Jupiter Research analyst Joe Wilcox believes that Microsoft’s upcoming major OS release will not be generally available until 2006, a year later than what Microsoft is currently predicting. He believes that the changes planned are too far-sweeping to be rolled out quickly, when you take into account that developers need time to prepare their software.
But I wonder if this will mean that there will be a “Windows XP Second Edition” in the mean time.
Could Longhorn be Microsoft’s Copeland? If so, we can probably expect to see a release no earlier than 2013
how much GNU/linux and FreeBSD progress by that time….I bet a lot, so MSFT better sharpen their act by that time because the competition will be fiery.
So where does this leave all those corporate customers who are paying the annual license fees?
This long delay is probably the reason Microsoft changed their licensing scheme. With no new OS to sell and no reason for anyone to upgrade Office, they need to make money somehow.
Man that would piss off a very large number of corporate folks, and rightly so. However, I do see many of them, lining up to take it up the ass again after this current crop of licenses expire. People are funny that way.
Could this be a contributing factor as to who Apple is creating a new enterprise sales group?
http://www.thinksecret.com/news/enterprisegroup.html
… the licenses dont expire. They just end up paying for Software Assurance for a period in which nothing new gets released (And you can’t buy SA except with the orginal license purchase so you can’t skip it for 2? years when you know Longhorn won’t be out till after that)
So how much does SA cost for Windows XP and is it better to not pay for it (twice) and just buy new Longhorn licenses when it actually comes out? And what can you make on that money by not spending it up front?
Maybe ti will actually be a solid release, by taht time it better be.
Still, I don’t think there is a need for WXP SE.
how much GNU/linux and FreeBSD progress by that time….I bet a lot, so MSFT better sharpen their act by that time because the competition will be fiery.
Why would linux and BSD progress at all, they can just whine to the government and start another “antitrust” trial if they don’t do so well.
The last antitrust was a stage show. I hope the next one is real.
Yeah, and Arthur C. Clarke predicted we’d have a fully functional space station and moon base by 2001. By 2010, he believes we will be able to safely travel to Jupiter and beyond. Hello?!? Future predictions are exponentially wrong based on the time difference ebtween the current date and when the event is predicted to occur. One year away is fairly inaccurate, three years is much worse.
So typical of them.
So what else is new. This isn’t news.
The antitrust trial was crap, MSFT was guilty, all the evidence proved it, it just turns out that there are a lot of scare dpussies in the government.
how is it possible that someone who does not work at Microsoft predict the software release better then Microsoft themselves. We have no idea how much technology has been implemented and just not interfaced (‘connected’) to the reset of the OS under-system.
He doesn’t, he’s just talking out of his arse.
But anyway here’s a really cool link about how microsoft works on windows development:
http://www.pcquest.com/content/technology/101120604.asp
and here’s an (weird) interview with Dave Cutler:
http://www.microsoft.com/windows2000/server/evaluation/news/fromms/…
Windoze and .Not is just too buggy to make any progress so M$ changes the gui shell with barbie pictures so people will buy it.
Good, Mac should be on OS 10.6 or 7 by then.. and Longhorn will still be trailing 10.3… sounds like it’ll take them at least 11 years to catch up to Apple… again.
(11 years = Time between 1984 (first mac) and 1995 (windows 95))
Maybe Microsoft will include X-windows with this new version of Windows, so I can run LyX and Fluxbox on Windows without downloading so much.
Probably not, Longhorn is supposed to be a step forward, not backwards.
The release of Longhorn is well over 2 years away (MS is always about a year late with Windows releases). Consider that 2 years is how long it took the KDE project to totally rewrite KDE between 1.0 and 2.0. 2 years is how long it took to go from kernel 2.4 to kernel 2.6. From a technology standpoint, Linux and KDE are already on a par with many aspects of Longhorn. It would not at all be surprised if KDE and Linux have a significant edge on Longhorn by the time it comes out.
