Slashdot Mirror
The Near-Term Future Of Open Source Desktops
securitas writes "eWEEK has two related articles on the growth of open source software. The first article is about the growth of desktop Linux, featuring Lotus and the Open Source Applications Foundation (OSAF) founder Mitch Kapor, who says (among other things) that call centers will be where the next wave of growth for desktop Linux happens and that 10 percent of global desktops will be Linux in a few years. He bases his statements on a report by Eazel and GNOME Foundation co-founder Bart Decrem entitled 'Desktop Linux Technology and Market Overview' (PDF) mentioned last week. The second story is about open source software growth in the government sector where government agencies like the U.S. Census Bureau have embraced OS software for projects like the State and County QuickFacts site. Based on Perl, Apache, MySQL and Linux, the site gets 200,000 page views a day."
Call centers just need something for their monkeys to use that works, and the cheaper, the better. Linux fits that nicely. Doesn't need to play the newest games, or run the newest Windows software, just deal with callers.
Once Linux is the main OS sold with new computers and Windows is the "optional extra" - then I'll regard it as a success. At the moment the market share of Microsoft means that most people know of one OS - Windows - and that is what they ask for with new computers....
Video Game cheats, hints a
Another place that could use this might be places like public libraries, where pretty much all you need is a working browser. Plus, a place like that could give some nice exposure to Linux.
and that 10 percent of global desktops will be Linux in a few years.
In a few years. We know the revolution is just round the corner. But how many corners do we have to revolve around?
"It's not your information. It's information about you" - John Ford, Vice President, Equifax
Based on Perl, Apache, MySQL and Linux, the site gets 200,000 page views a day.
:)
That's really cool... but in a different way. It makes me feel really proud of my Python, Apache, MySQL and (Linux|NetBSD) site (two locations) that gets 400,000 pageviews a day!
Enterprise features like layered transactions, replication, stored procs, load balancing, etc. are available using Postgres...but you can't find developers and cheap hosters that run Postgres anymore. Was it just the name "MySQL" that made it popular?
Just as long as my internet works, I don't care. Where I work, we use LINUX, Windows, and Mac OS X (the latter being used only on my machine). There is no liberation of the masses for the masses. They don't care. Two of us are LINUX fans, and the other 10 use windows because they just want to check their email. They don't want to ever touch anything in the command line, and I can't wholly blame them.
LINUX makes sense for the corporate IT infrastructure. The UNIX of old is expensive, and Windows is buggy and (also) expensive. As long as people can get sub-$600 PCs running Windows ME, they will buy them because they simply don't care. And their job and their life has nothing to do with computers other than that everything happens to need computers today. The end all is "If ain't broke, don't fix it." My computer checks my email. And lets me read slashdot.
on what the call center does. I'm sure it would work fine for dealing with customer accounts. However, many call centers are software support. It wouldn't make a lot of sense to put linux on a machine when you are supporting Windows or a Windows application. Credit card, insurance, and similiar industries probably could move to linux easily with the exception of required office applications. OpenOffice might be able to fill in that role though, and it would significantly reduce overall cost.
Mitch Kapor says it will be around 2007 it right in the article:
An "extra" that detracts... that's some newfangled mathematics you got there!
You see? You see? Your stupid minds! Stupid! Stupid!
... I would guess that very few call centers are computer or product support. Think of all the banks, government agencies, insurance companies, airlines etc. Pick any website on the internet that is not computer related, look for a phone number and call it. I bet you $1 that the person on the end is not sat alone in their office waiting for your call.
Although Linux is a good solution for call centre workstations that doesn't take into consideration corporate workstation standards. In a help desk type of environment, I would hazard a guess that most call centres are simply departments within a larger company. They probably don't have any say whatsoever in what kind of desktop OS - the internal corporate IT department does.
Except for today, when we line it up for a good Slashdotting.
To be more on topic, I wonder how much of this is chosen by the PHBs ("I've heard a lot about this Linux, maybe we should use it for this next big project.") and how much is chosen by the admins without PHB approval ("Well, we need this project up on a server, and we have this old PII-400 laying around, let's just throw Linux on it, fire up Apache and mod_perl and then take an early lunch."). I know that when I was in the Air Force, I saw the latter happen much more often than the former.
Karma: Marginal (mostly due to the border around the website)
What is this, some kind of new slimline case model?
