| This 62 message thread spans 3 pages: < < 62 ( 1  3 ) > > || |
|Open Source will *eventually* rule the desktop.|
Evidence, do a news group search for "linux desktop" on a year by year basis.
1994 -> 9
1995 -> 38
1996 -> 100
1997 -> 136
1998 -> 696
1999 -> 1,110
2000 -> 1,430
2001 -> 3,620
2002 -> 3,660 (so far)
As you can see, Linux is gradually growing in 'buzz', this, of course, does not translate directly into user growth but it us a indicator of how much Linux is on the conscience of the users of the web.
Five years ago *nix was in the realm of the (excuse the phrase) "ultra geeks". Today, webmasters and SEOs commonly chose linux as their operating system of choice. What will it be five years from now? I think it is wide open.
I don't see the numbers atrophying, there is no way Linux is going to lose ground, IMO the question is, "how fast will it gain popularity?"
Littleman's Nostradamus like prediction of the evolution of open source usage:
Stage I ->Hardcore programmer
Stage II ->CS majors, adventurous Webmasters, the curious
Stage III ->Academia, wannabe hippies types, government, many web professionals
Stage IV -> A full on horse race between Open Source and Closed Source software for market percentage
Stage V -> Closed Source lost 90% of the mass market software space
Right now, we are in between stage two and stage three.
My prediction is that we will be at stage V by 2017, if you step back, you could see it in motion now.
Good point. At one lawfirm with which I am familliar, the desktop is currently what is holding them back from using Free server software. They can't run Exchange on Linux. Period. And don't even think about training the sheep to use something that isn't Outlook. They're working on teaching them not to use the "recycle bin" as a permanent storage area. (I kid you not. I once got in rather hot water and had to be extricated with backup tapes for allowing Outlook to empty the recycle bin when I closed the program. It had been prompting him to delete all that stuff every time he closed the program for who knows how long, and he was still deliberately using it for long-term storage of imporant documents!)(I *did* get extricated. The parting was amicable and happened for reasons of the school year starting again.)
|It may be running some "server" type software (Of course I have Apache on my development workstation, it's part of my server-side deveopment environment.) |
I do not have this type of server software on my machine for several reasons.
1. I have a Development/Quality Assurance, and a Production Environment - two physically seperate servers.
2. I am a resource pig! I cannot stand any background services running eating up my memory or processor - hence slowing down my development.
3. I have had too many times that I have locked up a Development Server, I would hate to have to reboot my workstation.
4. If for some reason my Production machine needs to be serviced or needs some downtime I am able to use the Quality Assurance box temporarily until the Production machine is ready to go.
This may be why I consider a "Workstation" and a "Desktop" one of the same, or so closely related that it does not make too much difference to call them that way. As well all of my design/engineering software will run on my wifes, boss's, parents machines without a problem (Maybe a little slower in some cases).
I don't mean to beat a dead horse, I just like to share my point of view.
One thing I'll add to the discussion.
I hope I'm wrong on this, but I strongly suspect Microsoft and friends will launch a series of vicious attacks on open source soon using software patents as their weapon to scare companies away from open souce -- Linux in particular. I think this will be aimed at Linux specifically, while targeting open souce in general.
I suspect besides waiting till the antitrust trial was over before playing the "heavy", Microsoft is also waiting for their enemy Sun Microsystems to launch their Linux on the desktop efforts before springing into action with software patent lawsuit attacks and cease and desist order tactics to scare businesses away from using Linux. I think MS will use this as leverage in their other legal battles with Sun -- thus the wait until Sun allegedly violates the software patents Linux may be guilty of infringing on by distributing their own Linux distro.
Of course, MS may also be waiting to aquire a few more select companies...along with their software patent portfolios...to insure an even longer term anti-Linux FUD software patent attack campaign, too. Lets say they attack using one or two patents. Linux cleans up. Then they launch another. Linux cleans up. Repeat until patent infringement portfolio is used up. Several years of this with a few strategic lawsuits against buisnesses USING linux and thus violating software patents would be not only quite effective, but also in keeping with MS's nature, IMO.
