While CDE is the old standby, I would guess that very soon now both the KDE and Gnome user base is going to be larger than CDE's. There just isn't that many Unix workstations out there relative to people playing with and using Linux. In fact, FWVM may already have a larger user base than CDE just to it being in RedHat. --
The Halloween document was correct on at least one point: Usability isn't something that can be easily 'grafted on' to an otherwise finished system, any more than security or stability.
This is a very valid point - unfortunately, the whole KDE vs Gnome thing often gets boiled down to a pretty widget issue (see any comments on themes or icon sizes.)
For those of you who might not have touched a Macintosh in ten years, here are several technical features which are key to it's usablity. I'd like to see the KDE or Gnome projects rip some of these ideas off.
* You can drag an application or alias anywhere and it will still work. This works because the Finder automatically updates a "desktop database", so the system can go back and find that app. This eliminates many common problems newbie Windows users have until they figure not to touch anything. (Obviously a Unix implementation will need to accomote security and path settings, but it could be done.)
* Drivers load directly out of the "Extensions" and "Control Panels" directories, not some initilization file. If you are having problems with a driver, just drag it out of the folder, or drag an updated one in. (This is a "designed in" feature which is possibly impossible to do with unix. However, it would be nice to see a good virutalization.)
* Every file stores a "Type" and "Creator" value, which means you can have some JPEGs which open in a viewer, and others which open in Photoshop. Makes a double-click interface much easier for the user, because the right application always opens. (There should be no reason one couldn't do this in Unix, if a standard could be developed.)
* All applicaitions store their icons, windows, dialog boxes, text strings, etc. in a standard resource format. (Which is why Macs have the wierd two fork files that no other OS has.) Anyone can launch the ResEdit applicaiton and customize the GuI for their apps. I would assume that QT/Gnome has something similar, it's just not exposed to the user.
* Drag-n-Drop just works. For example, decompressng a file involves just dragging it to the Stuffit Expander icon. Compare this to WinZip or the KDE equivalent.
The MacOS is lacking in many places, but these little bits and pieces make it nicer to work with from a mouse-wielding user level. I haven't tried Gnome, but using KDE, CDE, LinuxConf, or even MS Windows, you always feel as if you are working with a similcrum of somehthing else which may or may not be decieving you, whereas the Mac is super predictiable enough from the mouse pointer level that you can actually feel comfortable there.
Agreed. Although in the Linux world, if you asked "Run unknown macro Y/N?", most would probably answer a resounding no. So there is a "security fix" in MS Office, it just relies on the great unwashed users to press the right button.
Thanks for being one of the few to point out that Linux is not somehow magically immune to this sort of virus, just because it's not from Microsoft and it doesn't have 90% of the market. The bottom line is that Linux doesn't have a built-in certificate infrastructure or code-signing mechinism either (like Lotus Notes does), unless you count a PGP sig here and there. Many of you would still run something off of freshmeat even if that sig wasn't there. I have.
The one thing Linux has going for it is that right now the projects are being run by engineers for engineers, so security is likely a higher consideration. Microsoft has been berated by complaints about the macro situation in Microsoft Office for years, and there is no solution in sight. The nice thing about Melissa is that now they're getting the complaints from the CIO rather than from some random system admin. This has been a huge PR hit for Microsoft, which they deserve for letting the marketing dorks choose the feature set.
The fact of life with modern computer systems is that code and data are getting mixed up all the time, and furthermore are getting pushed down to the client, and who's to say that KDE or GNOME or even emacs is immune from this sort of attack. If Microsoft is to blame, it's because they've lead this movement to push the logic down from the back end system to the client platform. The Linux's community's response has been to pick up that client-side ball and run with it.
Wait until someone releases a simlar virus that exploits emacs/pine/mutt/sendmail or whatever. Totally possible. Ooops - it's your fault for using Linux. Too bad, screw you.