Ok, Windows vs. GNU/Linux on the desktop is no new discussion, but here goes...:
I think that GNU/Linux in many ways are equal to or better than MS Windows considering apps. A recent GNOME or KDE provides a great working environment with good browsers, email apps, etc. etc.
A problem for GNU/Linux _from a joe user standpoint_ is the inherent security and multi-user nature of UNIX-like OS's. Windows has a history of insecurity, but that also means no hassle with passwords and the like.
This "hassle" and inherent security are of course Right Things, but Joe User just thinks it's annoying.
It would be great to get a serious effort to send "old" (meaning 200MHz and up) computers to third world countries, loaded with open source operating systems. Macs, PCs, whatever. The problem is that to send them all there would cost more than to just buy new ones from a local manufacturing plant :P Maybe we could load up a few shipping containers, weld 'em shut, and just drop them in the ocean. The countries where they wash up get the computers.
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
"the site gets 200,000 page views a day."
Why did they bother upgrading? Even 2000 server can handle that.
Near-term future of open source desktops?
They will remain shitty.
Computer manufacturers are sued or pressured into including onto their computers.
windows will die out eventually as the masses see the control and customazation power of linux, however in another note, so this will make sense to all those who don't quite understand, i call referance to the chewbacca defense!!!
About them Open Sourcers,
ain't they codin' warriors.
Writing under licenses,
coded by they lawyers.
Some code on a payroll,
but usually it's free.
BSD to GPL,
they code for all to see.
Them happy slappy Open Sourcers,
they rights shall not be lost.
Keeping libs bug free because
you just can't beat the cost.
How to be an Open Sourcer,
there's no hidden catch.
Find an open project,
'n submit an open patch.
A programmer is a machine for converting coffee into code.
Yes, ??? and Profit!!! apply to Microsoft only...
Linux is GREAT...M$/Microsloth/(any ol' abusive deragatory name for Microsoft) is EVIL!!!
Can I have my +5 Moderation now?
One that I helped set up had a series of applications talking to a printing package which central templates had been defined in.
People just entered the action on a screen, and the server sent a request to the printing package which printed a letter on a central printer.
All the applications ran through a browser. That company could move the call centre desktops to Linux very easily.
What cripples Linux is that major Internet-Service Providers (ISPs) like AOL do not provide a Linux client through which you can dialup a connection to AOL. Hence, the average consumer will reject a Linux desktop because she simply cannot get e-mail.
It's really that simple. Once all the ISPs provide a dialup client on Linux, then Linux will make a major dent into the desktop market. It is really that simple.
Unfortunately, we have a chicken-egg problem. ISPs will not provide Linux dialup clients because few consumers buy Linux desktops. As well, few consumers buy Linux desktops because they cannot obtain a Linux dialup client for their popular ISPs. And so goes the loop ...
Note: I like Linux, and I have a dual boot Linux/Windows machine at home. I've used UNIX, Linux, and Windows professionally (old school MacOS, too).
Windows has its share of troubles. The idealistic among us don't like Microsoft's market domination. The security-minded don't like the multitude of holes. But take both of those out of the picture, and you end up with a simple question: Is the Linux desktop experience, including applications, really significantly better than Windows in some quantifiable ways? In my personal experience, the OSS desktop environment developers have been playing a game of catch up with Microsoft. Sure, Microsoft didn't invent the GUI. We all know that. But it's not like Linux + KDE|GNOME is so much stunningly better than Windows that there's a reason to jump ship to it. At the same time, realize that there are many, many happy Windows users *and* developers. The anti-Microsoft angst is largely from a certain crowd. The end result is that this issue is largely a muddle. If you paint it to be a clear-cut battle, then it's not representative of reality.
Choice is good, yes, but realize that this choice already exists. Is beating Microsoft and getting everyone to use the Linux kernel a win for choice?
Mod him down, ladies and gents.
I'd rather you posted with an ID, so we know you're not just flaming. Anyone else backup the stats this guy is posting?
I'm contracting at a major corporation (one of the world's largest producers of wine). They are currently replacing hundreds of older machines, about 400Mhz, without brand new high end systems. All these older boxes are being donated to charity (public schools, I think). Unfortunately, the charities will want to run Windoze. They would barely get by with win98. If they try to run anything newer they won't be happy.