Again, I hope I'm wrong, as I'm a fan of open source and am now learning FreeBSD, which I suspect may not be a target of such suits. However, I think the software patent attack on Linux WILL happen and will be devastating to it's uptake by business and it's growing mindshare as a Microsoft alternative -- server or desktop.
If business is scared away from Linux use and distro makers like Red Hat, SUSE, Mandrake and others are sued into oblivion or at least a pale shadow of their former selves, the Linux juggernaut will be derailed. It will no doubt make a reappearance back on the train tracks, but only after much travail from time and machinations required to remove and replace the offending code that alledgedly violates software patents. Not only will this take time, but the lawsuits will likely take even longer to resolve -- the longer, nastier and more costly the better for Microsoft and friends, I suspect; all the better to maximize the anti-Linux FUD.
The *BSD will likely take the lead in growth and mindshare if this does play out as I suspect. Moreover, it's important to keep in mind that Linux is NOT the only viable open source OS out there.
One thing for sure, it will most certianly be a geek and business soap opera of epic proportions in the online "press" lasting a very long time. Unfortunately, lots of theater detrimental to open source in the minds of average people, too, due to the likely behavior and response from some Linux zealots/supporters.
Again, I'm not in favor of this happening and hope I'm wrong -- but I think I'm right on target. Time will tell.
Another release regarding Linux and marketshare:
I can't really see the stand alone version of Lindows 3.0 selling very many copies @ 109 bucks when you can get a PC, (albeit a weak PC) thrown in for another 90 bucks.
I think the guy missed the point though when he asked, "why would linux shoppers be at Wal-Mart"? Well, hmm, Windows shoppers (no pun intended) will be at Wal-Mart, and isn't that the target market? Surely they don't think Lindows will appeal to hard core Linux users...
|Surely they don't think Lindows will appeal to hard core Linux users... |
Eep! A distro that defaults to logging you in as root all the time? I think not ;) I doubt the author of that article would know what "root" was, much less why you shouldn't use it all the time.
People used to Windows 9x/ME, though, probably won't be scared off by that.
As for boxed Lindows, why would you buy it? The pre-installed market seems like a much better place for them to focus. Personally, I think maintstream availability of pre-installed Linux machines is an important step on the way to desktop adoption. As I've seen pointed out many times, most people think Windows is easy to install *because they've never had to do it*. (NT 4 installation makes FreeBSD look easy, and every Linux install I've ever done was easier than FreeBSD.)
Whatever claims have been made about the big manufacturors not being able to sucessfully market Linux offerings, I don't buy it. Every time there has been some big announcement that a major vendor was now offering pre-installed Linux systems, I've gone to their web site and tried to find any option of a desktop machine loaded with Linux anywhere on the site. I have always failed. If I can't find it even when I'm specifically looking, it doesn't count as an attempt to sell.
As for patents. I can only hope it doesn't go that way. There's a very small movement under way now to create a body of patents under the control of open-source people as a weapon for counter-strike, but I doubt it will work. Too little too late.
Sorry, littleman, I think you're way off base here.
In a utopia, you're right, *nix might overtake Windows in time. In the real world, it won't happen. Why? Because it's not the superior product that dominates, it's the company with the best marketing department.
- Betamax is better than VHS, but how many people go to Blockbuster to rent Betamax tapes?
- Laser Disks were better too, but they came and went until the DVD finally arrived with Sony, Panasonic and the major Electronics makers backing them.
- OS/2 was far superior to Win 3.1 or Win 95, but it couldn't make a dent in the desktop market.
- Palm Pilots are kiddie toys compared to PocketPC's, yet they out-sell them 5 to 1
- Mac OS 10 is the "best of the best" *nix as far as useability is concerned, and it has the biggest company behind it. Yet they can't make more than a 3% dent in the market.
History is full of examples. Great marketing will beat great technology every time. It's only when you have the best combination of both that your product will succeed.