PS> I’m less optimistic about GNOME. GNOME has an edge on KDE from a polish/UI standpoint, but the core technology isn’t as good. In comparison, KDE’s core technology rocks. Everything is componentized. Everything is DCOP scriptable. Qt and KDE offer some very powerful application development frameworks. Code reuse (and thus consistency and integration) is extensive. The power of the technology shows from how quickly KDE and its applications have been developing.
Good, Mac should be on OS 10.6 or 7 by then.. and Longhorn will still be trailing 10.3… sounds like it’ll take them at least 11 years to catch up to Apple… again.
(11 years = Time between 1984 (first mac) and 1995 (windows 95))
MS caught apple in OS design with Windows 3.0. It had a cooperative POS kernel much like the MacOS at the time. UI wasn’t as good but MS pulled away from everyone with OS sales regardless.
Kernel wise MS buried Apple when they released NT. Apple caught up with NT 6 years later when they released OS X.
It goes back and forth. Almost like a swing set! Weee!
Apple has never been able to compete in the coorporate environment. Apple’s “Enterprise Sales Group” currently has only 16 employees, hardly a “fighting force of extra ordinary magnitude.”
If Apple wants to compete with the likes of Dell and IBM, they need to look at the whole enterprise, from the $400 departmental servers to the $20,000 blade servers. The X-Server and X-Raid is a good start but I don’t think it is enough.
From a technology standpoint, Linux and KDE are already on a par with many aspects of Longhorn.
What aspects are you refering to?
Why would linux and BSD progress at all, they can just whine to the government and start another “antitrust” trial if they don’t do so well.
Learn some history please. The Antitrust trial was over netscape which at the time was a commercial company.
Learn some history please. The Antitrust trial was over netscape which at the time was a commercial company.
Hey assclown, where did I say the last “antitrust” trial WASN’T about netscab?
Linux will most certain progress in the next 2-3 years, but probably not so much that it would actually matter.
Even if there was no Longhorn until 2006 and there was no Windows XP SE, the only thing that’s going to stay stagnent in the next 2-3 years is Windows. The apps will continue to progress, which is really the only thing that matters, since that is the big draw to Windows in the first place.
thank god
Nothing new as usual 1-2 yers late
Kingston: But I wonder if this will mean that there will be a “Windows XP Second Edition” in the mean time.
They call it SP2. Remember now that Windows XP also serves the Professional market, having a pay-for interim release like Windows 98 SE wouldn’t be wise. Would probably be as successful as Windows Me in market.
Could Longhorn be Microsoft’s Copeland? If so, we can probably expect to see a release no earlier than 2013
Apple’s Copland was never release. And while Microsoft slips on every major release of any of their products (from MSN to Office), they don’t slip into the decades.
This long delay is probably the reason Microsoft changed their licensing scheme. With no new OS to sell and no reason for anyone to upgrade Office, they need to make money somehow.
You may not find any use of Office 2003, but a lot of people, principally medium to large organizations and those who wrestle a lot with information (principally numbers) in documents would want Office 2003. Office 2003 is in many ways a bigger release than Office XP (and some would say Office 2000).
As for money flowing in, as long PCs are sold, new companies are opening up and existing ones expand, Microsoft would get money. Especially since most of their Windows revenue comes from OEM and corporate sales, not the retail.
DCMonkey: … the licenses dont expire. They just end up paying for Software Assurance for a period in which nothing new gets released
I must be dreaming about that release of Windows 2003 Server… (they use servers too).
linut: Why would linux and BSD progress at all, they can just whine to the government and start another “antitrust” trial if they don’t do so well.
So true.
The antitrust trial was crap, MSFT was guilty, all the evidence proved it, it just turns out that there are a lot of scare dpussies in the government.
A lot of people was proven guilty as witched in those days in Salem… I guess you don’t understand antitrust laws. It has no clear-cut defination of an anti-competitive behaviour. It is all in the opinion of the judge. One of the big mistakes of the trial is to gloss over the “harmed” competitor’s business skills, which in my opinion the thing that cased them to tank.
me: Good, Mac should be on OS 10.6 or 7 by then.. and Longhorn will still be trailing 10.3… sounds like it’ll take them at least 11 years to catch up to Apple… again.
