Slashdot Mirror
Recommend Apple, Lose Your Job?
rocketjam writes "While examining whether outsourcing tech work to India is really cost-effective, Robert X. Cringely takes a look at the old conspiracy theory that IT doesn't recommend Apple solutions because they need less support, thus endangering IT professionals' job security." Cringely argues: "Ideally, the IT department ought to recommend the best computer for the job, but more often than not, they recommend the best computer for the IT department's job."
as I mentioned in a response to Bob.
It's an issue of comfort.
Everyone is comfortable with windows, even if they don't like it.
Many admins are comfortable with Linux/Unix. It's what has gotten the job done for years.
I have used lots of different operating systems, CPM/TRSDOS/OS-2/VMS/Unix/Windows but have
NEVER used a Mac, so I'm not comfortable recommending it. I expect it to be very different
from the CLI world I'm used to.
In order for me to get comfortable, I'd have to play with it. If MacOS ran on PC hardware,
I would consider setting up a partition to boot it, but that's not the case. It's expensive
to learn, and I have no incentive.
Robert
--- If it's worth doing, it's worth doing in Perl!
Used to work as a contractor for a major pizza company that used to be in Kansas. They were an entire Macintosh place, had systems for about 700 - 1000 users I think it was.
Then a large soda company bought them and felt that 'they all needed to be the same' even though the Microsoft Offices the platforms ran worked together.
So, we went from the two of us supporting 700 - 1000 users to 18 people.
And the user populace was not happy. The standard rebuild time of a machine went from 'when they got new ones' to once a week. We had device driver issues, and SLAs of getting machines back up and running in two hours so we ended up just ghosting machines over and over to clear up whatever went wrong.
Weird.
As a rock-in-roll Physicist once said, No matter where you go, there you are.
I'm an applications admin. ClearCase and ClearQuest specifically. I also support a host of other engineering applications. Most of those apps were never and probably will never be ported to the Mac and I'm willing to bet that other engineering shops are in the same boat.
Sure, I figure that Macs might have a place in a business or accounting context but not for engineering. Anyone got a counter-example?
Blaze a trail to the New World
Every IT dept I've ever encountered was overburdened to start with. I can't imagine they would not want a break so they could attend to more important things than Windows crashes. Not only that, but the techs _hate_ dealing with this stuff.
Imagine how much money you could save your organization if you had the time to verify all backups and replace old, failure prone disk drives before they crash.
There is always more to do in IT.
End tutorial. Should all be familiar from there.
Sort of facetious, but, well, not really. Try it. Take a half hour out of your time. If it's not that easy, well, then you now know you're making the right decisions instead of wondering "if".
The CPU? Don't know yet. The unit is too new to need any upgrades yet.
NetInfo connection failed for server 127.0.0.1/local
Don't kid yourself. Technology choice is rarely, in any organisation, done for purely altruistic 'what's good for the company?' reasons.
Instead, the technology chosen is one of two choices:
1) What people are comfortable with. A lot of people want an easy, safe, predictable decision.
2) Resume fodder. What do the decision-makers want to add to their resumes? What's missing?
Analyse the average IT department's choices and one of those two is almost always the cause. Let's face it, most of us would also be guilty of these; picking what we're used to and what we think would be fun or useful to learn.
Apple hardware has had horrible design mistakes.. but no Wintel hardware ever has had hardware flaws that ended up biting people in the ass? And don't only look at the price of hardware. The workforce needed to run a Mac network as well as support costs are almost certain to be lower than a Linux network. If you only look at the price of hardware for purchasing decisions, then you are dumb.
Feel free to mod me "-1 - Angry Jerk".
Whatever the conventional wisdom or the Microsoft marketing message, Macs aren't dramatically more expensive to buy and on a Total Cost of Ownership basis they are probably cheaper.
Cool! Where are the numbers to support that? Probably isn't going to cut it.
Then a little later....
