Slashdot Mirror
Microsoft's Chief Linux Strategist Interviewed
sl0wp0is0n writes "Computerworld has published an interview with Microsoft's chief Linux strategist, Martin Taylor. It's interesting to find out that Microsoft thinks and predicts Novell (SuSE) will be the dominant Linux distribution they'll have to compete against. The interview also has Taylor talking about indemnification, IBM and his realization that customers generally adopt Linux to get a better TCO than Unix, not Windows."
When you're getting something for free, [vendors] get a lot of "get out of jail free" cards. You see [people saying], "Oh well. We didn't pay for it anyway, so we shouldn't care too much about security. We'll fix it ourselves. Oh, there's no regression testing. Who cares? We'll do that ourselves." But once you start writing a check, you now have demands, and rightfully so.
And indeed, for me, this marks the start of Linux having the potential to be a threat. It means that if a commercial Linux is a viable option, then more commercial software will be written.
Get thee glass eyes, and, like a scurvy politician, seem to see things thou dost not.--King Lear
It's interesting to find out that Microsoft thinks and predicts Novell (SuSE) will be the dominant Linux distribution they'll have to compete against.
Novell? They'll be lucky....more like Knoppix!!
Could anyone explain that to me? This guy is explaining that people put KERNELS into DISTRIBUTIONS?
Personally I wouldn't pay attention to an organisation that can't even put together a web site properly.
/.
And no, I'm not talking about
At least as far as this interview goes, it's all about corporate strategies AGAINST Linux suppliers and integrators. Little to nothing about OSS's/Unix's/Linux's strengths. Again, they are fundamentally missing the point in the interview.
That doesn't mean they aren't using their legal and financial blunderbuss to defeat the Linux vendors/integrators the same way they wiped out Netscape, though. If so, they almost certainly won't talk about it in an interview.
I didn't RTFA though so troll me if you wish.
http://jayceecorder.blogspot.com
It could have been stranger:
An Interview With Microsoft Linux's Chief Strategist, Bruce Perens
I'm always amazed at the lack of any sort of CLUE that these microsoft representatives have. One sometimes wonders how they managed to get into their dominant market share with executives that think TCO is the most important thing to cause someone to switch.
As you can see here with this little nugget:
And you can end up with Linux not being Linux, but Red Hat Linux being different than Novell SUSE Linux, Debian Linux and Mandrake, or whatever the case is.
Very nice. His teacher at FUD school must be beaming now.
Oh well, did you really expect a MS Linux Strategist (nice title btw) to say or do anything different then what we read in the article? The same would be expected from a Linux MS Strategist (if there is such a thing) doing spin on Windows.
Circle of life...
"Leo Fender was in a 'state of grace' when he designed the Stratocaster." -- Paul Reed Smith
Anyone who thinks Microsoft is going to announce its GENUINE thoughts about Linux to the public world is deceiving themselves.
Not knowing anything about SUSE (other than its German focus), why would Novell choose them as a Linux distro? Does it have some capability that other distros do not?
customers generally adopt Linux to get a better TCO than Unix, not Windows
While that may be more or less true in the US, from what I've read it seems like a lot of foreign countries are switching to Linux from Windows for the better TCO as well.
In the US, it seems like a lot of big Unix companies are switching - but eventually there will be a large Windows to Linux switch here to.
Looking for hardware (Currently need: Large Etch-a-Sketch) Have one? See my journal!
Total cost of Ownership ?
I thought and it was Microsoft and its BSA/SPA satellite that software could not be owned, hence the EULAs.
So, they imply one might OWN a system ?
Trolling using another account since 2005.
to their 'independant studies', I can only conclude he is lying and knows he is lying.
How does it feel to sell your credibility, Mr Taylor?
they think of Internet Explorer..
No one pays for that, so Microsoft "Get Out Of Jail" for that? I think not...
And also according to those click-through licenses my rights pretty much include "up to but not more than $5".. so that's a comfort is it?
I love their TCO argument. If you read the MS EULA, you will find you have NO ownership of MS software at all, unlike Linux. So it should be MS's Total Overall Cost (TOC) vs Linux's Total Cost of Ownership (TCO).
Get the facts, MS owns you if you use their software.
Indeed, My guess is that this started right after you "funded" SCO's litigations and started to pantent every damn thing under the sun.
And you are surprised customers brought up something you brought on? Puhleeze..
-RG.
They're absolutely right. The major migrations in big corporations tend to be replacement of Solaris boxes, with I suppose HP and AIX getting a look in too.
The home user running Mandrake isn't what they're thinking about here, though I'm sure they spend some time on that too. No - they're thinking about datacentre stuff. But don't take my word for it - ask Sun. Ask 'em how their sales are recently, and why they've had to start offering Linux and x86.
Cheers,
Ian
I just setup a linux file server for my business. If I bought a windows server with enough licenses for the computers in my store, it would cost me $3000. Linux on the other hand cost me $0.
Now if a person who needed a server like this didn't know anything about linux, I'm sure he could hire someone for less than $3000 to set it up for him. $100 to hire someone for an hour would be reasonable.
I just thought I'd throw in that example...
Outdoor digital photography, mostly in New Engl
That's because Novell has withstood the onslaught from Microsoft and still managed to eke out a survival. The folks at Novell know how to fight back against Microsoft.
--
A neighborhood journal
Microsoft's anti-Linux strategist, on the contrary, will probably be recommending more changes to Microsoft networking to put more roadblocks in the way of the Samba people, more file-format changes to Word and Excel to screw OpenOffice, and stuff like that. It's rumored that Microsoft has in the past hired actors to behave like really obnoxious Linux fanboys at trade shows, damaging Linux's image - if it's true, no doubt he'll have a hand in that, too.
Just unreal. It sounds like he's basically saying that IBM, Novell and RedHat will start stabbing each other in the back, and fuck over customers in the process, pretty much for the sake of stabbing each other in the back.
In the real world, strategic alliances exist because you realize that by co-branding or working with another company, you can make more money, grow market share and benefit customers.
Apparently, that's not how it works at Microsoft.
Buy the President
"Chief Linux Strategist" eh?
I'm still waiting to Microsoft to grow a pair and try to buy out one of the Linux distributions. Although I really don't like Microsoft, it would be an interesting turn of events. Especially with Sun doing what it's doing with Linux(wait, what exactly are they doing? Re-branding SuSE and Redhat?), Novell buying SuSE and IBM funding RedHat out the arse.