Besides, it wasn't a Windows exploit, it was a MS Office exploit. Windows and IE actually have a certificate checking routine for ActiveX (although iit's pretty braindead), whereas Office has no security infrastructure, just some dumb Yes/No question. --
Throughout our industry, you'll find many, many examples of people who were "there in the beginning" who are unable to make the shift as a company becomes successful.
Except, that it's arguable whether Netscape has been successful, or if they have been riding the success of the early versions.
He points this out when he says that Netscape doesn't have the managerial or engineering talent to pull it off. And if you look at it, it seems reasonable - they've bombed in the server market (hope Sun can help there), and they turned their flagship product into such a POS that they've had to start over from scratch. Not even to mention all the crazy half-done products they started to only later drop.
I wish Netscape a lot of luck, because they're looking like they need it.
IE 5 has been released, and it's heckava lot faster and more stable than Netscape 4.5 on Windows32.
If you've got your nose in Linux or another Unix all day, you may well think "Oh, Netscape 4.5 isn't *that* bad", but it really is a big stinky dog compared to what you can get on Windows, esp. with IE.
As far as Netscape not being that far behind, maybe that's true for lowest-common-denominator Internet browsing, but for the really interactive CSS/DHTML stuff that you can get away with on an Intranet (where you can develop for one client), IE is way ahead.
Std Disclaimer: Sure IE has proprietary features, but so is any feature of Netscape/Moziila that no other browser has. At one time frames and tables were a proprietary Netscape feature.
The Circut City connection is why Divx will ultimately fail. It's a bizarre plan to push a new "standard" that's available through only one retail outlet.
Especially since (the article mentioned) that Circuit City has raised the price of their DVD players to make Divx seem more competititve. So, they are basically losing sales in order to push this crap.
Customer - "Why is this DVD player $100 more expensive than at Best Buy?"
Circuit City Salesman - "Because we have Divx which is better"
When Lotus 1-2-3 or MS Word came out in the early eighties, I'm sure most thought word processors and spreadsheets would never reach the status of commodity/free software. (Gnumeric is free or even the mighty 1-2-3 is sometimes discounted down to $50 or less.)
Enough corporations are pouring money into ERPs that sooner or later someone is going to scratch that itch. For example, wait until some university department decides that they need Peoplesoft (which isn't an ERP, but is still big $) but can't afford it. It wouldn't be that hard to homebrew something. --
Anyway, then, people who saw (Linus') kernel said, "Well Lets see what we can put together with this kernel, see if we can make a whole system." Low and behold, everything they needed was there. What they found was the GNU System. It wasn't all written by us, because since we had a higher order goal, "We're gonna make a whole system", which means, probably hundreds of programs.
By "GNU System", he's pretty clearly talking about the HURD + GNU user space stuff + everything else project. He's not refering to the Linux kernel.
Yes, but if the "GNU System" didn't happen to exist when Linus developed his kernel, there probably wouldn't be anything called a "Linux OS Distribution".
Rather you'd see things like "FreeBSD with the Linux kernel", that is, if anyone even bothered. I doubt the kernel would have been developed as extensively if it was an alternative on another OS distribution.
(I don't call it GNU/Linux myself, I just think that RMS made a pretty good case that the Linux project used a lot of code from the GNU Not-Unix project. Of course, RMS gave all this code away under the GPL, which is why he's a hero. Instead he gets called names and is treated dismissively on slashdot.) --
This one always gets me -- You want your operating system to recognize more than 640K - go spend $50 for QEMM. It hardly seems outragous for a memory manager to be "integrated" in the OS (unlike, say, a web browser).
This kind of news offers hope to the folks who think Microsoft might be indestructable. 10 years ago, they had no application market to speak of except on the Mac; they had to give Windows away for free for anyone to use it; and DR had reverse-engineered their goldmine (which was only a reverse-engineered version of their goldmine, to be fair).
(Before DR-DOS, there never was a notion of "upgrading" your PC's OS. You pretty much just ran the customized version of MS-DOS 3.2, or whatever, that came with it.)