I am not claiming that every organization should throw out its PCs and replace them with Macs, but the numbers are pretty clear
You mean, those numbers that you didn't include? How are they clear? Once again, Probably isn't cutting it here.
Macs reduce IT head count while Linux probably increases IT head count, simple as that.
AS PROBABLY SIMPLE AS THAT!!
There's that probably word again! Ok, so it's obvious he's a Mac user. I'd probably take him a lot more seriously if there were a lot less probablies and a lot more proof and information. I'm PROBABLY going to stick with Linux for my IT needs for now.
For every annoying gentoo user, are three even more annoying anti-gentoo crybabies. Take Yosh from #Gimp for example.
Know the whole world knows! All of us IT types are really all sitting around in a room filled with exotic dancers, pool tables, video games, and food. We don't really do anything. And it all gets so boring that we go LOOKING for systems to give us more work to do!
In reality, 99% of the IT people that I know would practically sell their left arm to have systems which required less of their time.
steve
Oh, you're not stuck, you're just unable to let go of the onion rings.
Let us be realistic for a moment. How often is one going to do anything to a rackmount other than add ram and perhaps add more/larger drives? Rackmounts are kidna meant to live on the rack until they outlive their usefulness. Besides that, do you really think its a good idea to upgrade the CPU in a 1U rackmount unit? Lets consider. A rackmount unit really has to worry about airflow and cooling, A hotter CPU may cause unreliable performance. And speaking of unreliable. You dont buy a rackmount system to putz around with it. Often they are mission critical units. You wanna upgrade a cpu on a mission critical mail server for fun?
Assuming the OEM even makes a model that has a faster CPU and you can stick in yours, why didnt you buy the faster one anyway? By the time you woudl get around to 'upgrading' your rackmount the net gen technology would have already rolled out the door.
Two years of labor will cost you about $100,000 (for the sake of argument). For that you can buy 250 CPU upgrades.
People are so much more expensive than hardware that the whole argument about saving money with Linux is ludicrious if you can find another system, such as Macs running OS X, that can do all of the same things as Linux and reduce head count by even one or two people.
The xserve is decidedly NOT enterprise-grade hardware. It uses cheap, desktop-quality IDE hard drives and doesn't come with basic options like redundant power supplies and hot-plug PCI slots. Not to mention the sluggish G4 processors, the same ones as used in desktop machines. And the memory only goes up to 2GB. Excuse me, but that's not an enterprise server. That's a low-end server. Perfect for a small department server or maybe for hosting a small website.
Dell Optiplex GX270: Celeron 2.00GHz, 256MB, 80GB HDD, combo drive, 17" monitor, USB WiFi adapter, v.92 modem, XP Pro: $1,352 after $50 rebate.
Apple eMac: G4 1.00GHz, 256MB, 80GB HDD, DVD-R/CD-RW, 17" flat CRT monitor, Airport Extreme, v.92 modem, OS X 10.2: $1,398.
I guess it depends on your definition of "significantly."
And if you thought the Microsoft OS costs were bad, looks at Apple's. OS X launched in 2001, and, if you were a 10.0 buyer, while 10.1 was a free upgrade, 10.2 wasn't, and 10.3 is coming fast! And from the end user perspective, these have all been largely mandatory upgrades -- many apps now won't work unless you are running 10.2, for example.
Windows ME and Windows 2000 were released very close together, if not at the same time, yet you were expected to pay again to go from one to the other. Every machine sold until the release of 10.2 still could run OS 9, and there are plenty of applications available there.
2) Usability. While there are a lot of things that work smoothly under OS X, there are still some issues, ESPECIALLY with Windows interoperability -- and any company of size is going to have a significant overlap. So you'd have to train IT folks (or hire new ones), and still have some userland issues.
Actually, I've found OS X to be easier to integrate into a Windows network than even Windows 95/98. People at my company who come from OS 9 and Windows alike find it very easy to log on to servers use printers. If your users don't like it, OS X can be scripted onto servers just as easily as any other workstation.