So if Microsoft were to consider buying a Linux distro, who does that leave? Ahh yes, Debian or Gentoo. Now *there* is a dilly of a pickle.
The Uber
http://www.tulg.org/
http://devurandom.livejournal.com/
From the linked article:
When I talk to customers and they say, "Hey, we can get better TCO with Linux," they're not always saying better than Windows. They're saying better than Unix.
Hardly surprising. For a customer migrating from a commercial UNIX version, the switch to the UNIX-like Linux will probably be much easier than the switch to Windows.
In this case, the difficulty of switching to a completely different environment works against Microsoft. But this merely balances out some of the Windows environments, whose owners find the switch to Linux too difficult.
C - the footgun of programming languages
Any other surprises? The surprising thing, a little bit, is how predictable our conversations are now with customers. ... One other thing that's come up more over the last 12 months is this notion of indemnification [against patent and copyright claims]. More and more customers are asking us, "Help me understand what you do from an indemnification perspective versus HP or IBM or Red Hat or Novell." That's weighing into decisions more and more. ... Customers began introducing it and asking me about it more than I was introducing it to them.
The FUD is working, and working well.
My Karma: ran over your Dogma
StrawberryFrog
IBM has had a deliberate strategy of not having its own distro. This guy obviously thinks that is a bad idea. He is implying that IBM has no idea of where it is going with Linux. He seems to like what Novell is doing though. Personally, I think he is totally underestimating the enemy (IBM). IBM has shown that it can totally re-invent itself if necessary.
You got to love the title.
It's like:
Chinese government's Chief human rights activist.
Vatican's Chief birth control strategist.
McDonald's Chief vegetarian strategist.
What a great title!
Do you have any lined up for the future?
They're going to continue to be around the scenarios that customers say are important -- TCO, security and reliability.
--END QUOTE--
Well, one out of three aint bad. MS don't have a super reputation for either reliability or security and even their TCO studies in favour of MS are very suspect. I guess these must be future goals for the company.
All those moments will be lost in time, like tears in rain.
It's kind of sad to me actually. If customers *are* talking (to Microsoft) about indemnification issues, then any of Microsoft's allegedly behind-the-scenes investments in SCO's legal actions have paid off for... for Microsoft at least. Another FUD issue successfully on the table.
And notice how the TCO issue is spun... "oh the real Linux TCO issue is versus Unix"... so one might overlook the savings one would have using Linux rather than Microsoft. Why do I run Samba rather than paying $1000 for Windows Server? Or Apache rather than $1000 for IIS+Windows Server? Why does Microsoft cripple their software so that "Software Update Services" (which allows me to check from a central workstation if the PCs running on our network are patched to fix *Microsoft's* security holes) so it only works with Windows XP Professional? In a small/medium business, I now have to run around to all the PCs to doublecheck them because Windows XP Professional on every desktop is one more expense we don't need. And one has to take care that all the laptops which come and go at the end of the day get checked. Compare that to remote administration of Linux systems where it's super-simple to login remotely in the middle of the day or scan programmatically...
Linux isn't strategic for businesses because it lets them reduce a few Unix expenses (although any shrewd businessperson will take what they can get)... it's worth pursuing so you don't end up beholden to one big vendor for all your software. Microsoft's prices *do* keep rising over the years you know...
--LP
I mean, when have you ever heard of any reasonably competent Windows admin (yes, they do exist!) installing, say, a service pack without some serious testing beforehand?
HAND.
When some company starts offering cash for tradeins of original Windows cds for a linux/bsd install, then you will know linux is a commercial threat.
Unfortunately unless I invoke the underpants gnomes, I can't quite see the business plan - unless I am IBM flogging services.
The Singularity is closer than you think
Quant
One other thing that's come up more over the last 12 months is this notion of indemnification [against patent and copyright claims].
...
Yes I wonder who is making it an issue.
More and more customers are asking us, "Help me understand what you do from an indemnification perspective versus HP or IBM or Red Hat or Novell." That's weighing into decisions more and more.
Yes because again Microsoft are trying to tie people down with fear that what they will touch they will loose again because the big Microsoft guys will spoil thier fun.
Customers began introducing it and asking me about it more than I was introducing it to them. And I began to say, "Wow. We really stand behind our technology in a pretty aggressive way.
Hahahah yes you are plenty aggressive, like a cornered animal, even the Ministry of Truth could learn from you guys.
We should make sure that we get credit for that compared to Linux in many ways." And it's actually been something that tips the scales sometimes when people are on the fence.
Is that the barbed wire elecrified fence of 10 year supply deal, licensing terms, special backhanders, propriatary formats et al.
Lets all hug this guy. Anyone notice how Microsoft are finding security holes in its own software right when it wants you to upgrade?
#hostfile 0.0.0.0 primidi.com 0.0.0.0 www.primidi.com 0.0.0.0 radio.weblogs.com
...."his realization that customers generally adopt Linux to get a better TCO than Unix, not Windows."
They wish that this were true. Oh how they wish it were true.
IBM has stated that the future of the RS/6000 is LINUX. At some later date they will be converting them from AIX to LINUX. LINUX is slowly replacing UNIX and making headway against Windows in the server environment.
I'm not sure, but MS Bob probably predates their love of the internet.
When MS first became aware of the importance of the internet (somewhere in 1995), they started up MSN. MSN was supposed to become a "Microsoft Internet", with all content provided by MS. Something like AOL or Compuserve before they connected to the internet.
Unfortunately for Microsoft, people prefered the "real" internet over a proprietary online service [1], and MSN had to be revamped into a regular ISP. Since they couldn't provide a real advantage, MSN wasn't very successful as an ISP.
[1] AOL, Compuserve and other services like them had to do the same.
WWTTD?
See for yourself:
R.I.P.?
I really find the opengl to be a far more worrying story, who will get linux for free, and pay EXTRA to play games on it because Microsoft want to huddle in opengl.
Someone should rule that Microsoft cannot buy openGL, just like big company ABC might not be able to buy other big company XYZ if they become to big.
Shocking.
#hostfile 0.0.0.0 primidi.com 0.0.0.0 www.primidi.com 0.0.0.0 radio.weblogs.com
He says that IBM thinks: "Hey, because of our global services business, we can cobble things together and try to veil that for the customer and deliver solutions."
"Cobble"... "veil"... Ouch, that hurt real bad. Can it be this guy bears a grudge against Big Blue - just a bit?