Now Microsoft pretty much owns the application and OS markets, has 50% of the e-mail market, and is trying to nip at Oracle and IBM's heels in the big league database market. A lot can happen in 10 years in this industry, it'll be interesting to see what happens.
Har Har. My guess is that in 1989, the average workplace desktop was not on a LAN, or if it was, it was a disconnected departmental network. Of course the CEO of Microsoft was hardly using a average computer.
The guy who mentioned MS's VAX system probably got it right - Gates might have typed this message into a terminal emulator. Or MS Mail with a VAXmail gateway or something. --
MUST READ, MUST READ NEAL STEPHENSON!
on
Tuesday Quickies
·
· Score: 1
Right on. An excellent appeal to the power user communities in the Mac and Windows worlds. (esp. the Mac one) --
I fully agree with you that NT is probably not less expensive to own than Novell or Unix. (Hence "supposedly"), and as someone whose a production unix/novell/microsoft admin, I'm aware of the pluses/minues of each platform, aside from the marketing issues.
However, you can go get numerous studies telling you that NT is cheaper to support than either Novell or Unix, by a significant amount.
Can anyone point to a study that indicates supporting Linux is cheaper than supporting NT? Maybe for a web server or a DNS server, but how about for a complex environment (file and print, mail, database) across multiple sites? Again, I'm not sure - I'd like to see some data?
Do you really think a Unix administrator is the same price as an NT administrator with similar skills? Call your local temp agency and find out. I have -- Unix admins are 2x the rate of NT admins.
It's interesting that other collaboration/e-mail packages such as Lotus Notes and Eudora are unaffected by these problems....
It would be almost trivial to recreate this kind of virus in Lotus Notes. Yet it's largely unheard of, despite Notes being the #1 commercial e-mail system. (I'm saying this not to encourage anyone, but to get Notes users to tighten their systems.)
The one protection that Notes gives you is that all code is cryptographically signed, and the user or administrator can define the rights code have depending on it's signature. In a properly set up environment, Internet orginated Notes macro viruses don't have a chance. However, many Notes shops have not implemented this feature.
Eudora/Outlook Express/Pine/etc. don't really have a automation interface to speak of, so they are probably relatively immune. I'm not an emacs wiz, but it may be possible there due to the lisp interpreter, although I doubt anyone would bother.
(sigh, why am I always the guy who is trying to stop the lynch mob.)
Note that Microsoft promptly released a patch for this problem, soon after it was publically announced.
As for "buried in their website", I hardly call http://officeupdate.microsoft.com/word/ buried.
If Microsoft is going to be damned on this one, it's for unleashing a full programming environment on 90% of the word processor users out there who have absolutely no need for it. The small number of people that do use MS Office based macro applications could be bothered to run a seperate installer to get the VBA environment. Of course a groupware package like Outlook needs a programming environment, so maybe Microsoft should take a look at how Lotus Notes minimizes/elimiates this problem (through a certificate infrastructure).
--
Good News for PC Vendors, Bad News for IBM and Sun
on
SAP invests in Red Hat
·
· Score: 2
Right now, if you were to look at the applications that move hardware for the big mini/mainframe vendors such as IBM, Sun, and HP, I would guess that way up on the list is ERP packages like SAP, Baan, and JC Edwards as well as PeopleSoft.
Sure all of these packages theoretically run on NT, but right now that's sorta the Loss Leader to get the customers in the door. Once there, they figure pretty quickly that NT (and MS SQL) is not up to the task.
Enter Linux - Possibly a way to get these huge applications to run on relatively cheap hardware. The big win is going to be for Compaq and Dell, and the folks who can get the integration services done.
Of course, a couple years down the road, when IA64 is established, there's going to be little or no reason to run these big apps on Solaris/IA64 or HP-UX/IA64 versus Linux/IA64. The big minicomputer vendors are running head on into the commodtized hardware market, where there's a perfectly good and free OS waiting for them. It's going to be interesting to see if they can stay in business. --
Right. Companies don't like paying for "service and support". For every $500 NT licence that a company buys, it spends thousands of dollars supporting that NT box.
