Slashdot Mirror
Wintel "Thin" Servers to Compete with Linux
GenePrescott
sent us a news.com story that talks about Microsoft and Intel
working together on a
thin server thrust.
They're going to try to use thin appliance type servers
to compete with the Threat that Linux poses to them. Interesting
article. Interesting tactic. Not sure if it'll work.
Nuff Said
Yet, Penguin Computing and VA Research are still selling 8x Xeon boxes. A better description of the 'Linux Effect' might be: the computing muscle needed to get the job done, no more, no less.
A family member is a developer for a company that dominates their specialty, non-consumer, database-oriented field. They are using MTS three-tiered database architecture. The concept is really great. I've talked about the possibility of using Linux for server(s) in this kind of setup. Of course, you could run the database server on Linux because high-end SQL solutions are available for it. The client-side stuff can conceiveably be run on Linux machines (through browsers). However, the middle-tier is (so far as I know) pure MS territory. For reasons concerning the development market and the kind of solutions the end-users want, it just doesn't make sense to try to desperately piece together some complete system that will run on anything: Linux, *BSD, other UNICES, Windows... What I don't think will happen is an *easy* way to do this. MS just won't play ball. And they can't just "not use MS's nonstandard standards." It's not a matter of being biased toward MS, it's just that MS is the only solution that works readily for the market. If anyone can suggest ways to get around this stuff, it would be interesting to know about it.
--posting anonymously becuase I'm not sure how safe it is for people to know what I'm talking about!
"Several differences:
You get all the benefits of open source on the engine, including bug fixes, ports to niche/old
hardware, new cool features, etc. These increase sales of your game files, and can improve your
game's marketing (your ads and demos show BeOS engine clips, while competing engines can only
afford a Win9x port and Win9x clips).
Others can write games for your engine, too, and sell them. That sounds like a disadvantage, since
you're giving away the engine for others to make money off of. But since it's your engine, your brand gets an advantage, since everyone else's box will be advertising your brand for you, thus giving your game files a competitive advantage. Mindshare is a powerful thing - just ask Red Hat. "
So I'm a little company that has some really innovative software ideas for a game. So we spend a couple million on the code and art work. Then we release this super game. So then the whole entertainment industry jumps on it and sinks $100 million into art work, while we could only afford $1million for ours...our artwork is no better than the rest of the clones out their, and hence we have little competitive advantage. Hence we create the possibility of all of the income from the game, but we hardly get any of it. Yaa. Sounds like a self-sustaining industry to me. Sounds real easy for a software company to start up with fresh ideas...
I hate it when people point to redhat as a shining example of open source at work. Redhat is a leach. They didn't pay for more than a tiny fraction of the development of Linux. Of course they're making money! Let's see a company invest it's own money to create a product and release it open source. Where are those companies?
Netscape? Nope, sells proprietary servers.
Cygnus? Nope, sells proprietry stuff.
Sendmail? Nope...
who does it? where are the examples? All I see are lots of leaches sucking away...
Furthermore, I think that a great deal of the work on Linux is funded by the texpayers of the world, or by private companies through the salaries of the people who work on Linux. A great deal of these people work for companies that produce proprietary software. I honestly have yet to hear a clear vision of a self sustaining marketplace based on open source.
" You can sell the games files separately from the game. If a user has the engine, buying new files can be a click-and-drag thing with the Internet. Of course, since it's your engine, your game files are displayed more prominently than others - if you even display others. Or, alternately, you can sell your CDs with a downloader for the engine, saving precious CD space. "
Major flaw here. Why is creating and distributing art any different than code. You've just put forth a big double standard. You're saying that it's ok for the users to trade the code, but not the actual game. Or did you mean to say that users could also trade their games? Ooops, so much for selling the artwork...
But who will write the new engines?
Your average small bussness has a large problem getting skilled, qualified system admins. The solution of having your friend set up Linux doesn't work very well because someone needs to administer and maintain the server.
A thin server gives the business a functional file or print server with very little administration costs. If they need to upgrade to a faster, bigger (more memory/storage) thin client they just buy a new one, or maybe trade it in. The hardware may cost a bit more, but hardware's cheap compared to hiring admins, and especially compared to the cost of downtime due to admin mistakes.
However running NT or Linux on the thin server doesn't make much sense to me. You don't need the overhead of seperate user and kernel space (especially when Linux doubble buffers data between them). You don't need to protect against poorly writen user applications. VxWorks excells in this environment, and the price is much more reasonable.