It's sensible in a 'spin' interview like this to focus on persuading people that windows is better than what's currently out there.
.NET API, Office/music/video DRM, putting increasing resources into the .NET versus Java battle, dropping the price of windows to emerging markets and encouraging the use of non-standard MS technologies by bundling new API's and apps into windows at every opportunity.
I'm sure they also have an anti-OSS strategy internally but this is likely to be very sensitive information. Probably their anti-OSS strategy includes creating new standards for the Web via Avalon/Indigo that become reliant on having the windows
These are the kind of strategies that are neccessary to discourage linux adoption. Every change to windows that makes it less easy to migrate to linux must be hidden as either eye-candy, ease-of-use or a DRM "feature".
Matt
They're absolutely right. The major migrations in big corporations tend to be replacement of Solaris boxes, with I suppose HP and AIX getting a look in too. I can tell you one thing... AIX running on a p690 is in a totally different league compared to linux running on the same hardware. Linux stability, scalability and especially I/O performance is nowhere near AIX performance on said hardware. It's the small to midrange Unix server market that gets a hit... but the high end servers are still Unix all the way. Linux doesn't even come close.
To make this quick (and hopefully readable/coherent), as long as there are quite a few Linux players (and even *BSD ones) competing with each other, multiplying centers of FLOSS development, M$ will have a hard time dealing with the FLOSS movement, especially if volunteers keep playing a significant role, because Bill & co. just can't wrap their minds around the whole phenomenon (sp?).
As long as the various Linux distros and the BSDs don't play in the "traditional way", in the way that M$ understands, as long as *anyone* can contribute to the FLOSS movement, "we" will stay an elusive, hard to kill target. This was said repeatedly over the years, that what makes GNU/Linux a nightmare for M$ is the fact that there is no single company to buy out or to Netscape (the "cutting the air supply" thing).
The minute you shrink the field to only two (big) companies behind GNU/Linux (doing the bulk of the heavy lifting in development, BTW), you've just ~agreed to play on M$'s terms. M$ understands other, traditional, companies following a traditional business plan and getting traditional results/objectives/whatever.
The minute M$ can understand you, the minute they can "frame" you, you are f**ked.
This is why I sincerely hope that Novell will only be one of many players in a field where the loss of one of these players will not be a significant loss to the FLOSS movement because it will be able to pick up and continue more or less as if nothing happened. The same goes for Red Hat.
I want to go back to a world where I can choose between 4 or 5 shrinkwrapped distros updated at semi-regular intervals, each contributing in his own way to The Movement but not being *the* cornerstone of FLOSS.
If Novell or Red Hat become too important, if they "become GNU/Linux", M$ will simply have to kill them off (which should be easy in the case of Novell, sadly) and simply sit back afterwards, reaping the rewards of having killed off yet another (potential) competitor.
We just cannot let M$ define the playing field and play by its rules. Not to sound too much like ESR, but prior/current behaviour on M$'s behalf leave no room for peaceful coexistence unless they've been kicked in the nuts very hard and brought down a few notches, just like IBM was in the '80s and early '90s. We, nor anyone else, cannot compete in any traditional fashion with M$: they only way to do battle with The Beast From Redmond is gerrila tactics, more or less like the FLOSS movement has done up until now.
Change tactics, start playing the traditional game and see your dreams go down the drain.
It should be Microsoft's Chief GNU/Linux Strategist, except if they feel that only the kernel threatens them.
Mono is the real interesting stuff behind Novell, not Suse (who wants to use Suse anyway ?)
To some extend they miss the point focusing on Linux only.
Firefox, OpenOffice on MS-Windows are very good mid term alternative on the road to the linux operating system.
Once a user is used to these FLOSS tools on MS-Windows, the cost of change towards Linux as the OS becomes marginal.
Know what really annoys me about this interview? How this guy continues to spread FUD while trying to make it look like something other than FUD by attributing it to his "customers". I.e., "our customers keep asking about indemnification" or "before Linux was commercialized customers were willing to cut it some slack for poor security". Nice try, Martin.
Actually, I was trying to be Insightful, not Funny.
Microsoft's patches have been known to create quite a few problems, you would hope that a company with the resources they command would be able to perform a relatively compentant test of a patch.
I offer this as a first question
You state that MS should get credit for how aggressively they stand behind their product. Are you referring specifically to Lawsuits and indemnity? I sure have never seen MS step up to bat about the damage to the internet, small and large businesses etc... caused by the uncountable number of viruses written to your platform. Please do clarify.
I'd guess that this focus will end up biting MS in the ass, in the end. Currently, Microsoft is trying (well, still trying) to get Windows on servers, datacenters, etc. If it's not a desktop, Microsoft is trying to put Windows on it. Why? Because they've saturated their growth in the desktop market (that came about as a result of the mass computer buying of the 90s). The only way to continue their growth is to diversify. And the biggest and most successful brand name they have is Windows.
The problem is, while they're busying trying to still penetrate the server market, which Linux is doing a nice job at expanding into (at the expense, mostly, of Unix machines), Linux has the real potential to encroach on the desktop market. I'm sure Microsoft realizes that. I'm sure they also realizing that "circling the wagons" to "weather out" the Linux threat won't work. That's the whole point of Longhorn. The fact that WinFS *still* isn't coming in Windows is a real disappoint/problem, though. It's both a sign of a core problem (backwards compatibility, both in the outside appearance but also in the code itself which is surely a major reason it was put on hold) and a sign that Microsoft's strategy of adding in tons of features (vapor or otherwise) isn't working.
In the past, the FUD/vapor of a perspective product launch, even if it kept being pushed back, would end up killing or crippling the competitor's product. Instead, Linux really hasn't done anything but slowly grow in the desktop space. Without an actual strategy to combat Linux, a sudden burst in people using Linux could severely cripple the Windows money stream for Microsoft. Then, Microsoft will have to use its massive cash reserve to try to come up with a way to continue to make money.
Of course, if Microsoft develops another highly profitable department, this becomes less of an issue. But, the only thing that's even close to that is XBox. Maybe that'll keep Microsoft alive, but then Microsoft will only be known as a #2 or #3 console maker. I don't think the CEO of Microsoft would like that too much.
Eurohacker European paranoia, gun rights, and h
TCO of Linux over HP-UX is exactly why my employer is moving 10,000 people to Linux. Everyone will continue to have their Windows XP box right next to their development station. I've not seen many labs replace windows yet.