The biggest plus is that you don't have to join a domain to access its resources. I had a Powerbook on a Windows-only network for 6 months. Not only was I able to log on to all of the Windows servers, I could administer them with Microsoft's terminal services client for the Mac, and still work through Outlook. Nobody had any idea that there was a Mac on the network--it was that compatible.
Another serious concern for IT has been how quickly Apple has outdated machines. Didn't we just see today that a number of machines aren't going to have proper functionality? Again, this is on fairly new machinery!
The people in my office who are still working away at their Beige G3's would probably disagree. I seem to remember the jump from the 286 to 386 to 486 caused the same issues (and complaints.) My 2000-vintage Pismo Powerbook was the machine I mentioned above. Not only was it able to be quite productive in a Windows-only environment, it has plenty of speed for what most people need it for.
Concerns have to be that Apple is quickly going to invalidate the G3 and G4 (over the next 24-30 months).
Why's that? I seriously doubt that they would shut out machines that are selling even now so soon. In fact, with the G5 becoming the new "high end" processor, it's likely that the G4 will become the new "low end." I expect eventual phase-out of the G3's because of new Altivec-ready applications released down the road, but those who need those applications will upgrade, and those who don't can continue to work with 10.2, or even OS 9.
Cost I'm willing to bet is more the case than some grand conspiracy by the IT world. Sure they cost less to support, and increase productivity, but when the initial outlay is more than a compatible x86 box, the people in charge of budgets start screaming. No one thinks about the saved cost in the future, it's all about now.
IE: Should I spend $3k (CND) to buy a swanky new apple powerbook (or more for a new g5?), or should I spend $1000 and upgrade my current x86 system to be a pretty kick ass gaming box, which can also act as a high powered linux server? Pretty easy choice if you see my bank account.
Sure in a year or so I'll want to upgrade again, or I'll have a MB or DIMM or hard drive go and will have to buy a new part, but that's ok. Because the cost is down the road, and therefor, doesn't exist.
Note: the last sentance was intended to be sarcasic or ironic, depending on your view.
Ten years ago, nothing beat appletalk from a user's prespective. Finding printers was easy and straight forward, as was finding servers. It was easy to share your work with others. And at the time, there was (and still really is no) equivilent in the Windows or Unix worlds.
/. users (preceived costs, lack of apps, unfamilarity, high cost of experiment, vendor lock, ...).
I've heard administrators horror stories of dealing with apple talk, but they don't seem to much worse than other horror stories. And even if they were the cause of a little more pain, isn't that the price of providing a good network solution to the users.
At my former company, when it was time to move away from appletalk, the network adminstrators jumped as fast as they could to replace it. But they didn't have horror stories of having to patch thousands of users computers, or bringing down entire networks as reasons for their JOY of seeing appletalk go away. Instead, it was that appletalk "slowed down the network".
So, we got a new network where we had to remember the IP address of any printer we wanted to use and any server we needed to access, and to share our work we had to tell everyone our IP address and hope that they wrote it down or you'd be telling them again, and again, and again... We went from a user centric network to a faster IT centric network.
Although I am a big Mac fan, I don't agree with Cringley on this issue. There are other reasons that Macs are being used in most businesses besides IT looking out for their own jobs. And most seem to be outlined here by
But whenever I think about the lose of AppleTalk, and now see it being blamed for Apple's shortcomings, I really wonder who IT thinks they work for. Always thought the user, but maybe not...
I am living proof of the Peter Principle
Who sold you that one, Bob? Did a bridge come with it?
"Flyin' in just a sweet place,
Never been known to fail..."
The XServe RAID has redundant power supplies, so it's likely the next XServe will too.
You might, from the fact that I quoted a line that said exactly that, conclude that I know that. I'm adding factors to the original TCO equation, not denying the factors that have already been mentioned.