If they would just switch to Linux and run one of the "lighter" (e.g. not KDE 3) desktop managers, they would get much better performance and save a fortune on software licenses.
On an annoying note, today a 400Mhz dual processor system was tossed onto the charity pile. I want it. It would be a great replacement for my aging server, but they not only do not have a system for anyone to purchase old equipment, they actively discourage people from asking! That's just plain stupid. The school that ends up with that box probably won't even know what it is and will deploy it as all the others - with win98! ARRRRGGGHHH!
-- Will program for bandwidth
Future of Windows desktop in the near future ...........
Based on Perl, Apache, MySQL and Linux...
:-)
These are exactly where my job experience is. If only every website would standardize on this, I'd be employed forever.
Cisco based call centers don't do Linux at desk or at the server: Desktop: http://www.cisco.com/en/US/products/sw/custcosw/ps 427/products_data_sheet09186a0080091de7.html
(This is the desktop light version - the CTIOS version will let you write your own clients - which "could" work on a non MS box, but Cisco won't support it)
At the server side, there is a "light" version based off an IVR - IPCC express
http://www.cisco.com/en/US/products/sw/custcosw/ps 1846/index.html
and the enterprise strength IPCC Enterprise:
http://www.cisco.com/en/US/products/sw/custcosw/ps 1844/index.html
That would be progress, how?
This sounds like you are running a kernel that does not do dma on your ide drives. Try a newer kernel. Most linux distros have a 2.4.20 kernel. The ide support in 2.4.21 may be a beter choice if you like the idea of compiling your own kernel.
/dev/hde /dev/hde: /dev/hde /dev/hde: /dev/hde /dev/hde: /dev/hde /dev/hde:
root@somewhere RPMS]# hdparm -d 0
setting using_dma to 0 (off)
using_dma = 0 (off)
[root@somewhere RPMS]# hdparm -t
Timing buffered disk reads: 64 MB in 31.28 seconds = 2.05 MB/sec
[root@somewhere RPMS]# hdparm -d 1
setting using_dma to 1 (on)
using_dma = 1 (on)
[root@somewhere RPMS]# hdparm -t
Timing buffered disk reads: 64 MB in 1.75 seconds = 36.57 MB/sec
[root@somewhere RPMS]#
IANALBIPOOGL (I am not a Lawyer, but I play one on GrokLaw.)
The use of linux and thin clients has been very effective in rural Namibia. More than 250 schools have been connected and an infractstructure of free educational websites exists. Its a little sparse at the moment but improving. Access to decent education sites actually can help students and teachers who lack resources. With linux you don't need the latest and greatest hardware to enable the masses. I support these projects because education is the one way we can break the cycle of violence that engulfs Africa. The whole of Southern Africa is peaceful now and starting to develop (believe it or not!).
Information should be free!
Sometimes I think that everyone is going about this in the wrong way. Yes... the Linux desktop needs some work before my 70-year-old Aunt can use it... but the tools are all there. Its not like the desktop WON"T do what she wants, she just doesn't know how to make it work.
.WMV for video, .PPT for presentations and .DOC for documents.
.MPEG or .HTML. I create presentations in .PPT all the time (crossover office)... but I save them as .HTML. Same goes for just about any other office document.
For Linux to succeed on the desktop, I think two things need to happen... somebody like HP, IBM, or Dell needs to step up and sell systems that are pre-configured so that people don't have to mess with them. Just turn them on and away they go.
Secondly, its the DOCUMENTS. The world needs to start using something other than
The only reason MS has a stranglehold on the desktop is because people have been convinced they need to use those formats. Everytime I turn around I see a website or some CD that is forcing people to use these documents.
The next time you are creating a document or file format.... even if its using Windows... force yourself to use
If anyone sends me a proprietary document format, I ask them to please re-save it in a format that I can use and send it again. Nobody has ever refused yet.
Just a thought,
-Kevin
-K.
Because Windows has gotten worse (XP licensing bullshit) or because Linux has gotten better?
Honestly, I haven't seen the latter happen, and this is from someone who runs only Linux on his home computer and, when a new game comes out, waits a few months for a Native or Wine-based port.
From everything I've seen, XP is better than 2K if only it weren't for the licensing bullshit. A strategy to defeat Windows (and this assumes there is a think-tank working to defeat Windows, and I don't think this really exists) would have to involve licensing, and right now Microsoft is vulnerable. Linux is better because it's open, and free. Period. Don't make technical arguments, make licensing arguments.
