so I'll carry both a phone and PDA if I have too.
As far as home appliances go, I don't see an all-in-one box their either. The set-top internet boxes for example sound great in theory (internet browsing from the comforts of your couch), but they have a lot of problems yet (IMHO) (ie not supporting javascript/java, low resolutions due to TV, etc..). I dunno.. I guess they're young yet, so those problems may be ironed out.
Howzabout that? Linus gets quoted on the upcoming devalutation of software the same afternoon that Ballmer admits that Microsoft stock is overpriced. Market tumbles.
Lacking <sarcasm> tags,
So you want two Net-TV devices instead of a heap of networked appliances? Maybe your paradigm is still locked onto the Intel/Microsoft "we will build what you need; next time for sure" mentality.. Only (near) monopolies, and in very mass production can afford to build general purpose devices, appropriate for many use.. It seems to me, PC's really good for Linux are still cheaper than a TV I really would love :)
Maybe the "multimedia convergence" finally really turns into divergence later on, and we will have specific devices much better suited to the task as a PC with a limited output (monitor, stereo speakers), and even more limited input (mouse, 10x key keyboard). Maybe the geeks of the future still will have a central remote control, which could well be more powerful than a current server, but still have his house populated by many more devices. After all, you can't have a single device for two people, it simply isn't fun to watch a PDA style device with another people in the same room. Why not have videowalls, remote controls, status displays, toys all around in the place?
IMHO it will be much easier to replace a faulty videowall, or control handset of all the wirelessly interconnected future room than to call a technician to fix the net-tv-holophone-security system, and find out that you can't open the fron door without the central unit :)
Remember, you can't predict paradigm shifts when you abhor change. But change will come :).
"Ten years from now, they could do it in a few seconds." -- The Racketeer of the Hellfire Club, 1993, Phrack 42
I work for a company that works with embedded systems, and it's surprising how much some of the embedded operating systems cost. QNX is damn-near unreal. They do have a good product, but I can't afford to purchase the damn development kit to even start working on a design, let alone make and sell the 50 thousand or so units needed to pay off the developers toolkit.
:) in server and workstation distros was absolutely correct. I think there should be another one for embedded systems.
I think that the embedded Linux operating system distros are the future for Linux. The article posted here a while back suggesting that Linux be sold (or downloaded
The major push for IPv6 is to give damn near everything (lights, toasters, microwaves, etc) an IP address, and if Linux gets a robust IPv6 stack and a little marketing, it will be in everything, simply because the manufacturers of those products don't have to pay any license fees for it.
I'm partly with you, partly with Linus there. Eventually the Unix model of small programs for special purposes joined together for all needs will reach appliances too.
We'll have separate units for
One should be able to combine those in appropriate ways to get what you want, from mobile to wall-planted. Once those devices recognize each other and communicate, funny possibilities should emerge.
I think, therefore thoughts exist. Ego is just an impression.
*Free* is Linux claim to marketshare. Linus is going to have to obsolete old solutions with something other than price. Many, many mortals before Linus have predicted the movement of markets... and were wrong.
Steve Jobs ~1993 predicted monolithic applications would move to component paradigms. Users simply load features needed, when they need it. Rulers, dictionaries, editors, etc... are simply an *install* away. Bad mesenger, you're right.
Dependancy upon web-based apps is not going to happen in the business sector. Productivity apps stay desktop bound. Sure prices could crash but why? Intuit is an excellent example of a marginal application upgrade with added new functionality. These apps are now cross-marketing vehicles more than productivity tools. Smart money has these *old* apps repurposed.
Embedded? SmartQuicken? People are going to buy into smartcards, BIG! It doesn't hold that *free* can crash the rest of capitalism.
The goals outlined in the GNU Manifesto are being reached. It takes time to change the world, but it would be done and is being done.
Free Software: the software by the people, of the people and for the people. Develop! Share! Enhance! Enjoy!
How billion dollar software houses are there?
(answer few)
How many billion dollar hardware houses are there?
(answer many)
Now, how many of the billion dollar hardware companies do not add value to their product with software?