Let's put a box together. The $799 eMac has an 800 MHz processor, 128 MB RAM, 40 GB hard disk, 32 MB video card, ethernet, firewire, modem, etc. A $599 Dell Dimension 2400 has a 2.2 GHz processor, 128 MB RAM, 80 GB hard disk, on-board video, and apparently no ethernet or firewire ports. The most important point here is that whether these two are equivalent computers depends on what you use it for. It should be clear that for a certain range of purposes, we can assume the two are equivalent.
Further assume that the eMac will be used for 3 years, and the Dell for 2 years. Finally, assume that the eMac will resell for $400, while the Dell will resell for $200. Doing the math, the Dell costs 55 cents a day over its life with you, while the eMac costs 37 cents. Therefore, the TCO of the eMac is actually less.
Now, note all the assumptions I made. The truth of the conclusion is dependent on the truth of the assumptions. Check the numbers out for yourself, because I just made up the usage years and resale value for this example.
C'mon. Is /. run on a single 1 GHz PC?
I don't think so.
So Apple has a "benchmarking" page:
http://www.apple.com/xserve/performance.html
Which shows Apache Web Serving performance -- where it's faster than a Dell 1650 (not sure if it's running IIS or Apache). Point is, even allowing for a little marketing hyperbole, OS X + Xserve is a fully capable web server.
Assuming of course you think Apache is up to the job?
Think for a second about how much the person using the computer makes. Multiply that by the company's personnel roster count. You really think an extra $1000 every two or three years matters if it keeps a $40k a year employee happily working during that time? A business trip costs more, or a weekend retreat. Heck, if it means that one or more IT people aren't needed, the cost could be absorbed. Or if 200 people lose 2 hours of work (virus or whatever) and make $20 an hour, that's $8000 right there...what happens if you lose a day or people make more? Fact is, hardware and software are *not* the most expensive thing with an organization. Labor is. So in perspective, $2500 isn't that much more than $1100. Not to mention that I see Macs stay in service longer than most PCs...
To all those who insist on putting the price of a PC up against the price of a "comparable" Mac...
Let's get something out of the way right up front: a Mac is a Mac and a PC is a PC. Sure, that's obvious, but it surprises me how little it's acknowledged in these kinds of discussions. $1500 worth of PC hardware won't give you a Mac no matter what you put on it. The same goes the other way; Mac hardware and software will never get you a PC no matter what combination you use. In the end, a PC is still a PC and a Mac is still a Mac. Play with numbers all you want, it won't change a thing. Folks who want a Mac will not be happy with anything but a Mac, not even a comparably spec'd out PC, period. The reverse is just as true.
Comparing Apple computers to PCs is like comparing Palm devices to ring-binder planner systems (nothing should be implied by the order in which those items were listed, by the way). Both serve similar purposes, and there are folks who use each who would never think of ditching their choice for the other. So would it be safe to say that all Palm users should ditch their Palm devices for ring-binder planners purely on the basis of a price tag? I think not. Palm users love their expensive Palm devices and binder planner users love their slightly less expensive binders, and neither is going to be wrong for sticking to their preferences.
I find the whole Mac vs. PC debate silly for the reasons described above. I use both, although I prefer my Macs to my PCs. That's just me. My wife loves her PCs and despises my Macs. Life goes on.
1) Pricing: Mac's are significantly more money. And if you thought the Microsoft OS costs were bad, looks at Apple's. OS X launched in 2001, and, if you were a 10.0 buyer, while 10.1 was a free upgrade, 10.2 wasn't, and 10.3 is coming fast! And from the end user perspective, these have all been largely mandatory upgrades -- many apps now won't work unless you are running 10.2, for example.
Now, your other points might be valid (though there are several companies that specialize in mac upgrades), but an unlimited license for OSX Server is what? $1000? Even at 10 times that, it pays for any hardware disparity with commodity PC parts with any significant installation.
How much is a 5000 seat WinXP contract? How many IT to support that, and how much do they make?
This argument makes the assumption that IT is ever properly staffed in the first place. IT people almost universally want to lessen their workload so it falls more in line with their actual [underfunded] workload capacity!