Q: "Is Linux better than Windows?"
A: "Yes, but in ways that you'd have to be a sysadmin to really understand. In other ways, it's worse. There are defintely going to be tradeoffs, and you'll take some time getting on your feet again."
Q: "Then why should I switch?"
A: "Because technology freedom is more important than technology, in ways that matter to everyone, not just programmers and not just budget controllers. Everything in your computer should belong to you."
People are responsive to this kind of argument, but it has to be presented honestly.
It's rare that you're presented with a knob whose only two positions are Make History and Flee Your Glorious Destiny.
May take a lawsuit or some other public pressure.
The latest thin clients (with no moving parts) are Linux based.
The ones I've been playing with are from neoware. Flash based OS and everything either X or Citrix off a central server. This is the sort of thing that call centres are actually using now, along with some fairly large industrial corps like Lockheed Martin (who I have to deal with).
You're right, I shouldn't complain about "having" to teach the Microsoft Apps... I did volunteer knowing I would be teaching some Microsoft Apps. I just wish I had more freedom.
*Apu voice* Thank you for calling Gateway technical support. Come again!
You make an interesting point about marketing. MySQL could very well be the VHS of OS databases to PostgreSQL's Betamax. PostgreSQL clearly wins out when it comes to a mature, stable feature set (the one perhaps sole place where MySQL shines in comparison is in being lightweight and fast at performing SELECTs--and even there the difference is undetectable in most situations). However, MySQL enjoys a higher profile and "better marketing"
Does this mean that PostgreSQL is doomed to the same fate as Betamax? Hell no! Unlike the VHS/Betamax case we are NOT dealing with proprietary technologies. Plus we are talking software--physically it is a bunch of bytes in files on hard drives, disks and the ether that is the 'net. There are potentially little to no manufacturing and distribution costs involved, as there are with VCRs, tapes and even boxed, commercial/proprietary software. It is the nature of Free software to endure despite lack of "success" in terms of widespread usage or revenue generation. So long as at least the developers use the software it will live on and potentially resurface publicly.
MySQL might be backed by a corporation and their aim may be to make money with it, but the driving force of most Free software is NOT driven by its money making capability--it was created and is maintained because it fills a technological need. Linux grew out of a personal project/experiment by Linus Torvalds who wanted a truly free MINIX-like OS to fill his personal needs--which happened to match the needs of many others. PostgreSQL was the academic follow-up project to INGRES--it's purpose wasn't originally to be an Oracle-slayer--it was to teaach CS studens about relation database concepts. When the Postgres project ended, it was in "production" use at several dozen (a few hundred?) academic sites, and thus there was a need for it to be maintained (originally by former UC Berkley students, followed by enthusiastic users with a vested interest in furthering the project). It's actually very hard to kill projects like *BSD or PostgreSQL even if Linux and MySQL are more popular, because they STILL have a LOT of users who are very much interested in keeping them alive for technical (and even emotional) reasons. They don't have to make money or dominate the market because they didn't from the start and they never were intended to. Those projects were simply meant to be damn good software.
OTOH, if Microsoft develops or acquires a product it is chiefly for the purposes of making scads of money (Windows, Office) or to establish itself in new markets with the intent of eventual domination (XBOX). If it doesn't make money and it doesn't destroy competitors (or, in the case of MS's 1st foray into "tablet computing" as an answer to products from upstart GO comupting, it destoys the competition but subsequently fails to make money), MS drops it like a hot potato. To hell with the fans the product might have. And why bother releasing source or schematics? 99.9% of MS customers are too clueless to comprehend them, and those that have any comprehension might track down and pester our developers, who MS would much rather have working as long hours as possible on a harder-to-crack next-gen XBOX or Office 2004 or Longhorn or whatever. Thus unpopular closed products wither and die very rapidly.
Incidentally, it wasn't just bad marketing that killed Beta. Oddly enough, one major factor was that it was "too proprietary" (Sony alone controlled licensing the technology and got a bit greedy. VHS eventually became maintained by a consortium of manufacturers, and it was much less costly to produce "approved" VHS products). The cost of being so proprietary was passed onto consumers, giving VHS the edge. The other reasons were marketing and exposure (such as space on video store shelves), plus the fact that most consumers preferred the much longer duration of VHS tapes at the same or lower costs over the better picture quality of Beta. If that lesson is missed by Microsoft et al in their zeal to develop and promote DRM, "trusted computing" and so forth, no amount of marketing will prevent Linux from gaining ground.