More and more customers are asking us, "Help me understand what you do from an indemnification perspective versus HP or IBM or Red Hat or Novell." That's weighing into decisions more and more. ...
"Well let's see first we find some really angry people, then give them tonnes of money to engage in a lawsuit on a completely spurious charge. Oh yea we get them to threaten ridiculous buisiness proposals if they win.
It doesn't really help anyone but um it won't hurt you as long as you join us.
>Linux doesn't even come close
That is until you put in n+1 active configurations. We went from a shop of purely big iron, SGI Origin 2000, Sun 6800, HPUX, Sequent. And have replaced it with Linux and have a higher overall stability, scalability and performance.
Compared just 1 linux to 1 big iron the big iron will beat it, but all I've got to do is buy 1 or 2 additional linux box and my availability is better (and been proven better over the past 2 years in our environment) than just 1 big iron box and a much better capital price point.
Another marketdroid Information Minister. As long as they're paying this fool, we have nothing to worry about. Clumsy attempts to differentiate between Novell and Red Hat. Bizarre statements about IBM. TCO pie in the sky. Regression testing my ass. Lame kernel jokes.
Keep this one, MS. We like him. Cute, clueless and cuddly.
insecurity asks the wrong question irritation gives the wrong answer
What I have thought for a while is that the hardware vendors will learn from the mistakes of the past. Instead of just creating hardware, they will also supply their own brand of linux with their machines.
Even today you see hardware manufactures creating their own software, look at HP and everything they are doing. IBM? APPLE....
I think companies like IBM, DELL, HP will be the first ones to do this, with NOVEL and IBM leading the way. It would give them a very high jump in revenues, not needing to pay for all the different licenses, allow them to bundle their own software (music store software the big thing I can think of first, then changed into a mulit media application like I see ITunes becomming), and partnerships with many companies (I cant belive AIM is not bundles with any manufacture PC's).
So yeah, pc manufactures will be the ones killng of MS (if anyone) because they will be the ones who change the home desktop.
TruePunk | Games
That's because Novell has withstood the onslaught from Microsoft and still managed to eke out a survival. The folks at Novell know how to fight back against Microsoft.
Or perhaps thats how they want it to be. One big target is much better than many small ones.
Think Napster. Its was much easier to sue Napster (they even disrupted the service), than to sue individual users in decentralized p2ps.
So, they probably expect one linux vendor to dominate the market, so then they hit it (patents, advertisement, whatever), and damage linux image (because whatever they do, we will always have non commercial distros).
http://www.computerworld.com.nyud.net:8090/softwar etopics/os/linux/story/0,10801,95988,00.html
Escher was the first MC and Giger invented the HR department.
"Red Hat can do stuff to their [distribution] that IBM doesn't get a vote on."
"IBM will be forced to compete with those guys because they won't have as much control of the kernel, of the [Linux distribution]."
Are these s'posed to lead me into thinking that IBM has a vote on what Microsoft can do to their software? Who is "try[ing] to veil that for the customer and deliver solutions", now?
"So IBM's going to be even more beholden to Red Hat and to Novell to do things in that [distribution] for their application stack to work effectively."
'Come be beholden to Microsoft, and we'll support you completely.' Yeah, right...
New for spring 2005 Windows XP for POWER processors!
I think what makes Novell a real serious competitor to M$ might be more probably Mono (http://www.mono-project.com/about/index.html) as an alternative to Microsoft .Net than SUSE as alternative to Windows.
http://www.computerworld.com.nyud.net:8090/printth is/2004/0,4814,95988,00.html
Escher was the first MC and Giger invented the HR department.
In other news, Microsoft has a Chief Linux Strategist!
Okay, let's look at the XP license:
That's really backing up your software guys.
FOSS == Free / Open Source Software. What's the "L" for in FLOSS? Are you just really keen on dental hygeine?
When you lose something irreplaceable, you don't mourn for the thing you lost, you mourn for yourself. - Harpo Marx
The MS guy was comparing what novell offers "stack-for-stack" to what MS offers. Novell has things like Directory Services (NDS) and ZenWorks that are as good or better than MS's Active Directory stuff. With the acquistion of Suse and Ximian they get things like OpenExchange & Evolution that also potentially challenge the Exchange-Outlook team. Add in the fact that Ximian's Mono could help break any MS stranglehold over .Net. The question is whether Novell can get their act togather and integrate all these *potentially* great things into a coherent and polished suite that would let you run a complete "Novell Shop" with a Novell server-OS (e.g. Suse), Novell manangment solution (e.g. ZenWorks/NDS), and Novell application servers (e.g. OpenExchange) in the backroom and a Novell client-OS (Suse Desktop) and applications (Evolution) on the desktop. Add in the ability to itegrate a "legacy" windows enviroment and tie it all togather with Mono. That is Novell's potential. We will soon see if they can live up to it.
Of course Microsoft would like to predict Novell would be their main competitor. Novell has yet to articulate a software licensing philosophy that truly embrassess Free / Open Source software ideals. Instead their CEO, Jack Messman, talks about a 'both source' future. Microsoft's biggest competetor is not Novell, or Redhat, or any other single company. Microsoft's biggest threat is the free software movement, as exemplified by the GNU GPL. Microsoft's nod to Novell is a transparent nod to the notion that "the enemy of my enemies is my friend".
There are a tremendous number of half-truths and misstatements in what is, in reality, a very short interview.
From bringing up indemnification, to the implication that IBM can only implement Linux because it has so many wonderful techs to throw at a problem-child operating system, through the implication that IBM, Novell and Redhat will begine infighting over the code, this interview is pure Microsoft FUD. It's a rather well-done piece, though, and it is easy to get confuzzled by it all.
I'd just like to point out that this is a mighty interesting trio of players Microsoft is whining about... IBM, Novell, and Redhat... Gee, where are those three tied together again?
The only surprise, really, is that there was no sniping against Autozone in this piece.
It was a joke! When you give me that look it was a joke.
Picking Suse as the primary target implies Microsoft realises they already lost a world market to linux. Big swarm of linux distros, so many movable targets they cannot compete with.
That implied the need to defend at least at home. I bet you can expect some big law about operating systems in the U.S.A. in near future, based on DRM control, with publicity motivated by terrorism, as usual.
There you are, staring at me again.
installing linux on hardware that's on the HCL is trivial and takes half the time of Windows
... if you choose a minimal install, it takes very little time to install -- however, you have to take the time and frigging deselect everything that is preselected for you, and then go through the million dialogs saying "xyz depends on this library" and turn a bunch of things back on you way in the negative as far as install time goes. If you choose a default install, be prepared to wait about 4 hours.
