Slashdot Mirror
Where Will IBM Drop Windows?
TurboProp writes "An article by the Associated Press on Friday (1/09/2004) Says that IBM has plans to abandon Microsoft operating systems on it's internal desktops by the end of 2005. The news originated from an internal IBM memo published by the Inquirer, a British technology news site. Further stories from the Inquirer, indicate that IBM May already have begun dumping windows. While this all bodes well for Linux users, and would seem to be a good PR move for IBM, executives at IBM seem to be trying frantically to put a much milder spin on the story. They say that the memo was taken out of context. I really can't imagine why they wouldn't be posting it on billboards."
..as would any organisation. No matter how much we might want them to get rid of Windows, they will only do so in areas it makes sense to do so.
In all likelihood, I would suspect that the vast majority of their servers already run Linux... Domino runs on Linux (and has for a while), and most of their webservers are likely to run Linux.
However, while desktops may get replaced, I would think that the engineers would be a prime target for Linux desktops well before, say, secretaries. But I wouldn't think it impossible that "where appropriate" will include Ms. Jane Q. Secretary in another year or two...
libertarianswag.com
First off, they won't be dropping it on their desktop or laptop lines anytime soon, so consumers can feel safe IBM is not abandoning them; Windows has a huge stake in IBM's business. However, it wouldn't surprise me to see IBM begin to try a move to desktops among their own company; however, I can state for fact this is nowhere near the implementation stage, there are a lot of barriers before this will become full-fledged across the number. IBM is too big a player to abandon Windows fully, but they may be making a little bit of a political statement in full support of their desktop, especially in light of the SCO case. IBM has their hands in so many honey pots, there should be no worries that the next ThinkPad you will get will be Linux-only.
If IBM drops Windows and adopts a Linux desktop, they're no longer tied to the x86 architecture. Of course, all the machines they have will still be useful, but what chip does IBM design and fabricate?
PowerPC
Now, I don't expect anything to happen immediately, but with the 970 aimed firmly at the desktop market, having a Linux desktop sure would make hardware migration a lot simpler--essentially transparent to the end-user...
My company uses two IBM-owned Windows-based products: Lotus Notes and Clearcase. (Yes, it is appropriate to feel sorry for me now) Will IBM continue to develop and support these products for Windows? Certainly. But will IBM but less priority on those products now that they're almost certainly not going to be using them in-house to the same degree? It's possible.
My company and many others don't want to hear that IBM is de-prioritizing their Windows products, and that's why IBM isn't making such a big deal out of this.
IBM is a very large organisation with a significant IT infrastructure. Apart from Microsoft, I suspect that every such organisation has a team somewhere looking at alternative desktop software vendors. But this is nothing to get excited about as such investigations, whether approved or not, have been going on for years. It used to be Apple, now it is Linux, who know what OS will be the hot alternative in a few years (Windows, I hope).
IBM is dropping Windows because it is a good business move not as a gimmick. They are not trying to impress Slashdot or open themselves up to abuse.
IBM is simply too big to not sell Windows so they don't want to make this rude. IBM needs to make money from the whole market to support IBM. They do not want the more childish Linux Element posting to the web every time they see a non Linux machine in the hands of one of their reps.
The move is hugely important because it will help them hammer out all the kinks in their understanding of the software. It will help them see where business oportunities lie, more than anything else. They will then be in a position to share how they do it with the rest of the world.
They are not doing it to insult Microsoft. They are doing it because they have more control and bigger margins when they sell Linux.
IBM will run _all_ prevalent operating systems as long as there is an IBM. They are just too big not to.
LS
It would be business suicide for IBM to ditch Windows on it's desktop systems in 2005.
While I'm far from an MS fanboi, Linux isn't ready to take the place of Windows on the desktop. It's not ready technically or graphically or from a design perspective, that is, the fluidity of the GUI. It's also not ready from a market perspective. Windows is the juggernaut. For IBM to ditch the market leader would be consigning their desktop division to massive losses.
I'd like to see a world where companies were free from market pressure to ship Windows or Linux or OS X and that all three could live equally in the computing world, sold to areas where they could leverage each OS's strengths.
But then I long for a world with ethical politicians and businessmen, where there will be peace in the middle east and that upper management will always make good decisions, so I don't think it's likely to happen.