Yes, because of the never-seen-yet Aero, Longhorn is just some clone of Panther? Hello? If they wanted to do that, all they needed to do is licenses DesktopX from Stardock and viola – they can release it now as oppose 3 years from now. Everything else from WinFS to Palladium – completely forgotten. Heck, they can buy over Stardock, intergrate everything (except maybe the ability to theme) into Windows, hire some really expensive artist and viola – something that should look a thousand times faster than OS X. Surely save a whole lot of trouble, no?
From a technology standpoint, Linux and KDE are already on a par with many aspects of Longhorn.
But then again, from what we know from all the WinSuperSite and BetaNews expose – not quite. The last I check, Linux still doesn’t have a database FS – but that’s okay, because KDE still haven’t fully maximizes of features XFS (and ReiserFS 4) has yet anyway.
It would not at all be surprised if KDE and Linux have a significant edge on Longhorn by the time it comes out.
I would certainly hope it that way. But then again, most people wouldn’t care, as they do now. Most people moving to Linux now (read: corporations/organizations – the enterprise) aren’t going for it because of features rather stability (run even Windows XP boxes on a large network and see what I mean), security and of course price standpoint.
I’m not saying Linux is free, but in the longterm, it sure is a heck a lot more cheaper.
MoronPeeceeUsr: Kernel wise MS buried Apple when they released NT. Apple caught up with NT 6 years later when they released OS X.
Then again, Windows 3.5 wasn’t really any good like UNIX… or its cousins OS/2 and OpenVMS.
Richard James: Learn some history please. The Antitrust trial was over netscape which at the time was a commercial company.
I doubt he meant that comment seriously, rather a snide sacarstic comment on Microsoft suing competitors.
Apple has never been able to compete in the coorporate environment. Apple’s “Enterprise Sales Group” currently has only 16 employees, hardly a “fighting force of extra ordinary magnitude.”
Do you know something about Apple that we don’t? secondly, most companies put together working groups to see if a particular idea will work. How do we know that this “group” is mearly to see the viability of them entering the enterprise?
If Apple wants to compete with the likes of Dell and IBM, they need to look at the whole enterprise, from the $400 departmental servers to the $20,000 blade servers. The X-Server and X-Raid is a good start but I don’t think it is enough.
Maybe there could be PPC 970 servers around the corner to co-incide with the release of Panther?
Personally, Apple should start firstly in the small business and work their way up rather than aiming for the top branch of the tree and continuously falling.
Kernel: 2.4 is already ahead of the NT kernel. If Microsoft wants to compete with 2.8 (which will be out by the time Longhorn comes out!) it will have to make drastic improvements to the NT kernel. The press-info shown so far do not indicate that they are making such improvements.
Filesystem: Reiser3 and XFS are already better than NTFS. Reiser4 leapfrogs all three of these. Plus, Reiser4 is designed from the ground-up to support small, database item-like files. NTFS, which WinFS is built on top of, was not.
Component technology: KParts and DCOP are used much more extensively by KDE apps than COM and WSH are by Windows apps. The result is integration that’s not just superficial (like the MS Office toolkit looking like the regular Windows toolkit) but reaches down to the roots of the system.
The only thing really missing is the HW-accelerated desktop, which might very well pop up (via xwin.org or XOuvert) by the time Longhorn comes out.
2.4 is already ahead of the NT kernel
How So? Please provide technical details.
Microsoft is famous for announcing things before they’re ready, and they’re counting on you to think that will be the case here so you can be taken unawares. In reality they’re going to come out with something Linux cannot imitate, and it’s going to hit unexpectedly soon.
>>DCMonkey: … the licenses dont expire. They just end up >>paying for Software Assurance for a period in which >>nothing new gets released
>I must be dreaming about that release of Windows 2003 >Server… (they use servers too).