However, NT thin clients will still sell, because managers recognize the names Microsoft and NT, and unfortunately, that's how a lot of business decisions are made.
... Intel liked GNU/Linux, because it helped them sell high-end. They certainly have no none-x86 competition in the thin client arena now that they've bought StrongARM.
Thin servers will be shipped preinstalled and preconfigured on proprietary hardware, with a customization program that may be run on a web browser to reconfigure it. Therefore, ease of installation is absolutely NOT an issue. If somebody really wants a Windows interface, you can write a Windows App to configure the thin server remotely, or use SNMP with an existing Windows-based SNMP manager. The PHB need never know what OS is running on the box iteself, his management interface still looks like Windows to him! (See http://www.watchguard.com/ for an example of a box with embedded Linux) Windows has zero advantages in this market, and it just is not going to do to well against the already established boxes, many of which are running Linux.
This announcement is just a marketing gimmick to keep their names in the paper.
They'll get distracted by something else instead and forget about it.
Transported to a surreal landscape, a young girl kills the first woman she meets and then teams up with three complete strangers to kill again.
Uh, that would be the plot synopsis for The Wizard of Oz, for those of you that didn't catch it...
You sound silly. May be slashdot should run on 486 too?
"I think I've read both from Linus and AC that the kernel is coded entirely in assembly, and if I'm not mistaken that is very platform specific... Think of how long it too to release the Alpha version of the kernel. "
Ahem. The "whole" kernel is not written in assembly. A great deal of it is written in platform independant C that calls platform specific assembly routines. That's why it is relatively easy to port to new platforms...in the ideal case it's just a matter of creating a new set of asm routines and creating the relevant low level structures for that platform. In more complex cases I imagine some small modifications could be required to the general routines, but only if the new hardware does something that none of the other platforms could do.
DDE
OLE
COM
DCOM
Why have you not leant the obvious lesson. Don't develop application on Microsoft "Standards"
Cobalt has been around, what? 2 years or so. And only now does Microsoft *plan* on developing a comperable product. But of course, what OS would it run? NT? that still needs a mouse/screen/keyboard to boot up.
Someone were to ask me, i'd call this a "rear-garde action". Microsoft is on the retreat!
Muahahahahahha!
There will also be a Windows 2000 64 bit
version.
Apple DOS 3.3 was from Apple.
MS DOS 3.3 was from Microsoft.
Two completely different products.
If that doesn't help Linux at Windows' expense, I don't know what does!
It's not about best of breed technology - it's about ongoing R&D, branding, marketing and commitment to customer satisfaction. I own a Sony TV and video and they're pretty average technically (and they were pricier than most other brands). But who's the biggest consumer electronics company in the world ? Ummmmm..... Sony. Microsoft are the powerhouse of consumer software and maybe the same rules apply to them. While all the Linux bigots here at /. were berating MS during the trial they (MS) were scoring all-time high scores from Joe Public as "most admired company" etc. OK so 100,000 Linux zealots won't buy an MS product - big deal, the other 4 billion consumers will.
As for your comment about the digital Walkman...... if Sony brought out a device tomorrow that used the most primitive encoding and technology in the business the Rio player would be dead in two days instead of one. Rio is an unknown outside of techie circles - Sony is a giant. You are looking at this from a "it's technically awesome so it must succeed" perspective - unfortunately the real world doesn't work that way. Sorry.
Business types are total morons, and are the only reason why windoze still lives ( Linux??? can that run word with all its way neato macro viruses? ), and small business types are too stupid to even make their business halfway decent in size. They use whatever comes with the box, and take courses in how to use excel. The only word they will notice in the entire post above is "cheaper".
not Microsoft
Sony WONT be releasing an MP3 WalkMan... it will be releasing a digital WalkMan using a proprietary encoding format, for which NO music will be availble for free on the net. Sorry, but this doesn't sound like a Rio-killer to me.
For a Whistle-type device, the internal operating system is totally transparent to the customer, why should they care what OS it's using? Linux has a big price-performance advantage in this market because 1) It doesn't matter how difficult Linux is to install -- it will come pre-installed on the thin server hardware. 2) No per-unit M$ tax. 3) No exhorbitant source licensing fees required before OS can be ported to non-standard (i.e. cost-reduced hardware). 4) Easier to port to new hardware. 5) Most thin server applications already available for free! 6) Greater reliability. 7) Embedable Linux is available NOW, not after M$ finally ships Win2000, another Win98 update, and a new WinCE and finally frees up some engineers to work on it. 8) How many products has M$ promised but never delivered over the years? Would you bet the success of your company on M$ vaporware (like Cool, for instance)?