My favorite thing about Internet journalism is the quality of writing:
"A quotable quote in the Wall Street Journal has an IBM spokeswoman claiming that it's...."
Riiiight. Is it just me or is the writing online much, much worse than in newspapers and magazines?
Ben
"I either want less corruption, or more chance
to participate in it." -- Ashleigh Brilliant
It is very simple, they actually want it both ways. They want to play up the fact that they are moving to Linux, but make it clear they are not abandoning Windows. Since they still sell Windows and sell consulting services for Windows customers, they want to make it clear they aren't biased against Windows if that is what the customer wants.
Comment removed based on user account deletion
IBM is not going to stop selling thier customers MS Windows hardware or services that want them. They are only trying to move as much of their internal systems to Linux. IMO, this could help leverage the monopoly playing field that MS has. Now, if IBM were to say that they will only sell Linux hardware and services, then yes, that would be stupid. However they are doing no such thing. I also think that this will be great for other companies out there that DO want Linux, but are a little worried about making a big change to their internel IT.
What in the world makes you think that IBM would be throwing away their MS Windows services just because they may end up using more Linux internally? IBM's internal IT structure has nothing in the world to do with their hardware sales or their global services.
If Tyranny and Oppression come to this land,
it will be in the guise of fighting a foreign enemy. -James Madison
I have it on good account that another major ($6 billion+ / year) software company will be dropping Windows O/Ss on their internal desktops within a year. Moving the SUSE, since Redhat dropped support for non-Enterprise versions.
I was doing an OEM deal for my employer at Research Tringle Park (RTP) near Raleigh North Carolina with IBM. RTP's entire server development group works out of RTP, I was shown off the blade center before it was functional etc. Oddly enough! No ethernet. Still token-ring, why give up on your own technology?! This great resource center of hard work and tech invention was most likely the ugliest most run-down facility one could possibly expect. I would say the time frame for an ENTIRE move from Microsoft desktops INTERNALLY is sometime after 2005.
if Longhorn is delayed longer than the early 2006 the best estimates have as of now;
If Apple & IBM make the Mac the fastest desktop during those years ( think dual core PPC980 based G6 @ starting >3.5 Ghz by this time in 2005 );
If Apple continues making OS X better at the same level and gets to true 64 bit-ness in time;
I think Apples market share will grow to 15~20% by the time Longhorn arrives.
A year ago, no one could even see Apple where it is now. They were universally dismissed as a cult and a joke. Now IMO they have the greatest mind share in computing right now, and it's only going to get better. HP decided to switch than fight the iPod, and if others get on that bandwagon, look out. Apple may actually win this time.
... in fact, they do like Israel does: they threaten to go full-opensource just to have a big rebate on their next round of Windows licencing !
..actions speak louder than words?
Maybe IBM realise that what they actually do counts for much more than whatever they announce loudly in a press release, or whatever "spin" they put on the news as reported by somebody else.
They get the benefit of guerilla advertising passed on the grapevine, and would presumably suffer *lots* less embarassment in the event of unexpected problems with deployment.
Ian
Will IBM start selling notebooks without the microsoft tax too? And with fully linux compatible hardware?
True genius is grasping a situation like a peice of fruit, and peircing it just right so that it drains dry.
This is exactly right. They're still embarassed and humilated by their OS/2 thing. When I worked at IBM in the mid 90's, everybody had OS/2 on their desktops, and used Word Perfect and Lotus Notes. It took a lot of gaul for them to use that stuff internally (I still think Notes was awesome, but Word Perfect was a pain in a Word world). So they've got the intestinal fortitude to pull this off, but they memory of their OS/2 humiliation is too fresh for them to fail to realize that it might not work out.
IBM could port Notes to Linux. They had native versions for OS/2 before. The server side is a console app, and the client side can be done on web browsers. I wonder if Project works under WINE? There's also the possibility of Citrix servers. You're right though, it'll be very difficult for them to be completely off of Windows by 2005. They did bend over backwards to get themselves on OS/2 though. But OS/2 had nice Windows emulation, which was key to their ability to do that. I've never played with WINE, perhaps it's as good as OS/2 warp was at the time.