The various *nix distributors (with the exception of perhaps Sun) are all very small, very new, and very inexperienced in the business world. They just can't compete with Microsoft on a marketing basis. Until they can, they won't be able to secure deals with companies like Dell, Compaq, Gateway and Sony to distribute Linux based PC's on an equal footing with Windows.
Microsoft is currently running on a huge profit margin.
With time, Windows descendants will get better and at the same time Microsoft will be forced to drop prices.
Don't discard them yet.
With Open Source you don't have to reinvent the wheel so things can only go better, it's really a matter of time.
See Shoulders of Giants -- A Paper on the Inevitability of Open Source Dominance [cyber.com.au] for more detailed discussion.
Aside the way the system clock is handled, the only real difference between Windows and most Unix varients is the brand name. Both OS categories are based on C, and run more or less the same way to do essentially the same things.
I just finished a project where I had to write an application that ran on both windows and linux. There was a difference of three lines between the two implementations.
Will linux ever take off on the desktop?
The vast majority of desktop users neither know nor care about what OS or hardware platform they are running. They just want easy with minimal crashing.
For an idea of the demographic, think of people that think your speaking Jargon when you use simple terms like "reboot" or "icon."
While Linux may be cheap, even the "easy" distros are A LOT more difficult to set up than windows.
Give it time, and throw some money at it, and the people working on linux distros may eventually catch on.
>Both OS categories are based on C, and run more or less the same way to do essentially the same things.
The problem is that one of those 2 OS's fails to do what it has to.
"The problem is that one of those 2 OS's fails to do what it has to."
An operating system has to give applications access to whatever resources they have rights to access, it has to make applications behave and terminate them if they don't and only if they don't (not because they were written by the competition).
Windows fails to meet these requirements and a few others.
Something that's bad in the apps (not the OS) is that a lot of people write the same code instead of reusing what's already there - this leads to more errors and slower response to bug reports.
"Give it time, and throw some money at it, and the people working on linux distros may eventually catch on."
Linux is just another unix variant and over the years there have been lots of them. SCO (Santa Cruz Operations?) was a well respected name of another highly regarded version. It was the Linux of its day. Anyone remember Xenix, another Linus of its day?
Newcomers see Linux as a new piece of software that is contending with Windows. I see Linux as a new variant of a product that has had dozens of variants. When Windows was the new OS, unix was already well established. The battle was fought for the desktop and Windows won. Anyone remember how badly X-Windows and Motif sucked compared to the early versions of Windows?
Yes, by being free, open source coders can successfully copy someones else's proven ideas for windows applications and file formats. Free does work.
We all are aware of pirated code is being run on a large chunk of PC's. The attraction is that pirated code is free. If this works for the pirates, there is no reason to believe it won't work for Linux. (Again only comparing on the cost of the application. Linux and open source programs are not pirated programs.)
A very likely scenerio is that a good distro will emerge with a clone of the most used parts of Word and Excel and a workable Browser. This can be successful on cheap machines. I see this timeframe as 2004 but it could happen sooner.
But don't be fooled, successful Linux and open source will rely on people making money on it. So how free is free? If the Total Cost of Ownership of a Linux desktop is truly less than a Windows desktop, Linux will gain some marketshare.
How long until Linux rules the desktop? Well, don't delay any purchases for critical application waiting for that to happen.
|The battle was fought for the desktop and Windows won |
You are forgetting to say against who and how was the battle for the desktop won. Your redaction style suggests that it was against some variety of *nix.
In 1987 windows 2.0 is announced, and IBM has OS/2. Where are the *nix then? IBM has AIX, AT&T buys Sun, and NeXT has NEXTSTEP. The only one of them even near to be a desktop system is NextSTEP. More on it later.
On that point the desktop battle was absolutely not Microsoft vs. Unixland, because the *nixes were not designed to be desktops, with the exception of NEXTSTEP, that eventually became the infrastructure where Mac built on.