Yes, but I was just refering to desktop OS releases. SA is per product AFAIK. Or do they make you buy SA for all licenses (server and desktop) if you buy it for one?
1) Filesystem. Reiser3 and XFS are faster than NTFS. Reiser4 is faster than all three, and is built from the ground up for database-like usage.
2) I/O scheduler. I have both kernel 2.6 and WinXP on my machine, and I have to say the I/O scheduler in 2.6 rocks. If a process is doing heavy I/O in WinXP, it can really slow down the system. In Linux, the system response degrades gracefully as I/O load increases.
3) Process scheduler. The Windows scheduler is full of hacks to improve GUI responsiveness. For example, it gives temporary priority boosts to certain processes. The 2.6 scheduler tries to provide GUI responsiveness the *right* way, through an algorithm that tries to recognize interactive applications and schedule them properly. Plus, it can handle hundreds of thousands of threads, while the NT scheduler chokes at several thousand (and BeOS choked at about 400 on my PII-300!)
4) VFS. Supports tons of different filesystems natively.
5) VM. The WinNT VM is a mess. Its fully of heuristics that can explode in a firey mess under certain conditions. In comparison, the Linux VM achives memory balance as a result of algorithms. Proper algorithms are harder to write than using heuristics, but are much simpler and more reliable under widely varying loads.
6) Networking: Tons of protocols, built-in firewall, support for a huge array of services.
There’s probably more that I forget. Just read the case studies in “Modern Operating Systems” to get an idea of the real weaknesses of the Windows NT kernel. It refers to Windows 2000, but not much as changed in XP.
The only thing really missing is the HW-accelerated desktop, which might very well pop up (via xwin.org or XOuvert) by the time Longhorn comes out.
I would be more inclined to think that some derivitive of the TransluXent hack would be a good bet for a HW-accelerated desktop for Linux and BSD. It looked cool, but it definately needs more work, and more than one developer!
http://www.stud.uni-karlsruhe.de/~unk6/transluxent/
Take a look, I’m sure it’s been covered on this site before.
I doubt he meant that comment seriously, rather a snide sacarstic comment on Microsoft suing competitors.
No sarcasm involved. MS competitors are much more lawyer happy than MS, and usually only towards MS. Don’t recall Apple sueing amiga for the GUI, or Eola sueing the other browsers (of course they may, but they may not, and MS already WAS sued..)
Why would linux and BSD progress at all, they can just whine to the government and start another “antitrust” trial if they don’t do so well.
>>>>>>>>>>>>>>
I’m going to ignore the question of whether Microsoft deserved the anti-trust lawsuit. I think they did (and apparently, so did the DOJ, who knows a hell of a lot more about monopoly law than you do!), because that can be debated seperately. I’m more concerned that you are insinuating that the Linux/BSD people would resort to “dirty tactics” to avoid competing directly with Microsoft. As it stands, its Microsoft that’s resorting to dirty tactics. The Linux folks are competing purely on product quality and price/performance.
Evidence:
– Microsoft’s lies about the GPL: Steve Ballmer has refered to it as a “cancer.” Jim Allchin has refered to it as “un-American.” In one of its EULAs, Microsoft implies that any code created with GPL’ed development tools falls under the GPL. There is just tons more, from Microsoft directly and from Microsoft-funded studies. Most are just not true.
– Microsoft buying licenses from SCO. The timing of this was crucial. Even if Microsoft isn’t behind the whole thing, they’re trying to legitimize SCO’s claims to the public. However, this situation is ridiculous, because most large companies, including major US government projects, don’t seem to be paying one bit of attention to SCO.
– Microsoft lobbying governments to restrict the use of OSS products, through the “Initiative for Software Choice.”
I thought this whole article was about the prediction that Longhorn will be released a year later than MS says… I didn’t see a line in the text comparing Windows to Linux…
Anyway.