You are ON CRACK if you think that installing Linux is trivial. Given the inconsistent, poorly worded, in some cases confusing text presented during Linux distro installs I'll take the XP cd any day of the week.
Same comment applies to install time
An XP install takes about a half hour.
Running it, well - it runs itself (especially for a file server, we're not talking about a fancy application server here).
If you take this view on any server, it WILL get owned eventually.
On the flip side, Microsoft is giddy and hoping it's Novell because they feel that since they've beat them before, it'll be nothing to do it a second time.
Novell will be ready for their sleazy tactics this time however.
I know some people _think_ rpm is standard, but if package management was the same for every distribution (read, they all agree on some packaging standards), then distros could flavour their Linux, and vendors could reliable create one package for their software which would correctly account for all dependencies on all distros.
I think this is the number 1 problem that needs to be fixed... and it is solvable with co-operation.
Like all pain, suffering is a signal that something isn't right
> [...] his realization that customers generally adopt Linux to get a better TCO than Unix, not Windows.
Of everyone that I know personally that has switched, not a single one did it to "get away from Unix", rather all of them were Windows users and totally ignorant of Unix until they tried Linux.
Once again Microsoft is getting its version of reality from somewhere unknown to most of us and once again they believe they can dictate the reasons users do what they do.
Wise up Microsoft, YOU'RE NOT LISTENING.
THAT'S why people are fleeing in droves.
I don't know the meaning of the word 'don't' - J
I like the way he stokes people's fears over the SCO suite by referring to the whole issue merely as 'indemnification'. Like they're not the ones responsible for the inferred threat to start with.
I can't stand slimey bastards that use indirect references to things to be 'terrified' of, whether it's a Microsoft employee or the neo-conservative basket cases running the show in the US at the moment. How about talking about the real issues, both of you!
Step one, ignore Linux.
Step two, bad mouth Linux.
Step three, file patent suits against anyone who uses Linux.
BTW, we're at step two now.
If someone says he and his monkey have nothing to hide, they almost certainly do.
Microsoft commissioned analyst firms to do reports to help you "get out the facts" about Linux. Are you still doing that? If someone says, "Hey, Customer X says, 'If I had this data, it will help me make a decision, comparing Microsoft to Linux.'" And I basically hop on the phone with all the folks [at the analyst firms] and say, "Hey, I talked to four or five customers in the last two months, and they all care about x versus y. It's something that I think people care about. Can you guys go do something?" And sometimes they come back and say, "Yup. We've heard that, too. We're going to go do some analysis." Or, they come back and say, "Actually, it's not that interesting to us, but if you care about it, we'll use our methodology and stand behind it, but you have to fund it, because it costs money to get the samples, get the customers, get everything." That's going to continue to be my process. If there are facts or things that are needed, I'm going to hope that I can entice the analyst firms to go do it on their own because they think it's also important. But if they don't, then I'll commission it.
I have an enormous amount of difficulty believing this guy when it comes to his answer to a question on Microsoft's FUD tactics. Him claiming that Microsoft is nice and easy going about the methodology used in Microsoft commissioned analyses and that Microsoft doesn't use financial pressure (or that Analyst firms don't offer to cook the report in exchange for cash) on analyst firms strikes me as a total lie.
For example, the most well known example of Micorosoft's lower TCO claim (and the one displayed prominently on MS' website) was made by comparing Linux on a mainframe vs. Windows on cheap x86 commodity harware. There was no mention of the reasons a customer would go for a mainframe (reliability, bandwidth, scalability), just the FUD about Win2k3 on a dell box.
I think this is just the new (old) MS tactic of pretending to be nice in public and fucking everyone over in private.
Oh, ok. Nothing to worry about, then. In fact, no point in having a high-paid "Linux strategist" on the payroll - get rid of that guy, it's a waste of shareholder profits.
"Somebody has to do something. It's just incredibly pathetic it has to be us."
--- Jerry Garcia
Those shops were already 80% microsoft, so they don't have the time or energy to invest in an unix expert. It's much cheaper to hire a MCSE dork who doesn't know any about unix and took some cert class and passed.
his answers show he doesn't have a clue about linux.
Yes, the lower TCO of Linux versus Unix is a valid argument. I agree that Linux adoption is seen as a means to lower the cost of providing services on Unix systems. However, these services are generally provided on Unix systems in order to provide sufficient power, at a lower TCO than a suitable Windows system. So, ultimately , Linux is just a cheaper Unix, which is cheaper/more capable than Windows.
There exists no way of exchanging information without making judgments. --Bene Gesserit Axiom
I adopted Linux because I decided I wasn't paying Microsoft ever again, and couldn't afford a name brand UNIX for workstations.
My new server is a SunFire V240 running Solaris...
Sleep is for the Weak
The major migrations in big corporations tend to be replacement of Solaris boxes, with I suppose HP and AIX getting a look in too.
Although I generally agree with you I think AIX gets more than just a look in. Where I work (Telecom company), for most smaller servers, the choice is SUSE's line of OS packages, in part because of the ease of administration as well as good support. For really big rock solid production systems the choice is AIX and to a lesser extent HP-Unix over Linux since our experience of AIX/HPUX vs Linux shows the former to be more stable if you need absolute reliability. We usually try to replace Sun OS with machines running Linux or AIX, and that is not just brand snobbery. We inherited some numbes of quite new Sun Machines in a recent merger so we have actually had a chance to make a balanced comparisons. Microsoft OS'es are only used for customers specifically requesting MS solutions, failing a specific request for an MS server the customer defaults to a UNIX or Linux system. Apart from that MS OS'es are only used on workstations and for proprietary measurement gear that does not run with anyting else. The general trend for Microsoft server systems has been to minimize their number as much as possible and fortify them heavily because of the disproportionate amount of work we have to sink into them. This has largely been due to security breaches even though we autopatch the MS boxes the moment the patches are posted by MS and generally make every effort to secure them. Incidentally the number of MS Workstation installs is on a (slow) downward spiral for the same reason. People are spending more time than they can spare dealing with adware, viruses, worms etc...
Only to idiots, are orders laws.
-- Henning von Tresckow
Do you have any lined up for the future? They're going to continue to be around the scenarios that customers say are important -- TCO, security and reliability.
So, when Windows "wins" any one of these, we know the research is pure bullshit.