Regarding call centers, I've come to the same conclusion over the past few months. Desktop Linux is a great solution.
What Linux needs as a killer-app in this environment is good VoIP support. By good, I mean cheaper than Windows.
Specifically, I'm thinking it would be feasible to add software echo-cancellation to some of the sound card drivers or as a separate module. That would easily shave another $50 off the price of a typical call center desktop, and probably more than that with the way people tend to break their $100 headsets.
Does anyone know if this is possible?
"I assumed blithely that there were no elves out there in the darkness"
Dvorak's latest on Linux (rejected as a story submission). I had been trying to figure out what might create a consumer rush to Linux, but for some reason, I had not considered the obvious: the development of the must-have critical application. This means an end-user killer application that runs on Linux only.
Oh GREAT any money from linux going to INDIA! That'll really move things along NOT! It's not going to get better, really. You know Linux can not be used by the averate idiot computer user, call center or not. I mean who is going to pay $50 or however much each time a problem, and there will be many, many, many, comes up in Linux on the desktop? Not many, that's who. Kapor needs to run it through his spreadsheet and see what comes out.
A million LAMP sites doesn't make one useable Linux desktop.
My dad took me over to his friend's house yesterday, because his friend was unsatisfied with his website. I got hired to redo it from scratch. After discussing how he wanted it, I somehow segued into open source software. My dad was complaining about his old computer and need for more storage space. I mentioned he could get a new computer off walmart.com for only $199. They were both shocked. I looked it up at wamart.com for them, and then the german (my dad's friend) pulled out his credit card and insisted I order one with his card to solve my dad's computer problems. He also wanted me to help him install linux on one of his computers, since he was frustrated he couldn't install windows XP on both (the install CD wouldn't let him). If he likes it he might install it on the other one, too. He kept saying "You can be free from Microsoft Windows??!!"
If open source is getting stronger in government, why is IIS about to overtake apache in the .gov domain? Ad why is IIS
still gaining market share in the .mil domain
having long since dominated the market?
It seems the government is what's propping up MS in the webserver area. Why isn't this ever mentioned?
Apple OS X has it all: kick ass applications, cutting edge software engineering, beautiful GUI with top-notch useability and the most stable and powerful Unix core available today. By comparison, Linux is amateur and nigh un-useable. Mac OS X users are growing by leaps and bounds, primarily from enlightened Windows XP and Linux users who are sick of playing "PC roulette" and have decided to move to the next plane of computing, where technology and superior people hang out! Baby!
In court! It's a monopoly, remember?
karma: Marianas Trench (mostly blub blub)
Having worked at two of those hellholes i can only confirm that they are cheap bastards doing anything to save a buck. Since they have so many workstations doing such simple things they are well suited to use linux. Only a handful of applications need rewriting. Funny thing, most systems i worked against was unix and even vax so using windows was in fact most "wrong".
HTTP/1.1 400
Okay, I loved KDE & Gnome in their first versions. They were better than the contemporary windows version. Recently I reinstalled both to see what had happened. MAN it is so slow it's unusable. I mean... windows is so much better than the current versions. What the heck happened? KDE & Gnome turned from being a light in the tunnel to being so slow and buggy that I really don't want to bother with it, and I LIKE to use Linux. I want linux to be a desktop competitor to windows.
I was imagining people saying
"I use Linux because I think KDE/Gnome is better than windows"
Now I mostly think it's
"I have to use KDE/Gnome because I need to use Linux"
What happened?? Why did they destroy the good start they had??
I thought I used KDE on the desktop. If you swap the linux kernel our from under that in favour of a BSD, Solaris, or heck, a Windows kernel, how would I even tell?
If you were blocking sigs, you wouldn't have to read this.
From the article at e-week -
putting pressure on Microsoft Corp Is this guy nuts or what?
Nathan Hanks, managing director of technology for Continental Airlines Inc., said his concern is making sure that he can turn the Houston company's airplanes around as quickly as possible. As such, the open-source-community concept is not as appealing to him. When the SQL Slammer worm hit earlier this year, Microsoft responded immediately and addressed the issue. Its executives also visited him to discuss the matter. This would not be possible in the open-source world, Hanks said.