I think that was the point of the article. It's about the CIO not the staff. Headcount is king, and from what I've seen, it really is. If you need more IT to keep up all the time, you just keep getting all these people under you.
Resume-- Mangaged a 350 head IT department for bigass corporation...
is much more impressive than
Resume-- Managed a 5 person IT department for bigass corporation....
Well, since you seem to be gunning for a response from a Linux/x86 person for some reason, I'll bite.
Your argument is well taken, and in fact I think it is an excellent point to use when debating Windows vs. Mac as a desktop machine. Apple has a beautiful UI, can run MS Word for those who need it, and is built atop BSD so I would expect it to be very stable. The extra purchase price of a suitable Mac with OSX for daily desktop use and as a network client in an office setting is more than offset by the smaller number of tech support people required to maintain it, resulting in a net savings vs. Windows. I wish Apple would make OSX available on x86 so it can run on the legacy x86 hardware that most places already have. Apple's currently exhibiting their biggest market miscalculation since they priced themselves out of the personal computer market in the '80s, by not releasing an x86 version! Heck, Longhorn and DRM and Trusted Computing and Subscription-Pricing and all that is an open invitation for someone to come in and eat MS's cake. If the option existed on existing hardware, I'd recommend migration to it as a standard desktop ASAP. I might well be running it at home right now instead of Linux; I'd at least give it a try for sure.
Where I think your argument breaks down a bit is in the server room. OSX is beautiful, and all that, but there is no reason to be running a GUI on the server. When it's loaded up with connections and is busy being a server, then you start opening windows with those flashy effects, it's going to bog the whole network down once the CPUs peak out. Your phone will start ringing off the hook with users helpfully telling you that their spreadsheets are taking forever to load. Unix GUIs have historically been so clunky and ugly to save clock cycles for more useful things, like serving up files and SQL data records to clients. When GUIs were used at all. This is part of why *NIX is eating MS for lunch in the server space (leaving security aside).
Being a BSD beneath it all, I'd expect there's a way to boot OSX into a command line and not use the GUI. If you do that, I'd expect it to be similar to running BSD on any other machine. As a server, OSX should require no more and no less of a knowledgeable maintenance staff than any other BSD implementation. Once you're underneath the GUI what's it matter whether it's Apple's BSD or OpenBSD or any other flavor BSD?
Now, given Apple is a sole-source outfit, if their stuff is well integration-tested prior to release, it might save a little bit of time, but that cuts both ways - I'm currently stuck with a used Beige G3 that I picked up (after consulting apple.com) last weekend intending to give OSX a spin, and now they've decided it's not going to be supported after all. The moral is, any savings in support staff must be weighed against the risk of getting locked into a marketing-driven, sole-source vendor for both hardware and software. So far, OSX looks very promising, but the very nature of Apple's business model presents the risk of Microsoft-style lock-in and a similar forced-upgrade treadmill without the benefit of commodity hardware pricing. Remember to take that into account. Given how much babysitting a Windows box requires, it might still pay off to switch though!
Additionally, Apple has very little of a track-record with enterprise servers. It wasn't until a few years ago that they managed to properly implement multitasking and multithreading, and to do it, they basically admitted to themselves they weren't ever going to figure it out so they copped the BSD kernel instead. I commend them for the decision, but it still leaves a seed of doubt as to their ability to follow through at the enterprise level. I've seen Apple change direction, suddenly drop product lines, and almost go bankrupt more times than I can count; you'll have to forgive me if I adopt a wait-and-see attitude for a while before I recommend bet
my xserves have never failed. redundancy doesn't mean shit if the product isn;t good to begin with. I got lucky with super micro and in fact the latest 2000 cpu cluster at my company is super micro too. but frankly the other copanies on paper were actually better.
the only reason I dont buy more apple xserves is that as long as I can get lucjy the linux boxes are cheaper. but if I had to pay anybody the same as an apple i'd rather have an apple since I know it will work. you can have your redundant power supplies.
the key if finding a good vendor and sticky with them.
Some drink at the fountain of knowledge. Others just gargle.