I think we shouldn’t judge any piece of software before it’s gone gold. You can’t judge the caracteristics of the new Ferarri by looking at the scale model made of clay, now can you?
It are the people who do judge software (read: Windows) before it’s released that irritate me. It’s good with me if you don’t like Windows, BeOS, Mac OS X or any other OS. But PLEASE, some of you really try so hard to prove that MS sucks, it kind of seems like a religion to me…
Go live on a deserted island somewhere, and don’t get in the way of people who judge software with pure objectiveness. You people spoil our fun.
Being objective does not mean that you have to ignore past history, nor current information about the status of the project. You might not be able to judge a new Ferarri by a clay model, but you can be pretty sure it will be a kick-ass car. Similarly, if you predict that next year’s Pontiac Grand-Am is going to have too much plastic cladding and poor interior materials, then you’d probably be right again…
Consider this: There is no way Longhorn will be any better than what MS says its going to be. From a hype-machine company like Microsoft, this just won’t happen. If it was going to be better, they’d turn the hype up a notch. (Am I the only one who remembers the Cairo hype circa 1994?)Microsoft is not making any statements about how they are making sweeping improvements to the kernel. Most likely, then, they won’t be. They’re not talking about new object models or more powerful integration/scripting techniques, or a faster filesystem, or anything like that. Most likely, then, you can bet that Longhorn won’t be getting such things.
Don’t think I finished my thought
Anyway, by comparing the current state of Linux, and taking into account clearly-defined future progress (we already have a good idea of what Qt/KDE 4.0 and kernel 2.8 will be all about, the dark-horse is really X) and comparing that to Microsoft’s blue-sky vision of Longhorn (taking into account their previous disparity between vision and reality) you can get a damn good picture of where things will stand in 2006.
You said
2.4 is already ahead of the NT kernel
But then you go on to compare NT to 2.6. Please stick with your original comparison.
Filesystem: Reiser3 and XFS are faster than NTFS.
Can you point to any benchmarks the prove this? What are reasons for the performance improvements?
Reiser4 is faster than all three, and is built from the ground up for database-like usage.
What does this mean? NTFS wasn’t built with data integrity as a design element? (Refer to page 700 of ‘Inside Microsoft Windows 2000’). Do Reiser4 or XFS has native support for file encryption, Quotas, Change logging, ACLs? NTFS isn’t perfect, but please provide some technical details on why the other file-systems are better.
I/O scheduler: Have you actually performed any benchmarks to confirm this? Is it possible there are other variables affecting the result?
Process scheduler: Have you actually seen the code to confirm that there are hacks or are you just guessing?
The “Hack” you are talking about isn’t a hack but a design consideration. It’s called the foreground quantum boost. It doesn’t effect the priority level of the foreground threads, but instead increases the amount of time they are allowed to run, before the scheduler interrupts them. This can be easily disabled. W2k will boost the priority of threads under certain conditions. This again is a design consideration that deals with things like CPU starvation, GUI threads waking up due to windowing activity, etc. The NT process(Thread) scheduler does support different priority levels for threads (32 to be exact). You can actually change the base priority of a process through task manager.
Plus, it can handle hundreds of thousands of threads, while the NT scheduler chokes at several thousand (and BeOS choked at about 400 on my PII-300!)
Show me the data from tests to prove this. What exactly do you mean by choked? What were the threads doing? Was the hardware identical? Are you comparing NT kernel threads to userland threads?
Aside from that, creating hundreds of thousands of threads will result in a performance degradation if there are not enough processors to run them on. This is why heavily threaded server apps in NT/Windows 2000 don’t perform as well as apps that implement thread pools. At some point, you will start to see diminished returns.
VFS: I agree that NT/W2k doesn’t support as many file-systems natively. However, the NT file-system driver specification is documented and you can write your own driver for your favourite File-system if you like.
VM:
VM. The WinNT VM is a mess. Its fully of heuristics that can explode in a firey mess under certain conditions. In comparison, the Linux VM achives memory balance as a result of algorithms. Proper algorithms are harder to write than using heuristics, but are much simpler and more reliable under widely varying loads.