Backlash
Microsoft shouldn't be expected to take something like this lying down. You can expect the folks in Redmond to dig into their bag of dirty marketing agreement tricks and find a way to punish IBM. Things like not giving them preferential pricing for installed Microsoft software unless IBM publicly renounces use of Linux on internel systems, that sort of thing.
Microsoft wasn't able to force IBM out of the OS and applications market by offering technically superior products; they did it by using their ability to set prices for their own products to punish anyone who dared to offer an alternative to Redmond. IBM can expect a price increase on Microsoft products (both those for internal use and for re-sale) as soon as they drop licenses for Microsoft products internally.
They that can give up essential liberty to obtain a little temporary safety deserve neither safety nor liberty.
Ben
IBM thinks about IBM first
1) the p970
this chip is a major contender, with apple helping with distribution and a probable p970 blade this year. they will need an os that can really use the chip (linux)
2) if you buy an IBM desktop your an idiot.
IBM is known for their servers (big iron) they don't really care about their small range servers they make their money on pseries and mainframes.
3) if linux can be specialized and modified for easily deployable office appliances and big servers than this could make IBM allot of money.
4) a linux version of OS IBM,
What if linux came out with a proprietary desktop, that could compete with sun, run linux and leverage the OSS movement. apple made that work so why not IBM think about it. would you pay 100$ for a linux based desktop with IBM backing it, has all the apps you need, and has a commercial interface.
This makes sense if you look at where IBM really makes its money, not desktops, not small servers, but the big guys.
Where I work, IBM makes loadsamoney on:
* Mainframes, mainframe software and mainframe support. DASD. Backup equipment (ATLs)
* RS6000s of various flavours (pSeries NUMA machines are $$$)
* AS400 still has a strong presence
* WebSphere (and MQ / MQ Integrator), DB2.
* Lotus Notes!
* Services - they charge mountains of cash if you want to outsource something to them. Ask them to tender, and weep. Then pay.
We don't use IBM PCs.
I reckon IBM make most of their cash in low volume / high margin. Linux is probably a way to turn the desktop into high volume / medium margin (support).
Yours Sincerely, Michael.
They were thinking that they weren't going to tie down their enterprise to a niche product that was before its time. IBM may be big, but technological hubris like that is what gave us such disasters as the Titanic. IBM has become a pragmatic company which thinks about more than making pointless ideological statements.
Now, Linux is more the sort of idea whose time has come. It has continued to thrive despite Windows, not instead of Windows, has built upon the mature foundations of Unix, and now is a well-developed (if still niche, at least on the desktop) platform. It actually makes sense for IBM to try this great experiment. It didn't make sense to do so with OS/2 Warp all those years ago.
Since I retired and went back to school, I have discovered large groups of people that use Excel as a general-purpose everybody-has-it environment for numerical computation. They make heavy use of tools like the embedded Visual Basic and Solver (general nonlinear constrained optimization). Some of those features, such as Solver, will be quite difficult to duplicate. If an OSS alternative uses a different algorithm, for example, it may have quite different convergence properties that cause the alternative to get different numerical solutions than Excel. Similarly, an embedded programming language that is almost-but-not-quite Visual Basic will break a large number of existing spreadsheet applications.
Are there OSS apps that provide "sophisticated" compatibility?
I was just thinking about companies (or countries) that cry "Open Source" as a means to get MS discounts. It's a good business strategy, perhaps, but it isn't good for F/OSS. Maybe these companies don't intend it as a bargaining strategy in the beginning; but if MS is able to convince people to renew their contracts with speacial deals it means that F/OSS isn't doing a good job of communicating its benefits (cost being only one factor of this).
I don't really think that IBM is toying around, but I am sure there are many smaller companies that would like to play that game.
They already committed in Munich. There's no going back. They either believe what they say or they don't take advantage of free software in house. They can't have it both ways.
The chances of failure, by the way, are slim to none.
Friends don't help friends install M$ junk.
According to the Mac News Network, independent sources put its latest sales figures of desktop systems at around 3.8% market share.
CNet News.com puts Apple's latest (3rd Quarter 2003) market share figures slightly lower: "Apple, meanwhile, saw shipments rise, but not as fast as the market. The company's U.S. market share is now 3 percent, while its worldwide share is below 3 percent."