Seems to me like it would be suicidal for M$ to go after this market -- especially if they come up with something as poorly implemented as their thin-client technology. I predict that M$ will get their collective asses kicked by Linux in the thin server market, just like they did by the Palm Pilot in the PDA market.
Linux is not an appropriate environment for an embedded system.
Excuse me? What kind of crack are you smoking? The WatchGuard firewall and several other embedded products are doing quite well running Linux. You don't have to be a strict "real time" OS just to service requests from the internet! Any latencies Linux introduces in servicing requests from the network are trivial compared to the other delays in the network! Let me ask you this: after my packet has gone through 3 routers to get to you, does it really matter if you service the interrupt from the nework card within a few milliseconds?
Sure, VxWorks is real time. But it ain't cheap. And yes, I develop for NT and VxWorks as part of my job -- and most of the networking code we're running under VxWorks has been ported from Linux!
How come nobody ever includes a backup device
in their "Super Hot Boxes"?
I wouldn't install a server without one.
I use PS 3.0/Win3.1. It was bundled with the traded-in scanner the local computer monger was about to throw out, but gave to (inflicted on?) me instead. PS is cool.
But what's even cooler is that I dual boot Linux on that box. You may need a Xeon to make NT run fast, but Linux is a happy camper on a 166 Pentium. I just don't need to buy another Intel product until this one wears out.
Linux IS a threat to Intel profitability; it means Intel has to compete on price/performance just like every other CPU manufacturer, rather than charging a premimum for it CPU because everybody is locked into the Wintel architecture. Intel's profit margins have already gone down (due more to Cyrix and AMD than Linux) and will continue to shrink in the coming years. This is a GOOD thing for PC buyers, bad thing for Intel shareholders.
Intel's existence is in no way threatened by Linux, just it's heretofore obscene profits.
"Resistence is futile. You will be assimilated" -- the big Borg from the Free Software Foundation (RMS).
To all those out there whining "but VxWorks would do this better than Linux" I must ask: Has Samba been ported to VxWorks? Remember, their going after file and printer sharing, i.e. Microsoft Network (SMB) server applications...
And how many copies of Windows have been sold? Let's be optimistic and say 5 million - hell, lets say 50 million. Linux will sell that number of operating systems in the first week. You seem to have confused "first to market" with "commercially successful".
Your pro-M$ FUD is easily dismissed -- ever hear the phrase "don't count you chickens before they're hatched?" You probably predicted M$ would sell millions of copies of MS-BOB too! Next time, consider waiting until you've got real numbers to back up your outlanding claims before you start slamming other people and gloating about Microsoft's success...
And NT, most certainly, is not.
I have a whole bunch of real world numbers to back up my "claims". Can you please tell me how many Cobalts have been sold ? Can you say "seven" ?
As for Bob - it sucked. Do your hear MS talking about it anymore ? No - you could be excused for thinking that it never existed. You are all bitter and twisted about something that happened 21 net years ago. Get real....
" Better parallel would be Chevy a threat to Shell Oil. If Chevrolet (Linux) comes out with a car that uses 1/10 as much gas, and has higher performance, you better belive that the oil company (Intel) is going to be right nervous. "
An even better analogy...
Chevy comes out with a car that runs on something other than gas, and it runs just as well.
you sound silly. maybe he doesnt run a server that thousands of people access a day, maybe he doesnt even run a server?
stop spreading FUD, please.
Intel feels no threat from Linux.
Intel manufactures Alpha chips.
Intel owns part of VA Research.
Intel owns part of Redhat.
Intel is funding the Linux port to Merced.
These 'thin servers' will probably have no problem running Linux as well as Embedded (ahem) NT.
Real men, of course, roll their own OS. Ask LT.
But what about pSOS, they used to be #1 in embedded Operating Systems, where are they now?
playing 'both sides' is like... good
because
1. theres no way to predict which way the
market will head
2. if you are in multiple areas, you can
switch over faster
of course this has the risk of defocusing
the company. but big companies seem to make money
doing this sort of double life thing.