Under what conditions? Provide details.
Algorithms and heuristic are not mutually exclusive. A more accurate description would be that the NT VM contains heuristic algorithms. Which means that the algorithms try to discover or learn the optimal path by evaluating and understanding the meaning of certain conditions (Engineering Cycle v. Empirical Cycle). I’m not sure if this is actually true, but your simplification doesn’t provide any insight into what the problems with the NT VM subsystem are. I’m not saying that there are not any, or that Linux isn’t better in this regard, only that what you have said is meaningless and provides no basis for your assertion.
Networking: NT/W2K also supports “Tons of Protocols”. I will agree that the Firewall capabilities in NT/W2K are weak, but W2K does provide native support for IPSEC. The huge array of services (NNTP, SMTP, TELNET, etc??) are userland services and have nothing to do with the kernel.
I’m not saying that NT doesn’t have its problems. It does. But to argue that the Linux 2.4 kernel is better by comparing NT to the feature set of the 2.6 kernel is simply wrong. If you make the assertion that the Linux kernel (any version) is better, then provide an actual reason why. Provide benchmark data or some technical descriptions of both implementations and why one is better than the other. I’m not being sarcastic. I’m really interested in the technical reasons why one is better than the other.
Try Google.
WinZealot.
OMG! You are teh winner of this argument for sure. Be sure to claim your prize.
“We’re not going to put something out there just to meet some date.” -Jim Allchin
Hehe.
“However, the NT file-system driver specification is documented and you can write your own driver for your favourite File-system if you like.
”
Does it mean that one can write “drivers” to support, for example, ext2fs, as NTFS or VFAT fs ? If so, why doesn’t it exist ( I really cannot see any reason ) ? Could you give me some links which talk about that ?
By the way, could it be possible to speak about MS without having stupid comments from some linux zealots who really don’t know what they are talking about ( talking about thread support of the kernel 2.4 is… funny, as threads are process with an other name, to simplify. 2.6 is much better on that point, though, but not programs on linux still don’t use thread so much ) : they harm linux more than they helpo them. Linux for sure has some strong points over NT kernel, but the contrary is true, too. I really don’t know how you can compare NTFS with XFS : with the NTFS driver for linux ? Or with a magic program which supports natively XFS with NT ?
From a technology standpoint, Linux and KDE are already on a par with many aspects of Longhorn.
What aspects are you refering to?
They both have mouse support.
[i]Then again, Windows 3.5 wasn’t really any good like UNIX… or its cousins OS/2 and OpenVMS</>
OS/2 was a fine OS. I agree.
Techically speaking, NT 3.5 was a vast improvement over the MacOS. Thats what I was making reference to. Apple had a version of UNIX during this timeframe but it wasn’t supported near as much as NT (even though NT support sucked in the beginning)
VMS I could care less about. Digital was a brilliant company with poor management the likes of Commodore Business machines. Rest in piece.
Component technology: KParts and DCOP are used much more extensively by KDE apps than COM and WSH are by Windows apps. The result is integration that’s not just superficial (like the MS Office toolkit looking like the regular Windows toolkit) but reaches down to the roots of the system.
People still interface with COM ? I thought everyone who had any brains had moved to .net components by now.
COM is dated and MS is already trying to forget about it.
What do KParts and DCOP being used more extensively than COM have to do with the technical merits of the technology ? I’m sure when I was creating COM components I could have used them more extensively and wasted a good chunk of time exposing everything as an interface.
So what ?
WinXP/2K is ahead of any other OS…
btw, do you think unix OSs will finally got rid of the X disaster by 2006 ?
Do you think PPCcomputers will be fast enough for MacOSX to scroll windows contents “fast” enough ?
Hum…
Leo.
“btw, do you think unix OSs will finally got rid of the X disaster by 2006 ?”
Sounds good, doesn’t it? “No more X Windows.”