Linux advocates may scoff at the "supportablity" of NT, but one of the reasons that NT has taken over the workgroup server market its that it is supposedly much cheaper to support than Novell or Unix.
Linux has solved the $500 software licence problem, but I'm not sure they've solved the $50,000 "service and support" issue. People want "easier" and "cheaper", in most cases much more than they want "better". --
And what's cheaper still is to have general motors produce a production line version of cars with sunroofs.
This argument has more to do with the difference of business models between (say) Sun and Microsoft than it does with Linux
Sun will charge you $50 for a GUI PPP program or $10,000 for their application server.
Microsoft makes a crappier PPP program and application server, but their both free with the OS licence.
The Linux "market" (I don't mean students and hackers!)involves waiting around until someone writes a GUI PPP program or application server, because if you really, really needed it you'd probably buy it off the shelf before you'd write it yourself.
If Linux takes hold, the consulting possibilties are huge to fill in these missing pieces. And it will probably be mostly commercial. (Take a look at the price of a NT disk quota package. It's pretty high, but it hasn't stopped many from installing NT. Again, if a company needs it, they'll pay for it. Human Nature will take care of the rest.)
But all of this Linux consulting is all essentially lost consulting buisness for the NT/Novell/Unix shops. At one time, a good chunk of EDS (which is #1 system consulting/outsourcing, I think) was Certified Novell. Now they are all Certified Microsoft. You don't think they would quickly get Certified RedHat, if the market started to shift?
Slashdot Mirror
While CDE is the old standby, I would guess that very soon now both the KDE and Gnome user base is going to be larger than CDE's. There just isn't that many Unix workstations out there relative to people playing with and using Linux. In fact, FWVM may already have a larger user base than CDE just to it being in RedHat.
--
The Halloween document was correct on at least one point: Usability isn't something that can be easily 'grafted on' to an otherwise finished system, any more than security or stability.
This is a very valid point - unfortunately, the whole KDE vs Gnome thing often gets boiled down to a pretty widget issue (see any comments on themes or icon sizes.)
For those of you who might not have touched a Macintosh in ten years, here are several technical features which are key to it's usablity. I'd like to see the KDE or Gnome projects rip some of these ideas off.
* You can drag an application or alias anywhere and it will still work. This works because the Finder automatically updates a "desktop database", so the system can go back and find that app. This eliminates many common problems newbie Windows users have until they figure not to touch anything. (Obviously a Unix implementation will need to accomote security and path settings, but it could be done.)
* Drivers load directly out of the "Extensions" and "Control Panels" directories, not some initilization file. If you are having problems with a driver, just drag it out of the folder, or drag an updated one in. (This is a "designed in" feature which is possibly impossible to do with unix. However, it would be nice to see a good virutalization.)
* Every file stores a "Type" and "Creator" value, which means you can have some JPEGs which open in a viewer, and others which open in Photoshop. Makes a double-click interface much easier for the user, because the right application always opens. (There should be no reason one couldn't do this in Unix, if a standard could be developed.)
* All applicaitions store their icons, windows, dialog boxes, text strings, etc. in a standard resource format. (Which is why Macs have the wierd two fork files that no other OS has.) Anyone can launch the ResEdit applicaiton and customize the GuI for their apps. I would assume that QT/Gnome has something similar, it's just not exposed to the user.
* Drag-n-Drop just works. For example, decompressng a file involves just dragging it to the Stuffit Expander icon. Compare this to WinZip or the KDE equivalent.
The MacOS is lacking in many places, but these little bits and pieces make it nicer to work with from a mouse-wielding user level. I haven't tried Gnome, but using KDE, CDE, LinuxConf, or even MS Windows, you always feel as if you are working with a similcrum of somehthing else which may or may not be decieving you, whereas the Mac is super predictiable enough from the mouse pointer level that you can actually feel comfortable there.