Excuse me, but can you explain to me how you can use a $200 video card and a $100 OS, and still meet a $200 price point for a settop box?
What if Microsoft merely ports over some of its server apps (exchange, IIS, etc) to Linux? With some mildly clever programming, they could make the GUI look like NT. Especially if the thing's going to be web administered anyways.
The thin server has the benefits of a stable OS with the server apps that businesses are used to using.
AFAIK there is no Open Source Exchange Server project. Therefore, no threat.
It's free as in "free source". Let's face, Linux gets ported to a lot more hardware a lot more quickly than Windows, because there are far fewer impediments to porting it -- the code is better organized, better documented, and most importantly, available freely and immediately to anybody with an inclination to experiment with it. In terms of hardware platforms supported, Linux has already kicked Micro$oft's ass!
I'm not sure that extra RAM would cure the swapping. I'm sitting here at work with an NT machine (*help me*) that has 132M of RAM. As a little test of the system I decided to open a 69M text file (lets just say it's needed). During the attempt the machine was constantly thrashing and eventually complained about lack of virtual memory. In fact there is a problem with virtual memory that was posted to bugtraq a while back. Namely if you set the virtual memory to something trivial like 3k/5k (min/max) the system upon rebooting (for settings to take effect) becomes unusable and complains about lack of virtual memory to open even explorer. Even if the machine has a healthy amount of RAM like say 132M NT still won't work properly with minimal virtual memory settings.
The implication is that one must (not may) set aside an amount of disk space never to be exceeded whereas Linux can run quite happily without swap files or swap partitions if there is plenty of free RAM.
" First of all, Linux places NO threat to Intel. Does Intel REALLY care what kind of OS runs on their chips? "
Think please.
If Linux gains ground, and people find that they can run the same software on an alpha, strongarm, whatever CPU for cheaper and faster than on an Intel box, then Intel will lose market share.
Of course Intel doesn't care what OSs run on it's CPUs -- just as long as they don't run on anyone elses.
Oh yeah, I can see how being totally GUI-based is a big advantage in a thin server... NOT! A thin server shouldn't even need a monitor, keyboard, or mouse... it can be remotely administered from any other machine on the net using HTTP or SNMP! Worst case, you write a windows Ap to run on another machine to administer it.
Seriously, we're talking about single-purpose systems shipped preconfigured on cost-reduced hardware... seems like Linux has a BIG price/performance advantage in this market... why would anyone preinstall Windows on their firewall or router? What's the advantage that justifies the additional cost of the microsoft tax? Doesn't the same analysis apply to other thin server applications? What's easier to customize for cost reduced hardware -- Linux, with freely available source, or Windows, where you pay at least $100,000 for a source license? Seems to me, this initiative is going to take off like a lead balloon...
" Example: Netscape releases the source code for Navigator. Let's say you are a developer for Navigator. Suddenly you just lost your job because of this free software thing. It's ok though. There is a big ISP near by that uses netscape as it's default browser. Now that it has the source code for netscape, it hires you to hack
on it and customize it for that ISP. So, you are still hacking on navigator like you were before it went free, still getting the same pay maybe even more, but instead of working at Netscape Inc. you are working for the ISP instead. You don't lose any jobs by going open, they just transfer from software vendors to companies that use the vendor's software. "
Except that oops, most people want more standardization and not custom built software. Further, your idea is an incredibly worse value proposition in the long term than the existing model.
Explain who would write the first version of netscape? Who would foot the bill for this? Netscape/mozilla is not a small project, and we're seeing now the difficulties in it's development, even with commercial backing from other proprietary software. Where would this funding come from without the proprietary stuff?
I had a chance to play with one of their Instant Internet Boxes. It's just a rack mounted box that acts as firewall, web server, CSU/DSU, mail server and a few other things all wrapped up in one. Does DHCP too.
I wasn't on the job long enough to find out what the thing was running on, but for a small cash strapped company with minimal technical expertise on hand it's ideal.
Of course once you get the technical expertise in place Linux is just as good or better. The box was really aimed at the market of 100 user or less networks.
One more time, kids: A thin server DOES NOT need a GUI Want a GUI interface? Connect to it with a web browser, or a proprietary Windows App running on another machine. By definition, a thin server is connected to a network! The only advantage M$ has (a GUI look-and-feel that almost everybody is already used to) absolutely does not apply to the thin server market, which is why the Wintel solution is going to lose miserably.