Damn, it’s time for a replacement.
Can somebody please tell me what the bad thing about X is? No, not “the network transparency makes it slow” which is not true, no, something else.
Besides, on my P350 webpages in Konqueror do scroll smoothly, even with anti-aliased fonts! The only thing that is bad is IMO the lack of transparency, which can be implemented within X, as seen by this URL posted earlier:
http://www.stud.uni-karlsruhe.de/~unk6/transluxent/
Additionally, throwing away will also hurt Linux/BSD, as Gtk+ and QT will have to be rewritten for the new windowing system, which means they will improve *less*, not more.
Instead, I think development should focus on making hard things easier, for example auto-things at USB device attachment (mounting mass storage, starting Kooka for scanners and such) and maybe a really nice application-based frontend to apt-get and pkg_add.
And no, driver installation also isn’t very important, as you either install drivers as a package, and if you don’t, they should work *automatically*.
Does it mean that one can write “drivers” to support, for example, ext2fs, as NTFS or VFAT fs ? If so, why doesn’t it exist ( I really cannot see any reason ) ? Could you give me some links which talk about that ? <?i>
It probably has a lot to do with the fact that the development kit costs $900 and includes a license agreement that may restrict developers in some way.
http://msdn.microsoft.com/library/default.asp?url=/library/en-us/gs…
explains that the Installable File System Kit documents the requirements for developing a file system driver or filter.
[i]File system driver
A file system driver handles I/O independent of any underlying physical device. File system drivers include drivers for the system-supplied NTFS and file allocation table (FAT) file systems. On the NT-based operating system, file system drivers are kernel-mode drivers. File system filter drivers provide additional capabilities above a standard file system driver, typically by implementing such value-added services as virus screening.
(from the above link)
come back OS2 all is forgiven
because Microsoft didn’t set a final release date yet, right?
I just love to hate that company. They’re so entertaining.
Thanks for the link, but the license agreement link is dead… Try to figure it out. But it would be so great to have
ext2/ext3 support for windows… ( Or real NTFS support for
linux ).
It is not possible to write one driver and distribute it without fee, I assume.
What version of Windows do we use until then?
“”Can somebody please tell me what the bad thing about X is?””
xlib
I am no windows fan, but why do people hate Linux
1. One word “XFree86”, but Xcouvert has made a promising start and hope it continues with the same tempo
2. “Next – Next” software installation. Agreed RPM is a grear way to install software once you where to get the requiredlibraries. Till then pray to GOD.
Frankly, once somebody works on Linux for a month or two, I’m sure he wont turn back, But its needs to be reduced from a month to a hour:)
Anybody ready to bet that Longhorn will be the waterloo of Microsoft?
It’s a just a hack, how are you going to implement true alpha blending between windows that are over each other when the server doesn’t even remember what is underneath. It’s an architectural problem, so is the lack of responsivness due to x crap event management.
Darth Gates: When will Longhorn be ready?
Sinister Henchman: Do you want to use it or sell it?
DG: To sell it, of course, you fool!
SH: It will be ready on time, Evil One.
Try Google.
WinZealot.
Was this directed at me?? If so, what part of any of my posts make me a zealot? Please provide details. I merely suggested that reasonable technical details be provided to backup the claims that the Linux 2.4 kernel is already ahead of the NT kernel. Instead, you insist that I should be responsible for proving the assertion that I DID NOT make. I’m looking for an objective technical discussion on why one kernel is better then the other. Your response is to call me a WinZealot. That must have required and enormous amount of grey matter. Thank you for adding your invaluable insight into the discussion.
<Your logic>
DOS is a better server OS than SunOS because you’re a zealot
</Your logic>
Honestly, If there’s anyone here that can provide some accurate technical information about why the Linux 2.4 kernel is better than the NT kernel, I would love to hear it. I’m not being sarcastic. I really want to know.
linux kernel is _pure_ monolithic crap by design(2.4 or 2.6 doesn’t really matter) but is micropatched by people from all around the world. Linux is toy for geeks(doesn’t make it bad though).