Get this clear: *nix never lost a desktop battle because it never fought. OS/2 did. And NEXTSTEP was way better that the Motif you want to shun. Why wasn't it as popular as Windows or even as OS/2? Hardware issues. NeXT's hardware division was never very sucessful, and eventually bited the dust. NEXTSTEP continued as OPENSTEP, but it was simply not very popular without the black box. Until it became MacOS, it is.
For a sample of NeXT's interface just look at WindowMaker at any *nix box.
Resuming, would you care to elaborate on exactly when did *nix lost the desktop war? That is simply not clear to me.
Your using as a point of referrence Window's 2.0. I am looking at the world even before Window's 1.0. The players were IBM on the mainframe, AT&T owning unix, Apple, Radio Shack and CP/M making up the PC world.
I went down to Xerox at some point and saw the Lisa graphical interface somewhere around the time IBM introduced the PC with DOS. Microsoft was a smaller player until that time.
With the assistance of IBM, Microsoft developed Windows. IBM contributed its knowledge of virtual paging. Having experience with the MIT developed VM (virtual machine)mainframe OS. Years later IBM would take the Windows code after they stopped joint development and clone it into OS/2.
Apple would get to market with its graphical interface built on the concepts from Xerox. Microsoft and other began development of graphical interfaces.
IBM was the dominant player in the computer field at the time so much so that they fell afoul of antitrust laws. AT&T the unix owner, dominated communication so much so that they also fell afoul of the antitrust laws. So real giants were competing in the market place. Microsoft beat IBM, AT&T who put unix in the public domain, Xerox that had the first graphical interface.
They beat all these big guys when they were small. They have grown and beat other big guys too. Novell once owned networking but gave away the marketplace by insisting on ISP/SPX rather than IP.
There was a time that Microsoft was as small as Red Hat or the other Linux Distributors. The won marketshare with a superior product.
Let's see what happens with Linux? It's major problem is it is a 'me-too' product relying on the open source movement to develop 'me-too' products. All the geek-speak in the world does change the marketplace. People want Windows on their desktop. It works consistenly. It is afforable and reliable.
It's biggest hurdle commercially is that it takes a cost and has a risk of failure to convert to Linux desktop. When you get there all you get is a 'me-too' product.
This is seen as having no up-side and possible down-side. Why would any career minded manager do this?
Please note that I'm talking Linux desktop. That's the only subject of this thread. Unix and the Linux variant have a whole bunch of ways of competing in the server market.
"Resuming, would you care to elaborate on exactly when did *nix lost the desktop war? That is simply not clear to me."
I really can't give you a date. It was around the time that Window's was running on 90% of the desktops.
>People want Windows on their desktop. It works consistenly. It is afforable and reliable.
I wouldn't agree that Windows wokrs consistently and reliably, I probably wouldn't have hurried so much to try out Linux if it did.
Anyone remember Xenix, another Linus of its day?
If I remember correctly, then Xeinx was a Microsoft product, and they failed miserably with it. I'm not sure if I quite see the analogy to Linux here just yet...
Anyone remember how badly X-Windows and Motif sucked compared to the early versions of Windows?
Actually, I remember the opposite, which may be a matter of style and what you were used working with. Motif died in the trenches of corporate politics even before the first useable Windows version (3.11) came out. But in my experience, all the early X desktops (sunview, openlook, Motif, even plain X) easily ran rings around the early Windows versions. Their main problem was that they only did so on horribly expensive hardware. Those were technical systems, not "desktops" in todays meaning of the term. Saying they lost the desktop wars is like saying that 16 wheeler trucks have lost the formula 1 circus.
Will Linux eventually rule the "clueless computer user"s desktop? Honestly, I couldn't care less, as long as it does its reliable job on mine...
When you get there all you get is a 'me-too' product.
I think that is where you have a misconception, The Open Source phenomenon is not just a product it is a decentralized social and geo-economic force.
|saw the Lisa graphical interface somewhere around the time IBM introduced the PC with DOS |
Are we talking about "graphical interface" or about "desktop"?