And from what I've heard, Exchange makes NT look reliably and easy to administer... I sincerely hope they DON'T port Exchange to Linux!
I think that's just the point. Obviously MS couldn't target a zero cost market, but they could target a low-cost market based on turning old machines into thin servers. In order to do this, however, they would have to market a low-cost scaleable server OS that can run on old PCs, not an embedded OS designed to run on special purpose hardware. (Of course Intel wouldn't like this approach...).
The Register also has an article about Citrix possibly creating a Winframe client for PlayStation II.
http://www.theregister.co.uk/990407-000019.html
I've used Winframe clients with Windows 95, Windows NT, and Macintosh, and I've very been pleased with the results. Really makes administration at lot easier, especially with database apps.
Agreed. But I don't think that tax professionals represent a majority of Linux users today.
My point is that many of the reasons why people have traditionally said they needed windows are starting to disappear.
Sure, it'll take a while before linux can be all things to all people.
I just wish I had a good powerpoint viewer for Linux...
you can do it through the web. Doh!
http://www.turbotax.com/webturbotax/
as more apps are web-based and users are comfortable with the security, it just won't matter, and then windows will be immaterial. Microsoft already knows this and the must see the writing on the wall.
They're betting the farm on Rupert...
And how many Cobalt boxes have been sold ? Let's be optimistic and say 5000 - hell, lets say 50,000. Microsoft will sell that number of thin servers in the first week. You seem to have confused "first to market" with "commercially successful".
Let me know if I'm typing too quickly.
How can you except _any_ company under _any_ business model to compete with the Linus Torvalds of the world who give away their products and don't have to pay their "employees?"
The notion that software should be free because it costs next to nothing to distribute it is as plausible as the notion of paying on artist for the cost of his canvas and a can of point! Of course people should except to get paid for software they write. Not everyone is an artist capable of turning a bucket of paint and a blank canvas into a sight to behold!
A simple analogy:
the Rio player has been a moderately successful device in the consumer marketplace. Cobalt's success has been so-so (to put it kindly).
When Sony launch an MP3 WalkMan it'll be a HUGE hit. When Microsoft release their thin server it'll be a HUGE hit (when compared to Cobalt and all those other variants).
First isn't necessarily best or most successful.
Another attempt to make a new Windows CE Thin Client:
ATI once again redefines the future of
set-top graphics and video with new
Set-top-Wonder(TM) II
http://www.palmsizepc.com/apr99-8-1.html
It was both, so we have,
DDE
OLE
OLE2
COM
DCOM
DAO (access/jet)
ADO (the 'everythings a db' tool)
DirectX
coming soon Bastardized Open GL
hehehe BOGL
jmr
VxWorks excells in this environment, and the price is much more reasonable.
Uh, more reasonable than free? Your argument is valid against NT, but not against Linux. Linux's zero cost, easy development, and portability make it much quicker, easier, and cheaper to come up with a product based on Linux than on VxWorks -- and I know, I've done both. These advantages offset the costs of using a general-purpose OS. Oh, the other advantage of using Linux is that practically every application you would ever want to run on a thin server is already available -- for free. Sure, porting these to VxWorks with it's Unix-like structure is possible, but non-trivial.
Has Samba been ported to VxWorks yet?
Agreed, thank you for clarifying. I just couldn't figure out what you were trying to say.
=]
Hah! I've got a larger network in my dorm room. /."
Besides, what the hell are you saying? That you don't have the problems slashdot posters say, or you do? --
"(we) don't incur the lack of NT stability issues many allude to on
Tanks! (spelled wrong purpose on, you for.)
calx
in circles...
I ate my tag line.
-=Ellis (D)25=-
Linux = Software
Intel = Hardware
So I do not think so. I perfer AMD over intel, higher performance, lower price.
I ate my tag line.
-=Ellis (D)25=-
The free aspect primarily just allows it to survive in the marketplace while it matures into something that a particular user might think is appropriate. Conventional platforms cannot do this because that requires money. Atari's and Commodores and even Apple's need to remain profitable enough to survive.
Free Software takes that need to survive as a commercial entity out of the equation and at last puts an OS on equal footing with WinDOS in that respect.
In that respect, Free Software is just a means to an end: healthy competition in Operating Systems.