--
That's a Netscape IMAP bug.
--
And currently none of the browsers available are standards compliant.
If nobody follows the standards, they aren't really "standards" are they. I guess you could call them "W3C Feature Proposals".
--
Agreed. Although in the Linux world, if you asked "Run unknown macro Y/N?", most would probably answer a resounding no. So there is a "security fix" in MS Office, it just relies on the great unwashed users to press the right button.
--
Seems more stable to me. "Active Desktop" hasn't crashed once on me since it's release, which was a daily problem on this hairball NT box here.
--
Thanks for being one of the few to point out that Linux is not somehow magically immune to this sort of virus, just because it's not from Microsoft and it doesn't have 90% of the market. The bottom line is that Linux doesn't have a built-in certificate infrastructure or code-signing mechinism either (like Lotus Notes does), unless you count a PGP sig here and there. Many of you would still run something off of freshmeat even if that sig wasn't there. I have.
The one thing Linux has going for it is that right now the projects are being run by engineers for engineers, so security is likely a higher consideration. Microsoft has been berated by complaints about the macro situation in Microsoft Office for years, and there is no solution in sight. The nice thing about Melissa is that now they're getting the complaints from the CIO rather than from some random system admin. This has been a huge PR hit for Microsoft, which they deserve for letting the marketing dorks choose the feature set.
The fact of life with modern computer systems is that code and data are getting mixed up all the time, and furthermore are getting pushed down to the client, and who's to say that KDE or GNOME or even emacs is immune from this sort of attack. If Microsoft is to blame, it's because they've lead this movement to push the logic down from the back end system to the client platform. The Linux's community's response has been to pick up that client-side ball and run with it.
--
That's What You Get For Using Windows
Wait until someone releases a simlar virus that exploits emacs/pine/mutt/sendmail or whatever. Totally possible. Ooops - it's your fault for using Linux. Too bad, screw you.
Besides, it wasn't a Windows exploit, it was a MS Office exploit. Windows and IE actually have a certificate checking routine for ActiveX (although iit's pretty braindead), whereas Office has no security infrastructure, just some dumb Yes/No question.
--
Throughout our industry, you'll find many, many examples of people who were "there in the beginning" who are unable to make the shift as a company becomes successful.
Except, that it's arguable whether Netscape has been successful, or if they have been riding the success of the early versions.
He points this out when he says that Netscape doesn't have the managerial or engineering talent to pull it off. And if you look at it, it seems reasonable - they've bombed in the server market (hope Sun can help there), and they turned their flagship product into such a POS that they've had to start over from scratch. Not even to mention all the crazy half-done products they started to only later drop.
I wish Netscape a lot of luck, because they're looking like they need it.
--
IE 5 has been released, and it's heckava lot faster and more stable than Netscape 4.5 on Windows32.
If you've got your nose in Linux or another Unix all day, you may well think "Oh, Netscape 4.5 isn't *that* bad", but it really is a big stinky dog compared to what you can get on Windows, esp. with IE.
As far as Netscape not being that far behind, maybe that's true for lowest-common-denominator Internet browsing, but for the really interactive CSS/DHTML stuff that you can get away with on an Intranet (where you can develop for one client), IE is way ahead.
Std Disclaimer: Sure IE has proprietary features, but so is any feature of Netscape/Moziila that no other browser has. At one time frames and tables were a proprietary Netscape feature.
--
The Circut City connection is why Divx will ultimately fail. It's a bizarre plan to push a new "standard" that's available through only one retail outlet.
Especially since (the article mentioned) that Circuit City has raised the price of their DVD players to make Divx seem more competititve. So, they are basically losing sales in order to push this crap.
Customer - "Why is this DVD player $100 more expensive than at Best Buy?"
Circuit City Salesman - "Because we have Divx which is better"
Customer - "Huh?"