I love the talk about indemnification, too. People are worried that they won't be indemnified, so they'll run to Microsoft. Brilliant. Is that the same Microsoft as here? Surely it's another Microsoft we're talking about...
For those who don't want to click:
Do you have ESP?
Because SuSE is basically in bed with AMD, and the support for AMD from SuSE makes RedHat look like a garage effort.
My prediction: RedHat will slowly go away if they don't adapt to what SuSE is doing....namely, bringing better hardware support. Case in point: 64bit. RedHat's 2.6 kernel 64bit support sucks it on many levels.
no martin taylor, you are mistaken. we ARE syaing that Linux is better than Windows. we are even willing to learn OS internals, compile our own kernels and applications, contribute our (better, more stable and more secure) applications to our comnnity for FREE rather than use your putrid excuse for a product and have to swallow all of the PR (in most polite terms) that your outfit spews. please keep your miserable opinions to yourself.
What weird invention from Microsoft (thru SCO) : the concept that you can be sued for using linux. But each move from China proves that patents and lawyers are irrelevant.
It's not "Linux versus UNIX", it's "UNIX versus Windows" with Linux being a very cost-effective version of UNIX.
Microsoft depends on expansion, they're running out of places to expand on the desktop, and they're going after the UNIX server market. The UNIX market is firing back with Linux, but it's such a big fluffy diffuse market that Microsoft is trying to (and in some cases succeeding) convince people that doing a complete conversion from UNIX to Windows is going to be cheaper than converting from one UNIX to another.
Which is just as ridiculous an idea as the one that you automatically save money in the short term by doing a sudden switch to Linux on the desktop: platform swaps cost money. Even Microsoft took multiple tries to do *theirs* at Hotmail, and eventually ended up using Interix to run existing UNIX software on NT... so they didn't do a full conversion after all.
Talking about Linux as if it's an "alternative to UNIX" is just playing into their hands. They know they have to split the competition, that's why they're pushing so hard on the whole SCO case and "indemnification". That's why they did this strategic interview to keep cranking the old SCO FUD machine...
Well, that explains the papercuts on Steve Ballmer's face recently.
"America has done some terrible things. But I know that Americans don't cheer when innocents die." -Dave Barry
...why I needa know about Microsoft strategies around Linux?
BTW I did not RTFW.
http://www.pbs.org/cringely/pulpit/pulpit20040916. html
How about poisoning USB for linux?
h
Patriotism is a virtue of the vicious
If I may revert back to high school, "No, duh!" Man, it is totallly revealing how clueless that Microsoft is regarding Linux that it took them this long to figure that out. It was obvious to me about four years ago, despite ALL of the industry rags saying otherwise (i.e. saying Linux is a threat to Windows) that Linux's first victim would be flavors of Unix that had ossified and weren't innovating but were charging huge fees.
.NET). The biggest threat to the current installed base of Linux is generally recognized as .NET. Linux developers need to develop a competitive offering (Mono, Java, whatever) as a purely defensive move to maintain share, assuming .NET allows developers to do things that they cannot do on any other other platform for a comparable price.
The primary reason is that the people supporting these ossified Unixes already had the skill (for the most part) to support Linux. As Linux gained the requisite features it was a relatively simple substitution for the Unix in question.
In order to switch from Windows to Unix, all of your admins would need to be trained or replaced and their salaries would go up. The cost of salaries can in some cases (especially in small to medium sized deployments) add more to TCO than the licensing. That's why some of the first companies to switch to Linux from Windows were huge companies that were paying millions of dollars in licensing fees. They couldn't care less if they were paying a few hundred thousand more in salary when they were paying millions less in licensing fees.
Of course, this begs the question of why they were using windows at all, but it may relate to the cost of development on windows. It is still easier to develop on Windows than on Linux or Unix. That's why many developers prefer Windows and that's why Windows is so appealing. It has tons of software available. Therefore more people are willing to deploy it. That's why Billy Borg Gates is always saying "it's the API, stupid."
Anyhow, Windows will only move upmarket where Unix and Linux rule now, if it can lower its licensing fees, which it is doing (note Malaysia Thailand, etc) or get such a critical mass of software developed on its platform that customers feel compelled to deploy it, which it is doing (note
If Linux wants to eat Windows' lunch, it has to become easier to develop on. An IDE needs to be developed that is comparable to Visual Studio. Once the software is easy to develop it will start to happen. It also needs to be at least as easy to use as Windows 2000. People can point out all of the flaws that they want about 2000, but it is good enough and it wins on ease of use for most people. Linux is getting there on ease of use, but it's not quite there yet.
Although, I have to admit that ease of use is less of an issue than getting developpers. Incidentally, this is why Apple hasn't grown share. There is nothing special about MacOSX other than ease of use and that is not enough to get it in the door of any corporation. Apple hardware and software are more expensive and in many cases cannot do as much as the competition or are simply comparable and not significantly (i.e. order of magnitude) better.
So, in sum, it's not Linux that will kill Microsoft. It is the insular, narcissistic, navel-gazing culture that has its blinders on to the rest of the world. They were blind-sided by the Internet, then Linux, and most recently by the "search paradigm". Linux just needs to not fall into the same trap. It can't be just software written by geeks for geeks, assuming people want Linux to succeed, where succeed means being ubiquitous and spreading freedom to everyone. Of course, on technical grounds, Linux in itself is already a success, but so was the DEC Alpha. Listen to the customer!
If everybody would converge on Debian we'd still have the distributed development going but without the cross-distro compatibility problems.
Installed the Bubblemon yet?
You don't have a clue about some distros. Linux Mandrake is easier than Windows XP Professional, so your statement isn't true by any means.
Perhaps they talk about SUSE because they want to lower RHAT's stock price?
It's obvious from the first 2 sections. You know it's FUD when Microsoft executives start telling us what "the customers" are asking them...
:-/ This is categorized under the Uncertanty part of FUD. Uncertant about all the positive press Linux is getting with regards to being cheaper than Windows.
First he says that it's about Linux TCO vs UNIX and not Linux TCO vs Windows. He tries to solidify this point by saying that when customers are telling them they're getting better TCO with Linux, that it's not always about Windows. Why would a Microsoft customer, say to Microsoft that they are getting better TCO on Linux vs UNIX? Remember also, they don't have to prove any of this and can make it up as they go. Heck, they do that in court too.