Open Source allows you direct access to the developers not some suit in an anonymous department in Redmond.
Remember the SSL bug in IE5 and Konqueror? MS were still denying it was a problem weeks after the KDE team had patched the bug out. The slammer worm was also the result of another bit of crap coding.
For christ sake remind me not to fly Continental if I visit the states. If all their staff had their heads that far up their arses their pilots wouldn't be able to see where they're going and remember that 2k was built on NT and XP is built on 2k and Server 2003 probably has been in development since long before Microsoft's "Born again" security review. The software you are using is based on a 20 year legacy of piss-poor programming that will take a damn site more than a 3 month security training course to cure.
Hmmmmmm..... Deep fried and look like Squirrel.
Hell, I can get a PPP link up about 10 minutes after a fresh FreeBSD install. I just have to edit /etc/resolv.conf and /etc/ppp/ppp.conf, then edit /etc/rc.conf to enable dial-on-demand and NAT.
And if you don't get your rocks off hacking config files, use KPPP. 100% schmuck-proof, guaranteed.
No. Windows is what they're given. Honestly, when was the last time you walked into a CompUSA or Frys and actually heard someone asking for another OS?
You have no need to worry. The distributions you like most likely won't be favored by the masses because they're not really designed for the masses. Mandrake, RedHat,etc. will be more accepted by the masses for the very reasons you state.
http://quickfacts.census.gov/qfd/img/wordmark.gif
Um. Anyone have any ideas why the logo has some letters highlighted and others just plain black text?
USCENSUSBUREAU
Just curious/paranoid.
Karma: NaN
In The Innovator's Dilemma, the case is made by point out case after case for companies destroying themselves by listening to their customers.
All the mainframe companies did - their customers didn't want minicomputers.
All the minicomputer companies did - all their minicomputer customers rejected PCs as a solution.
And Microsoft must be listening to their customers and getting the same wrong information. But when you talk to customers who manage corporate computer infrastructures built on paying Microsoft millions of dollars, how many are going to say "Microsoft, make the next release free with optional support so that I can cut my budget from $20M to $5M."? Zero! Why? Well, a CIO saying that might as well be saying, "hey, give my boss justification for cutting my salary by 50%".
But why bother? I've not seen a computer (either Mac or PC) take 2 minutes to copy a 17 MB file unless something was seriously wrong with it or it was doing something horrible (e.g. virus scanning) in the background for a loooong time (the slowest machine I have around -- 450MHz Cube -- just copied 10MB in "about 5s").
I will grant you this: traditionally Macs tended to lag behind PCs in file operations but were considerably more responsive (my 2.4 GHz Dell desktop with 512MB RAM and XP can really lock up during what would seem to be pretty minor file operations). This seems to have been largely addressed as of OS X 10.1 (I don't tend to benchmark stuff so this is based on "feel"), but again Macs err on the side of responsiveness.
OTOH: (1) the Finder window that is the destination of a file operation does not lock up the way it does with Windows; (2) your Mac, in general, will be more responsive than a PC during file operations in Finder; (3) your Mac will generally provide some kind of indication of its progress through a file operation (versus telling you the progress in copying the current file -- utterly useless information -- as per Windows).
As for the remark I've never seen a Mac that has run faster than its Wintel counterpart let me observe that during Apple's darkest performance days (i.e. pre the PowerPC transition), I remember comparing the performance of HyperCard 2.0 with color XCMDs running on a (then) two year old 25MHz Quadra 700 (with 36MB RAM) to Visual Basic running on a brand new IBM 486 DX2/66 (with 64MB RAM) both running a multimedia project with 8-bit graphics -- the Quadra stomped the PC into the dust.
These days, raw performance is usually far less of an issue, and the fact that the Mac just works and the PC just doesn't work is the reason some people think the Macintosh is a superior machine.
My reason is this: PCs tend to crash a lot on the hardware (system errors) and on the users (bad user interface). Linux doesn't crash the hardware much but it really crashes the users a lot (terrible user interface). Old Macs crashed the hardware a lot (system errors) but hardly ever crashed the user (great user interface). OS X hardly ever crashes the hardware OR the user.
Crashing the user is FAR more expensive -- in general -- than crashing the hardware or running slow.