While it is true that Apple has taken as much as 7% market share in the laptop market of the US in some months (see this MacLinks article) this only translates to 4% worldwide share. And HP is still the world laptop leader.
Apple, despite what you might think, has not significantly improved its market share over the last few years, except in laptops. You could make the argument that they have managed to keep market share from declining (which was the trend up until 1999), but they have failed to translate their new products into market share leadership.
I know it's hard to believe all of this when you get your news from Slashdot posters, but luckily industry analysts don't.
And, up until Photoshop 5.5,Adobe was writing Photoshop for UNIX. If it was running on UNIX, it wouldn't take much time to re-compile it for Linux. Likewise, QuarkXPress was ORIGINALLY released on UNIX, because neither M$ or Apple would release the API's to Quark to program for those platforms. UNIX documentation is overly abundant, so there is no excuse for these companies to NOT support it. Once you have UNIX support, LInux isn't that far behind. Now, we just need all Linux Distributors to follow the Linux Standards Base perfectly and make it easier for thes commercial companies to start for Linux. Once we have a standard to agree on, everyone can follow behind and get some real work done.
The music is all around us. I can hear it. Can you?
SOunds very familiar.
:
... your statements sound alot like someting a decade ago involving new risc more powerfull powerpc's and powermacs vs delays in WIndows95 called chicago back then.
... then and maybe then %20 of all new pc will be risc powerpc based with macOS/OS/2 real soon!
"IMO it depends on three things
if Longhorn is delayed longer than the early 2006 the best estimates have as of now;
If Apple & IBM make the Mac the fastest desktop during those years ( think dual core PPC980 based G6 @ starting >3.5 Ghz by this time in 2005 );
If Apple continues making OS X better at the same level and gets to true 64 bit-ness in time;
I think Apples market share will grow to 15~20% by the time Longhorn arrives.
"
Flashback to 94/95
1.) If Chicago aka Windows95 is delayed longer and summer 1995 is the best estimate.
2.) If apple made clones and increased its marketshare
3.) IBM continues to make os/2 available to the mac and powerpc for true 32-bit trueness
4.)If Motorolla gets those powerfull 604 processors and makes them 133mhz by 1996
I hated MS with alot more passion back then because their products were much much more flakier and unbearable back then. Does anyone remember how to get out of an infinite loop in Windows 3.1?
Hit the reset button. What if it took 5 minutes to log in due to a crappy network configuration? 3 infinite loops in your program cost you 15 minutes of time! incredible! This was my highschool by the way a decade ago.
I assumed the world would switch to OS/2 and MacOS to escape this hell that is Windows/DOS and these primptive CISC processors.
You know what?
THe world chose Windows and pentiums got faster. Same will happen here. It already is happening with AMD64.
Windows will be here forever and ever and will never go away. Not to sound trollish but how many million upon billions of VB code and MCF C++ code is lyeing around offices world wide? How many corps invested millions to upgrade their networks for Active Directory?
As hobbiests its one thing But IBM cores market is WIndows and they want a single platform they can support. Now which will have they chose already? You get the picture.
Same in 94 and same in 2004! Hell more so today with legacy Windows systems around.
They do not want Linux. Companies use Linux secretly or because they need stability or security. WIndows is improving and with palladium will be secure enough to kick it out. Stability wise Windows2k and 2k3 are as stable as Unix. Ask anyone who adminstered any MS boxes. Yes NT4 was not.
IBM needs Windows and will be hurt without it. Why oh why did they make that horrible deal with Microsoft 25 years ago? why?? They are still stuck and can not leave.
http://saveie6.com/
I'm afraid we're still talking about different groups (and the areas where we work reflect this):
:)
We use the Aventail client for VPN connectivity from the customer site. Yes, I know there's a Linux version. No one at our site has yet gotten it to work.
That's something that about six thousand IBM employees where I work don't use at all ever.
To get into the client's network, I need to have access to the VPN they use
again, not applicable to the tens of thousands (hundred thousand?) IBMers that never visit customer sites -- actually everything I've said and we don't agree with applies to this group.
YES, emphaticly. As I mentioned before, our client likes Microsoft Project.
But you see, here you're talking about switching IBM's customers to Linux, not internal deployment of Linux.