NT kernel is on the other hand one of best by design(academic one, good though) but isn’t properly implemented because you can have all the money in the world(MS) but they won’t stand for zillions of eyes looking if comment is written properly for example =).
*BDS kernel is also ahead of linux, historical reasons also(academic origin), good one also.
Solaris is winner for multi-processor architecture kernel design.
@others:
i. you have ext2/3 on windows as you have many others, just look for e.g. total commander plugins.
ii. MS doesn’t advertise kernel improvements because that sort of advertisment won’t sell the product more or less. Customers(product managers etc.) reading Longhorn advancemnts are required to hear/read another kind of improvements which MS understands and gives. That doesn’t mean that there aren’t any kernel improvements, they are _many_ imho most than since NT4.0 SP3. And also Longhorn would be a serious API/code cleanup, biggest since… ehm?
iii. COM is still used extensively in windows programming model enviroment although not directly but through ATL/WTL template toolkits. All in all, it’s still ongoing updated technology and has much cleaner design than say MFC. And will stay longer(at least until MS decides to do very clever move and next winXP SP2 and win2k SP5 would include .NET framework). Most of huge MS products are COM based and aren’t going to be rewritten to .NET anytime soon. And COM is also your good choice when looking to create win32 application. And I don’t think KDE’s object model(whether it’s Kparts or Qt’s one) is used more due to many reasons(less applications and so on).
iv. ReiserFS4 is really cool.
v. Win2k3 are as fast as freshly installed WinXP with everything turned off(Win2K3 is that way by default). Just to put all that review-FUD in balance.
waterloo maybe not, but something more like the seige of stalingrad IMHO.
Long drawn out slow decline. Novell comes to mind here.
In 2010, Microsoft will exist, but it will be far weaker than today.
Just my humble opinion, and anytime you speculate about the future you’re taking a stretch. We also have other major changes coming up, optical comuters, quantum computers, and IMHO, OSS will likely jump on this very quickly.
Microsoft can only sell their OS for a profit as long as OS’s doesn’t become a commodity, so how long do YOU think that will last?
In 2010, Microsoft will exist, but it will be far weaker than today.
Far weaker in the server market? Yeah. Farweaker in the desktop market? Maybe. Far weaker overall? Unlikely. Microsoft already entered the handheld market, got a significant amount of market with MSN (and who knows, maybe get some profit out of that!), and may get some market power from Sony with XBOX 2.
Once Microsoft stops moving, you can count it death. Business is like a bicycle – you stop, you would fall out. You need to go on.
1) Why are there so many anti-Microsoft, pro-Linux and pro-MacOS X posts even though there is nothing in the article about these two competing products?
2) There has been no information regarding Longhorn. The information that has been spreading around the internet are baiscally educated guesses based on what Microsoft has released publicly and what their current products look like plus a little guesstimation based on wild ideas regarding what people would like to see.
3) Microsoft isn’t going to die. Many people said that about IBM 15years ago, however, 15years later IBM is stronger than ever. Even if the operating system becomes a commodity item, the value isn’t in the operating system but the strong intergration between the different products and the operating system itself.
Basically, Microsoft would become the software equivilant of IBM, using a commodity operating system as a vehicle for their middleware business. IBM is doing the same thing.
Ultimately in 5-10 years time, what operating system is used will ultimately be a non-issue. The battle front won’t be operating system but middleware.
Linux is usable, but it needs someone to focus it. Redhat is going is the right direction with Bluecurve; Lycoris is too. Linux needs standards. Choice is nice, but it’s counter productive. I don’t need 6 different email clients installed automatically. Just give me the best defaults, or a way to make the best default, and an easy way to install other apps when I want to. Be did a lot of things right.
Longhorn MS’s waterloo? Not hardly, that was Win98.
One more thing, a Zealot free vacuum would be nice too, since we’re talking about what Linux is missing.