Next was how he was saying that MICROSOFT CUSTOMERS are asking Microsoft about protection from patents and copyrights. Is SCO going after Microsoft or something? This just seems silly for a Microsoft customer to be asking them. Especially with all the Microsoft licenses they have to agree to in order to use the software. IMO, this is another on of the "the cutomers are asking" PR stunts to try and add credence to the SCO vs Linux issue. ie, the Fear part of FUD.
I could go on, but it's pretty obvious this is just a PR presentation and ComputerWorld offered up their stage for it.
That part about Novell just means they now have a target they can shoot at. Especially since Novell is once again going after the desktop OS market( Ray Norda started this back in the mid 1990's. With Linux too! ). Anybody else notice how they've been using 'birdshot' in their PR gun against Linux/OSS the last couple of years? They are no better off today though. IMHO.
LoB
"Anyone who stands out in the middle of a road looks like roadkill to me." --Linus
Suse is a typical German product. It's well engineered, and after The Second Big Argument, they've learned to make things field servicable. I'm not so sure it's going to be top dog, as China has all their people working on their version of Linux.
Honestly, people are so worried about Linux that they fail to remember the applications. You make a GPL application with more advanced filtering algorithms than Photoshop, Pinnacle, Premiere; and then you make it easy to use and adaptive. Sure there would be a MSFT port of it, at some point. But in the meantime, the Operating System would become completely arbitrary for a whole bunch of businesses. Such as newspapers and universities.
The Cost of Pwnership (pwning MSFT, that is) is the applications.
He says the conversations are predictable, he says they are saying those things.
THIS DOES NOT MEAN IT IS HAPPENING EVERYWHERE.
Just because this guy says there is a trend toward these FUD items, does not make it so. What else is he going to say?
Don't get me wrong, I think the FUD is having a negative effect. However, you can't simply take his word for anything really. He is a paid spin doctor. The first interview posted here demonstrated that.
Blogging because I can...
I propose a new law:
Whenever a Slashdot poster mentions MS BOB (released almost a decade ago) the thread is officialy over.
Of course one could drive a large truck bomb to m$ and co and do the same thing.
Well, I'm ready to get modded to -1 coz reading ur post made me see a similar analogy, so here goes -
MSU.S.
FLOSSterrorists
Same battle concept right? The first does not know how to deal with the second because the first does battle in a TOTALLY different way.
Now, just to clarify things, I'm a Linux guy myself and I use OSS wherever possible. So PLEASE PEOPLE, I'm not saying that FLOSS are terrorists, ok? I'm ONLY TALKING about the battle plans here.
Find a job you like and you will never work a day in your life.
I remember watching a talking heads technology show on PBS in the late 80's where a spokesperson from IBM was describing how they were going to put the WinTel threat to rest with their PS/1 line of computers with the advanced MCA bus running OS/2.
This guy's entire line exhibits the same level of brand-centric myopia. Hell, he just noticed that there were different distros of Linux:
Now the challenge will be [that] they're going to need to do stuff to differentiate themselves from Red Hat, which then means that they need to find ways to basically almost have a customized distribution. And you can end up with Linux not being Linux, but Red Hat Linux being different than Novell SUSE Linux, Debian Linux and Mandrake, or whatever the case is.
For this analysis Microsoft pays this guy a 6-figure salary? I'm no genius, but I knew that Linux was not a specific distribution (e.g., "...Linux not being Linux...") in 1994.
They need to wake up before they become another DEC.
"Rocky Rococo, at your cervix!"
Maybe they only see Novell as becomming successfull because they want to split in the Linux industry - and make it so all the players are fighting and trying to destroy each other rather than competiting and posing a genuine threat to Microsoft.
What a softball interview. When he started talking about analysts doing studies of TCO, wtf didn't the interviewer call him out on that bogus study that was recently condemned in Britain as false advertising? The one where they showed Linux costs more than Windows, but failed to mention in the advertisement that they included the hardware costs and had Windows running on a dual Xeon system and Linux running on a huge IBM mainframe?
Enabling and encouraging people who don't know programming to write code for your platform is no better than encouraging people who don't know civil engineering to build bridges for your roads.
Yes, but to do otherwise is requiring a civil engineer to paint your house.
The internet is a beautiful example of what happens when power is divulged to the many. You may argure that this isn't really a good thing, but nobody would dispute that the internet will be fading away anytime soon.
That was his point.
The last thing we need is a flood of insecure, buggy crap giving Linux a bad name.
Hello? Ever taken a look a sourceforge? I hate to tell you this, but insecure buggy crap is already prevalent. The argument is whether or not allowing as many people as possible to write code easily will move Linux forward, and I agree with the orignial poster that it will.
If the Linux development community needs anything, it's fundamentals -- a deeper understanding of computer science (as opposed to code monkeying), relational database theory, functional languages, interface design (not that MS is much better)
At it's core the statement is directly against the GPL, and related philosophies. Limiting access to reasources through either hiding the code (ala closed source) or requiring vast amounts of technical know how is directly contradictory to having the freedom to control your own computers your way.
I know I stretched your arguement a little to make a point, but the road you're on is very rocky indeed.
You forgot "hiring a Linux strategist to instill fear in customers who take SCO's claim seriously" and "pay SCO a 'licence fee' to fund their 'operating system'".
It's called OSX.
NT being a winner had nothing to do with infighting, and everything to do with the www.
Hmmm. Depressing isn't it.
For enterprise, it's easy: Our products it's supported on RHEL 2.1, 3 and Novell Suse Enterprise 9. That's it. Using Mandrake, Debian, TFM Linux? Sorry, not our problem.
To bring back the conversation to something more on topic: we have to acknowledge that competition to M$ will be more often than not a smaller entity than M$, with less financial & marketing means. Let's face it: we're all small fry compared to Bill and co. Billg made sure of that, as no other company has the same kind of constant revenue stream they have, nor the same war chest (*entire countries* don't have the money Redmond has in the bank (!), which should say something).
Since individually we can't even put a dent in M$'s armor, we have to use the "death by a thousand cuts" approach. A single cut won't hurt M$, but start multiplying them... This is why the number of players in the FLOSS market/movement must down shrink down to just one or two. You become easier targets for Bill's efforts to cut off your air supply while not even having a chance in hell to take market share away from him.
(note, when I write "USA" I mean the government, not the people).
Why do they have to compete "Against" insted of simply "Innovate", creation i'snt in the Redmond vocabulary ?
You mean a lower total cost of 0wnership.
me, too.