The denomination of desktop doesn't come simply from exporting a display to make it available for the applications. If it were Linux would claim superiority right now. The X windowing system is way better than the Windows GDI. That is no big achievement, of course.
"Desktop" for the most of the people means a "canned" environment where anything can be done on the abstraction of converting the computer's interface on a almost tangible object. Cyril's lenghty rant fails to even skirm over that concept.
I'll say that again: Microsoft has not ever had real competition for the *desktop* out of OS/2. NeXT didn't go anywhere. Macintosh doesn't really care as long as can sell its luxury hardware, that will never go mainstream because that's not in the interest of the company.
Linux has a real chance as long as someone cares; I by myself don't care, I am happy with my barebones sawfish and half a dozen of virtual terminals. But lots of people really care. Don't understimate them.
|The vast majority of desktop users neither know nor care about what OS or hardware platform they are running. They just want easy with minimal crashing. |
Sorry, but I think you're way off base here. Brand name is a huge factor.
Take the average person off the street, and ask them if they want a car from Ford, or a car from Widget Co. They'll take the Ford everytime.
Is it because Ford makes the best cars on the planet? No. It's because they've never heard of Widget Co., and they'd rather take their chances with a "brand name" than take a risk with an unknown.
Every user has heard of Microsoft -- even people without a PC. Very few have every heard of RedHat, Debian, or any other *nix variant.
Again, it's Marketing, not Technology that *nix has to win the battle with. They just don't have the cash....
Yet, if Widget Brand Co. has an outstanding reputation and costs 1/15th the cost of Ford then people will chose Widget Brand.
And if Ford spent several million dollars a year spreading FUD about Widget Brand it would only slow down the transition -- especially if with Widget Brand you got free insurance, gas and upgrades.
>>Yet, if Widget Brand Co. has an outstanding reputation and costs 1/15th the cost of Ford then people will chose Widget Brand.
Maybe, but if you're applying that analogy to the Open Source argument it doesn't fly. Open Source on the desktop has the reputation of being a pain in the arse, confusing, time consuming and certainly doesn't garner the "outstanding reputation" adjective among Windows users.
It's still about marketing and public perception. Free doesn't equate to "good" or "reputable". In fact, free is often looked upon with suspicion. Even "cheap" and "less expensive" products need to have the marketing behind them to succeed.
Many open source porjects are better than their commercial counterparts and in time this will be even more the case.
Time is on Open Source's side.
|Many open source porjects are better than their commercial counterparts and in time this will be even more the case. |
As an example, I offer Evolution. It's hard to claim that Evolution didn't start off as a straight-up Outlook knockoff. Windows users had a nice integrated PIM solution that, despite its many flaws, was much more usable than anything with similar power. I think that was the case regardless of what platform you were using. I actually used to keep two machines at one of my places of work, one running Windows for Outlook and one running Linux for everything else. Outlook is buggy and full of security holes, but it's pretty easy to use. We wanted something like that, and while Helix Code / Ximian never, to my knowledge, explicitly said so, it was clear from the get-go that Evolution was supposed to be Outlook for *nix.
I've now been using Evolution as my main mail client and PIM for some time, and quite like it. Durring that same time, my father has continued using Outlook. We talk quite a bit, and lately I've been noticing a rising count of features Evolution has but Outlook doesn't. It was even more acute a few months ago when I started a job at which I have to use a Windows machine most of the time, and configured Outlook to use the same accounts as my Evolution installation at home. I was often trying to do exactly thesame things in each of them and finding that the Evolution way was much more straightforward or even that I just couldn't do that in Outlook. It happened so much, in fact, that I've given up on Outlook and now use VNC to connect to a machine where I can run Evolution.
More recently, I've upgraded from Evolution 1.0.8 to 1.2, and things are even better. A few menus have moved, and I've found them by just seeing them there when they were apropriate, rather than deciding that I wanted to do something and having to search. (Try *that* with Outlook settings!) Some new features exists that I discovered the same way. There was a well-labled widget in an obvious location that did something useful in exactly the way I expected. It may have started as an Outlook clone, but now I find it much better on all the fronts where I used to give Outlook credit, and still better on all the fronts where I criticise Outlook.