I work in hardware development, and even the hard-core researchers and developers maybe use Matlab (which does run on UNIX -- not positive about Linux versions), Notes (everyone uses that!), and a web browser for purchase orders, and administrative stuff in general -- that's it. Maybe you know tens of thousands of people where the cocerns you listed apply, but not where I work nor for 50, 60, 150 thousand (I really have no idea here) IBMers.
That's also whom I refer to in terms of remote administration. IBM IT (internal IT, the armies that serve our computers, upgrades, and all the people our internal standard software installer services) would be whooping with joy if they saw what they could do remotely with Linux (from the other side of the country even!) -- heck for most administrative tasks the thundering horde of engineers, researchers, secretaries, managers, lab technitians, operators wouldn't even know that an admin is at that moment updating several packages on their computers as they work (except with a kernel update, but that's about the only exception).
Like our internal AIX support groups (for Catia design, mostly, and for some of the hard-core computer modeling we do in our development) are aware of what a difference it is to work with an OS that has had remote administration with mind from the beginning. I remember having issues with an AIX box running Catia when it was first setup for me (no root access for me! But that's OK, I'd probably screw up the box seeing if I could get xmms running or something), it was eerie to have the person on the other side of the phone saying stuff like "I'm in your computer right now, OK, I'm renaming these and these files, I'm editing some config stuff that's broken, ok, I'm about to restart your program -- cool looking desing, BTW, what are you working on?" That's just simply impossible with Windows, and we all know what calls are like with that OS! " ok, click here, what's is say? Ok, now look here, you'll see a button. Press it. What's it say now?". No competition -- if I were in charge of computers within IBM, I know I'd be pushing hard for Linux just for the remote administration. When IBM's customers also switch (and the world runs X), can you imagine how much nicer your life will also be?
Of course, if you are logging on to customer's sites (and IBM's, IMHO, good in that regard; the customer's always right!) -- you will deal with what they use and an internal deployment of Linux by IBM won't affect them.
Internal deployment of Linux makes a ton of sense fo the people I described -- and yes, I place POs, do my PBCs, access my benefits page all within Mozilla and within Linux.
I have to deal with Excell, Word, and PDFs, and that's about the only thing (recently, more esoteric electronic design file format called Gerber, and the wine that comes with C4EB dealt with the "windows only" install.exe file identically as it would have worked within Windows, click, install and use -- I never even saw a command line!)
I understand, and you have very eloquently stated why Linux simply won't work for you. But, it has worked for me for a yea
I can think of one excellent reason:
Presume they want to have Linux running on most boxes, but not all (theres plenty of potential reasons). But the media (and places like slashdot) wrongly perceive that they are trying to run entirely on Linux.
Imagine the damage if word got out that they still had Windows running on a few directors' boxes. It would undermine everything about the switch, and give great ammunition to the pro-Windows argument (eg: "See, even Linux-friendly IBM still needs Windows!")
"And that solves the mystery of the missing ring" - Bender
We run test cases, and check that we get the same results as last time. We do this with all safety related closed source programs, when each new version comes out.
i tt lebit))),dodgy result, correctresult)
There is no way that we could read enough of OO's source to check that it was always telling the truth, we'd still have to test it. And every time we got a new version, we'd have to check all that source code all over again.
I agree, to some extent, that we shouldn't use Excel for safety related stuff, but, not for your reasons. Spreadsheets are inherently uncheckable - any random cell in a 30 Mb sheet could include the following pseudocode
=if(and((somecell>anumber),(somecell(anumber+al
The way I get around that is to write robust cells that handle all exceptions, and copy them down the whole sheet. Also, most of my work is correlation based (ie testing analytical models against real data)- so algorythmic errors would tend to show up.
Matlab would probably be a better bet.
Perhaps a more fundamental question is why do I trust Excel more than OO?
Well, if you see the discussions of OO here they usually claim it is a reasonable Excel substitute - despite its obvious lack of speed, stability and features. If the people making those claims are being honest then they can't be pushing it very hard. Therefore they aren't really qualified to comment.
I'm a bit puzzled by people's problems with various levels of Excel, I use 97 at home and a couple of different versions at work, my sheets and VB seems to work fine on all 3 installations.
There again my stuff is big rather than fancy.