This is the key... today we have commodity Intel boxes for the simple reason that the vast majority of users run windows and need to be able to run windows programs that are generally *only* available on Intel. (Yeah, I've got alpha, and PPC NT boxes, but they hardly count for anything in the consumer PC market)
So... suddenly the Linux community has hardware mfgrs thinking that they can innovate again instead of just taking whatever motherboard Intel hands them and whatever crud MS hands them and slapping them together into an bland, undifferentiated white box.
Suddenly hardware is fun again. If you don't need backwards compatibility with DOS, windows, or X86, you can do a lot of innovative and interesting things.
I don't play games on my PC anymore, and the only time I use windows at all is to view/edit powerpoint or excel spreadsheets. So, what do I care whether my box's cpu is an Intel , Alpha, SPARC, ARM, or ??
I build all of my stuff from source anyway, except some security patches to the OS.
I think we're going to see a growing number of users who don't have any vendor loyalty for any parts in their computer. They just want the fastest stuff they can afford, and they want it to be standards-based and compatible -- or they'll just want a really cheap *disposable* PC. (If your TV, VCR, or microwave breaks would you fix it or buy a new one?)
Personally, I think that Linux on WinCE notebooks could be a killer product. WinCE seems sort of crippled compared to 9X, NT, and the machines seem underpowered... BUT THEY'VE GOT THE RIGHT PRICE POINT... Put linux on them, and sell them for $800 or less and see how many/few people complain that it's not running an intel CPU.
So, if we could pull the blinders off of some of these CE laptop mfgrs and get them to at least consider that their success isn't directly tied to WinCE or Microsoft, we might see some interesting stuff.
Also, look at the embedded market -- many embedded systems are still DOS based, but with Linux, you can choose the most cost-effective architecture, recompile, and run with it.
I don't think Intel is going away, but should they be scared of how well Linux runs on cheap Compaq Alphas? You Betcha!
I thought this stuff had already been done before... Web based management, simply plug in and turn on to set up, easy to use, cheap, embedded OS...
This sounds like a pretty cheap ripoff of the Cobalt Qube/Raq to me...
And embedded NT?? That's hilarious! Aren't embedded systems supposed to be cheap, small, and most importantly, stable?
-- Dave
Doesn't that sum up the whole Windows operating system anyways?
"Thin appliance type servers"? If they're a small office, they can just use a spare workstation as a server by installing FreeBSD or Linux on it. No need to buy some special device suited for a special purpose.
Not that I think Windows 2000 will pose any threat in this area, mind you..
So this is what, the third time Microsoft has proclaimed the coming of Windows "thin clients"? Uh huh. Wake me up if they actually get anywhere with them this time.
-- brandon s. allbery, sysadmin @ cmu electrical & computer engineering "Think, youth, THINK!"
The Cobalt Qube is an all-in-one, easy to set up low-end server solution, runs Linux, and costs $899. Why the hell would anyone want to pay $1000 - $2000 for something that runs on pricy Intel hardware and unstable MS software?
That's out there too. Gads. Talk about putting a bandaid on a ruptured spleen...
...phil
...phil
"For a list of the ways which technology has failed to improve our quality of life, press 3."
The sad part is that it'll probably work. After all, it worked back in the days when IBM did it all the time against rivals Amdahl and Fujitsu, and I don't see that the buying public has gotten any smarter in the meantime.
-- Eric
Send mail here if you want to reach me.
Posted by The Mongolian Barbecue:
I like how these are being called "thin" at $1500. I built a linux based server for under $500, which will handle email, web services, etc. What a joke.
What NT does have going for it is that if you want to use its nonstandard standards like DCOM, this works best in Windows. But that's circular reasoning, saying that Windows is a better development environment because you insist on using tools that only work well in Windows.
Don't label something "offtopic" unless you know the topic well enough to tell what's on topic.
So assuming that they actually follow through with this, it looks like they'll actually be trying to compete against Linux by offering something better suited for their target market than what they're currently offering.
It sounds like fairly steep money for what's basically an embedded server, and there should be plenty of room for Cobalt et al to knock down the price point. NT's still a heavyweight operating system, and it will be interesting to see how much they can actually excise from it while still maintaining the necessary functionality. My money has it that it's easy to build a much lighter weight server based on Linux than around NT. And while disks and RAM are cheap, the low end's a savage place where $10 savings on disk and $20 on RAM and $50 or whatever Microsoft's going to charge for the OS can add up fast.