His whole perspective reduces people to cash cow-asaurus. The guy doesn't even live in this world. He comes across as someone needing to be beaten with a clue stick.
And you know what, that's perfectly OK. Windows made
its headway in the server market in the same way, chipping away at the low-end and making its way onto
the midrange as enhancements were made.
Since a move from big iron Unix to Linux is easier than moving to Windows, Linux will get the enhancements it needs to compete.
If you doubt this, then ask yourself why IBM would bother pushing Linux so aggressively when they already have a better OS with a larger installed base on the high-end hardware.
Pain is merely failure leaving the body
pot, kettle. Black.
It was a JOKE!
Now, back to a job that's safer than feeding trolls.
"We don't know what we are doing, but we are doing it very carefully,..." Wherry, R.J. Personnel Psychology (1995)
Off topic, I know...
/* should be OK in file scope */
/* can't initialize with function in file scope in C */
I lost count of the number of times I've traced problems with porting stuff to linux (redhat advanced server 3) to changes in glibc. This is code that works fine unmodified in Solaris 5.[68], AIX 4.x and 5.x, and under RH AS 2.
Trying to build it on AS3 with the newer glibc has revealed no end of problems.
While a number of them were legitimate problems with the code being ported (masked by the libc implementation details on the other platforms), quite a few were problems caused by changes in the glibc implementation. In each of these cases, I looked on various mailing lists and found numerous other people who had found the same problems. In each case the glibc maintainers haughtily dismissed these problems as not being bugs.
When I can write (C):
static FILE *arse = stdout;
And have it work fine with every other version of libc I've encountered, but with the newer glibc gcc gives me the same error message as if I wrote:
static FILE *arse = func();
Yet, the following works fine with the new glibc:
static FILE *arse = _IO_stdout;
I call that a bug, and lot of people agree with me. I don't care what is undefined in the ANSI specifications. The library maintainers should care about compatibility. In fact, I think they have moral and social responsibilities to take it very seriously.
A library maintainer who responds to this kind of backward compatibility problem with such outright derison as I've seen directed to others with this problem should be beaten with a stick before getting kicked off the project.
Bullshit. The companies I've moved over to Linux are SME's dying to get out of the lock Microsoft have on them with things like Windows Server, CAL's and then the client OS's. There's a lot of money in small business, and a lot of companies who've never heard of Unix who're ready to save a buck or two running Linux/Samba etc.
No.
Linux is a MAJOR threat to MS, and all of the major Linux vendors acting together would be *death* for MS.
Which is how the whole SCO thing is panning out. The big boys are realizing that there is no co-existance with Microsoft other than in the fashion that a slave "co-exists" with its master.
Because of Microsoft's predatory behavior, it'll either be hunted down like a man-eating tiger and killed by the members of the village, or it will simply eat the villagers one-by-one.
The MS "Linux Strategist" is hoping/dreaming that he can start some in-fighting among the villagers to prevent them from finishing the organization of the tiger hunt.
The boon/reward for cooperative behavior in this case, is that once the tiger is dead everyone will be able to freely compete for thier chunk of the 90% market share that MS currently controls on the desktop.
That's a powerfull incentive to cooperate against MS, and the examples of what has happened to every single one of the "Microsoft Partners" is a powerfull DISincentive for any of the big players to play nice with MS.
MS will be have to be incredibly lucky to get more than 10% to 20% market penetration on the desktop with Longhorn... and that will be the begining of thier end.
...it's the control, stupid! When MS Windows WILL NOT LET ME DO WHAT I NEED TO DO MY JOB and Linux will, guess which I will specify.
Oh, I thought it was a bank of monkeys surfing the MS Tech Web for the first tier suggested tips.
For second tier, all demoted MSFT contractors with marginal skills.
For third tier, all actual hard-working MSFT demoted employees with fair skills.
For fourth tier ($$$), all MSFT "non-employee, but should be employee"
For fifth "executive" tier (lifetime clients only), you talk to the developers directly.
-- Those are my experience only and not a reflection on anyone else. If your mileage varies (which can only get better)... stuff it.
Throw away an entire useless industry and something better ALWAYS comes along and replaces it.
-- I leave it to your mind to peg that "useless" industry.
Microsoft Linux 2005
Stolen from Open Source
Closed Source
Engineered to screw you into generation II of Microshaft.
Milk NT until it rots then pdo what Micrshaft has always done, borrow - steal and market. Lock'em in baby!
Besides, running two p690's, while attractive from an IBM sales point of view, is not exactly something we want to do.
And, on a side note, the "Linux is good for everything" attitude is exactly what is putting off so many people.
In our situation it's not clusters, but farms. Each node is completely independant but they do the same tasks.
You do realize of course that a p690 is what exactly what I was referring to as a "big iron" system. I'm talking Dell, HP Proliant, IBM xSeries type systems $7k. You may be doing some extremely niche fluid dynamics that require a single image; but the majority of what high-end Unix systems are being sold for do NOT require that, a small percentage do but not all. You take your one p690 and replace it with a bunch of nodes of *SMALLER* systems (not two p690's), enough to handle your entire load then throw in a few extra. You increase your availability by being able to lose a node, you increase your performance by adding some extra nodes, and your incremental growth costs are some inexpensive nodes you throw in once a year for growth (no more forklift upgrades that finance hates to hear).
Never said it was good for everything, but that we were able to prove to our management that we can get better: performance, reliability, scalability than the traditional monolithic systems. It seems that you don't want to admit that low class linux systems could possibly fullfill the same requirements as your p690 system but at a much better cost point.
Maybe to better illustrate the point is that you don't require linux, you could get some of the low-end IBM boxes, etc. and do the same procedure scaling horizontally rather than vertically. You won't get as an attractive price point but your availability, reliability increases significanly beyond what you have today with one system. You might say I've less of a "Linux is good for everything" attitude and more of a "horizontal scaling is good for almost all business, and combining Linux & x86 make it an easy dollar win". What's even better is now you could drop hardware support contracts, lease for 3 years and get a 3 year warranty, by some extra for spares and you've removed another significant dollar cost from your organization (provided you have semi-technical staff) repairs also happen much quicker than having to get parts expressed in from the vendor.
It may not work for everyone (extremely niche single image systems), but we've had extremely good luck in replacing our mainframes, our high-end Irix and high-end Solaris systems (high-end being 16+ procs) and getting much more performance (upto 2 fold per CPU) more reliability, and more availability and it cost our business much less.