(answer few)
Software is the big business.....most of them add some hardware to the software to add value.
So it is a NO BRAINER about how the future is embedded processors. (Like the 'market' for the Transmeta product) Writing for embedded environments is VERY BIG business.
Now, the embedded market is ripe with IP (Intellectual Property) With GNU/Linux you run the risk of having your IP released. Or, a torrent of flamage , like that heaped upon Corel.
Linus IS correct about the embedded market, but wrong about how GNU/Lunix will lead the charge. The BSD license is better suited for the embedded task.
If it was said on slashdot, it MUST be true!
The same reason people keep coffeemakers next to their stoves - stoves make coffee, too, but coffeemakers make it better, faster, and with a more appropriate UI.
But all of those are things that can be overcome. No one I know keeps a manual typewriter next to their computer. A lot of people with good internet access don't keep cd players, cassette players, or radios next to their computers. I no longer keep a newspaper in my house - it has been completely integrated with my computer. I know longer keep a TV guide next to my television - even though the TV guide has more information and a better UI than my digital cable, it's easier to find the TV than the TV AND a current periodical.
By contrast, I don't watch TV on my computer - that's too frustrating and expensive. I still keep maps in my car - again, the alternative is too expensive. I don't send e-mail through my cellphone, and even though I wish I could, it wouldn't be the only way I sent e-mail.
In summary - my girlfriend uses a knife to cut food, and scissors to cut paper and cloth, because she often has to do things requiring the precision scissors afford. I will use a knife to cut anything, because it's easier than looking for scissors, and I'm a lazy sod. There are many people who don't want all-in-one devices, but a lot of them will use them once they have them.
-=Best Viewed Using [INLINE]=-
Software price crash? I'm not so sure.. Well, depends on how you look at it. My favorite example for why software companies should fear GNU/Linux is the GIMP. Bye bye Adobe Photoshop. That won't be a price crash (ugh.. hundreds of dollars to touch up images?), it will be a total switchover from one product to another. :) It depends on the software, I suppose.. But still, it would be more of a drop than a /crash/..
The only crash I might suspect would be in the hardware end of the computer industry. After all, the Wintel connection will be for naught if Microsoft bites the GNU/Linux bullet. I mean, if you don't constantly have OS upgrades that are totally monolithic and suck ever-increasing amounts of resources, what's the incentive to upgrade your hardware? Mind you, I'm talking about the desktop. Scientists and other people who require /extreme/ computing power will always need more.. Richard Stallman in particular finds less and less of a need to upgrade all the time as hardware gets more powerful and his computing needs stay around the same. Hardcore gamers might find fault in my theory, but not all end-users are hardcore gamers. :)
Basically, if you run an OS that actually bothers to manage memory efficiently (contraty to the belief of the Windows-lovers I've met in my time, that is /not/ one of them) you eventually get to a point where getting the latest badass processor just isn't that big a deal. Therefore, in order for, say, Intel, to get you to go buy one, it has to be just a /little/ less than $500 or so.
~ Kish
I agree with LT that software prices will go down, but I cant see how software prices can fail because of a reduced number of updates. The current update system is more like renting software. When you buy Windows, for example, you update it every 2-5 years, and pay for the development by doing this. Essentially this is renting. If there wouldnt be a need for updating, MS had to demand a much larger one-time-fee, so prices would go up.
The reason I can see for lower software prices is increased competition. Large cooperations are fighting for market share (see Netscape vs. IE and MS Office vs. Star Office), and increasingly Open Source software takes market share from commercial software. Someday the Open Source model will win, not only because of its lower prices, but also because it offers advantages like the possibility of custom modifications. Commercial software cannot beat this, its only chance is better marketing or throwing more programmers at a problem (resulting in products that are bloated and usually worse). In the long term this wont help, and software companies will get their revenues by support&training, doing custom modifications for customers or developing software that is so specialised that the open-source development is not economical.
Try http://www.techweb. com/printableArticle?doc_id=REU19990923S0001
-----
Free P2P Backup, Windows & Linux