Some other examples: I think Galeon is better than any other browser I've ever used, though 07 comes close. I wouldn't do without some flavor of Emacs - as long as you're not using it in a terminal window, it's sufficiently guified that the initial learning curve is no greater than notepad, but it has so much more power. More to learn if you want to, of course, but you can take that at any rate you like.
For me, those are the three apps I need every day without fail, and all three have open-source versions that I like better than any closed-source alternative I know of. Browser, PIM, and text editor. I use a word processor not more than once a month, and spreadsheets less often than that, so I can't really comment on them. "Presentation" apps like PowePoint don't have a place in my life. If I need that, I'll do it in XHTML, CSS, and The GIMP.
"Cyril's lenghty rant fails to even skirm over that concept"
I do run on, don't I? Sorry.
I hope that I've gotten across the point that as a server product Linux is likely to emerge as the dominant unix variant. If Linux does dominate, it will make for much stiffer competition in the marketplace and may begin to regain lost marketshare.
I am more doubtful of Linux on the desktop in the near term.
``But we've still got a few years to go.''
Reading that article made think about the real costs of operating Linux vs Windows. Seindal on msg #14 [webmasterworld.com] said there is a "Lack of hardware support."
If you run Linux, you will need hardware that provide Linux drivers. With the exception of network cards, one would have to buy brand name video and sound hardware to work under Linux. While I could spend all of $10 to buy a decent video and sound card. It would cost more for Linux-compatible hardware.
|one would have to buy brand name video and sound hardware to work under Linux. |
Nah. Just have to know what chipset the card really uses, so you can use the right driver for what it really is. Not that the self-important know-it-all at Office Max will be any use for that, but you can figure it out with a little bit of research on the net. It may change which cheap generic card you buy, but usually not how much you spend on it.
I've put Linux on all kinds of PCs, including Emachines. and junk HP. In all cases I've been able to get everything working sans the WINmodems.
On my desktop I have sound, CD RW, a scanner, printer and all the other things I need to get my work done. In some respects you are right, I did have to get a higher quality Epson scanner which cost me $150. But my CD RW is a generic brand that set me back all of $19.
I'm amazed people are getting in such a huffy about such vague terms.
Open source doesn't neccessarily mean linux. I use plenty of open source software such as mozilla, 7zip, apache and others on my windows 2000 based system with no problems.
I don't think anyone's defined what desktop is supposed to mean. There is nothing wrong with using open source software. And there is plenty of quality open source software out there for most any operating system's "desktop" that you can think of.
At the rate OSS has been going in this regard, I could see how the argument can be made that OSS could overtake the "desktop."
|I don't think anyone's defined what desktop is supposed to mean |
|"Desktop" for the most of the people means a "canned" environment where anything can be done on the abstraction of converting the computer's interface on a almost tangible object. |
The key on that is the abstraction. A 'computer' can be or do just anything. But for the common of mortals a computer should not be a number-muncher; they need something where pulling a lever or switching a knob does what is expected.
That people expects to buy a computer and use it just like if it were a physical device. Just like a desktop. If a document is inside a drawer on the desktop the physical abstraction is to pull the drawer, take the document out and read it.
Contrast: on a desktop you have an icon representing the desktop; go inside the icon, there are several drawers; open the drawer you want, locate the document and read it. On a more "to the metal" abstraction do 'less /path/to/file'.
On a computer the desktop abstraction is not really needed; these are superfluous steps, not the optimal ones. But when people expects their computer to work like physical, not logical, entities, they demand an abstraction like the "desktop" we are discussing. For them it's easier.
[edited by: Duckula at 10:38 pm (utc) on Nov. 29, 2002]
| This 62 message thread spans 3 pages: < < 62 ( 1  3 ) > > |