--
When Lotus 1-2-3 or MS Word came out in the early eighties, I'm sure most thought word processors and spreadsheets would never reach the status of commodity/free software. (Gnumeric is free or even the mighty 1-2-3 is sometimes discounted down to $50 or less.)
Enough corporations are pouring money into ERPs that sooner or later someone is going to scratch that itch. For example, wait until some university department decides that they need Peoplesoft (which isn't an ERP, but is still big $) but can't afford it. It wouldn't be that hard to homebrew something.
--
Anyway, then, people who saw (Linus') kernel said, "Well Lets see what we can put together with this kernel, see if we can make a whole system." Low and behold, everything they needed was there. What they found was the GNU System. It wasn't all written by us, because since we had a higher order goal, "We're gonna make a whole system", which means, probably hundreds of programs.
By "GNU System", he's pretty clearly talking about the HURD + GNU user space stuff + everything else project. He's not refering to the Linux kernel.
--
Yes, but if the "GNU System" didn't happen to exist when Linus developed his kernel, there probably wouldn't be anything called a "Linux OS Distribution".
Rather you'd see things like "FreeBSD with the Linux kernel", that is, if anyone even bothered. I doubt the kernel would have been developed as extensively if it was an alternative on another OS distribution.
(I don't call it GNU/Linux myself, I just think that RMS made a pretty good case that the Linux project used a lot of code from the GNU Not-Unix project. Of course, RMS gave all this code away under the GPL, which is why he's a hero. Instead he gets called names and is treated dismissively on slashdot.)
--
This one always gets me -- You want your operating system to recognize more than 640K - go spend $50 for QEMM. It hardly seems outragous for a memory manager to be "integrated" in the OS (unlike, say, a web browser).
This kind of news offers hope to the folks who think Microsoft might be indestructable. 10 years ago, they had no application market to speak of except on the Mac; they had to give Windows away for free for anyone to use it; and DR had reverse-engineered their goldmine (which was only a reverse-engineered version of their goldmine, to be fair).
(Before DR-DOS, there never was a notion of "upgrading" your PC's OS. You pretty much just ran the customized version of MS-DOS 3.2, or whatever, that came with it.)
Now Microsoft pretty much owns the application and OS markets, has 50% of the e-mail market, and is trying to nip at Oracle and IBM's heels in the big league database market. A lot can happen in 10 years in this industry, it'll be interesting to see what happens.
--
Har Har. My guess is that in 1989, the average workplace desktop was not on a LAN, or if it was, it was a disconnected departmental network. Of course the CEO of Microsoft was hardly using a average computer.
The guy who mentioned MS's VAX system probably got it right - Gates might have typed this message into a terminal emulator. Or MS Mail with a VAXmail gateway or something.
--
Right on. An excellent appeal to the power user communities in the Mac and Windows worlds. (esp. the Mac one)
--
I fully agree with you that NT is probably not less expensive to own than Novell or Unix. (Hence "supposedly"), and as someone whose a production unix/novell/microsoft admin, I'm aware of the pluses/minues of each platform, aside from the marketing issues.
However, you can go get numerous studies telling you that NT is cheaper to support than either Novell or Unix, by a significant amount.
Can anyone point to a study that indicates supporting Linux is cheaper than supporting NT? Maybe for a web server or a DNS server, but how about for a complex environment (file and print, mail, database) across multiple sites? Again, I'm not sure - I'd like to see some data?
Do you really think a Unix administrator is the same price as an NT administrator with similar skills? Call your local temp agency and find out. I have -- Unix admins are 2x the rate of NT admins.
--
Can you say for sure that Groupwise/Wordperfect is immune, or just that nobody really cares about writing viruses for 3% of the market?
In the old days, there were WordPerfect and 1-2-3 macro viruses. I don't think anythings technically changed to stop them.
--
http://msdn.microsoft.com/library/specs/richtex
--
It's interesting that other collaboration/e-mail packages such as Lotus Notes and Eudora are unaffected by these problems....
It would be almost trivial to recreate this kind of virus in Lotus Notes. Yet it's largely unheard of, despite Notes being the #1 commercial e-mail system. (I'm saying this not to encourage anyone, but to get Notes users to tighten their systems.)
The one protection that Notes gives you is that all code is cryptographically signed, and the user or administrator can define the rights code have depending on it's signature. In a properly set up environment, Internet orginated Notes macro viruses don't have a chance. However, many Notes shops have not implemented this feature.
Eudora/Outlook Express/Pine/etc. don't really have a automation interface to speak of, so they are probably relatively immune. I'm not an emacs wiz, but it may be possible there due to the lisp interpreter, although I doubt anyone would bother.
--
(sigh, why am I always the guy who is trying to stop the lynch mob.)
Note that Microsoft promptly released a patch for this problem, soon after it was publically announced.
As for "buried in their website", I hardly call http://officeupdate.microsoft.com/word/ buried.
If Microsoft is going to be damned on this one, it's for unleashing a full programming environment on 90% of the word processor users out there who have absolutely no need for it. The small number of people that do use MS Office based macro applications could be bothered to run a seperate installer to get the VBA environment. Of course a groupware package like Outlook needs a programming environment, so maybe Microsoft should take a look at how Lotus Notes minimizes/elimiates this problem (through a certificate infrastructure).
--
Right now, if you were to look at the applications that move hardware for the big mini/mainframe vendors such as IBM, Sun, and HP, I would guess that way up on the list is ERP packages like SAP, Baan, and JC Edwards as well as PeopleSoft.
Sure all of these packages theoretically run on NT, but right now that's sorta the Loss Leader to get the customers in the door. Once there, they figure pretty quickly that NT (and MS SQL) is not up to the task.
Enter Linux - Possibly a way to get these huge applications to run on relatively cheap hardware. The big win is going to be for Compaq and Dell, and the folks who can get the integration services done.
Of course, a couple years down the road, when IA64 is established, there's going to be little or no reason to run these big apps on Solaris/IA64 or HP-UX/IA64 versus Linux/IA64. The big minicomputer vendors are running head on into the commodtized hardware market, where there's a perfectly good and free OS waiting for them. It's going to be interesting to see if they can stay in business.
--
Right. Companies don't like paying for "service and support". For every $500 NT licence that a company buys, it spends thousands of dollars supporting that NT box.
Linux advocates may scoff at the "supportablity" of NT, but one of the reasons that NT has taken over the workgroup server market its that it is supposedly much cheaper to support than Novell or Unix.
Linux has solved the $500 software licence problem, but I'm not sure they've solved the $50,000 "service and support" issue. People want "easier" and "cheaper", in most cases much more than they want "better".
--
And what's cheaper still is to have general motors produce a production line version of cars with sunroofs.
This argument has more to do with the difference of business models between (say) Sun and Microsoft than it does with Linux
Sun will charge you $50 for a GUI PPP program or $10,000 for their application server.
Microsoft makes a crappier PPP program and application server, but their both free with the OS licence.
The Linux "market" (I don't mean students and hackers!)involves waiting around until someone writes a GUI PPP program or application server, because if you really, really needed it you'd probably buy it off the shelf before you'd write it yourself.
If Linux takes hold, the consulting possibilties are huge to fill in these missing pieces. And it will probably be mostly commercial. (Take a look at the price of a NT disk quota package. It's pretty high, but it hasn't stopped many from installing NT. Again, if a company needs it, they'll pay for it. Human Nature will take care of the rest.)
But all of this Linux consulting is all essentially lost consulting buisness for the NT/Novell/Unix shops. At one time, a good chunk of EDS (which is #1 system consulting/outsourcing, I think) was Certified Novell. Now they are all Certified Microsoft. You don't think they would quickly get Certified RedHat, if the market started to shift?
--