The big issue for any vendor of such a system is competing with the big guys on volume, methinks.
Agreed! Besides, as bad as NT is normally, there's something VERY frightening about NT in a sealed box where I can do exactly NOTHING if it crashes in some way that the mini web server won't come back up.
The Cube in contrast appears to be quite ready for emergency procedures even though it's a lot less likely to have an emergency. Personally, I'm fond of the simple LCD interface for initial setup. Get an IP assigned by customer's admin, pre-configure, drop it in, and look like a genius!
My web server was a 486. Now its a p200 running IIS until I can figure out this httpd which will run again on the 486. Actually, it'll host 3 low intensity sites. My Internet connetion is a 384K pipe which allows me to ftp at 25K/s and host the few hits I get. We're a $40M trucking company, the site is only for phone/fax numbers and an intro.
I suppose it is because Linux itself is processor agnostic and large amounts of the more valuable software that runs under Linux is available as source code and not just Intel binaries.
Phil Fraering "Humans. Go Fig." - Rita
(currently testing something about signatures here)
Whatever kind of "thin" hardware they develop for said architechture, Linux will certainly run on it. It will turn out to be another scenario where Linux could demonstrate that it is a better server OS than NT. So, they might actually be helping us a bit in the end.
Go figure...
^D
In Soviet Russia, Jesus asks: "What Would You Do?"
It gives me a chuckle every time I hear that phrase...
^D
In Soviet Russia, Jesus asks: "What Would You Do?"
Is it just me, or are big companies (IBM, M$ for example) having a hard time not competing with themselves.
NT Embedded v.s. CE, NT WS, NT Server
NT WS v.s. Windows 9x.
IBM is backing competing *NIX variants for IA-64.
This whole "Thin Server Appliance" is not a "Thin Server" issue is crazy, as well. What is the difference? And why would anyone want to buy a stripped down version of NT? What would it come with? An empty hard disk.
WinNT can't compete with Linux as is, so what makes the think that a stripped down version can? In an enterprise (I hate using busswords, forgive me) situation, support costs will be comperable. Actual software/licence costs are dramitically different. Linux is much more scalable than NT. You can throw in, or cut out just about anything you want. Just try to get rid of the NT GUI??
--
(I count systems with the same essential underlying tech with a few slight differences as "flavors".)
And you know what? Everyone's worried about the Linux market fragmenting!!! What a laugh.
So, what's the difference between charging for the art or the entire package? Allocate your money however you want, you are still "selling" it. ("Call now and will give you this advanced digital watch FREE! ... with only $45 shipping and handling") :-)
Several differences:
Cobalt Networks has exactly what they are talking about. Easy to use, setup, configure, file server, print servers, with a web interface. I evaluated a cobalt raq that retailed for $1500 and it took me literally 10 minutes to get it up on the network using DHCP and half of that was me pulling out of all the packing material. The webserver took another 2 minutes to configure then add content. Comes configured with samba and printers are a breeze.
Go figure Windows and Intel always following in some elses tracks.
Stick
"Nuke it from space it's the only way we can be sure"
What do you despise? By this are you truly known.
That's one of the things I learned from working at the Iowa State Daily. (Apart from the fact that I didn't want to be a journalist.)
Not that it's necessarily bad for an article to have an opinion. (It's only good if you expect to sell your news to everybody.)
-----
The only problem with your scheme is that a company of 10 people probably wouldn't have the in-house skills needed to setup/configure/maintain a unix machine. And they probably wouldn't have to money to hire a competent one either. Enter the device any idiot can use (course we could just make web admin stuff for linux... oh, didn't they already do that with the cobalt machines?).
i mean, if they can't set up an NT box (even poorly) what chance do they have of setting up a linux box? very little, imho.
What is up with C|Net and CNN that give you these damn 2 paragraph stories? Why do they even bother? Why not just give us a head line, with no text underneath it?
As for the story itself, this is a doomed concept.
Let's say I own a small business and I have these choices:
A) A Wintel machine that the saleman says will be easy to install, but will perform only very limited function and come with no support except the phrase "Buy a bigger server"
B) A friend of the company who says,"I'll just set you up with Linux." Small Business Owner: "With what?" Friend: "It's cheaper, it works, I'll be done quickly. Oh and it's fully functional, so don't worry about expansion."
Only a PHB buys Microsoft because of the name, but most PHBs don't start small businesses. Thus, no market for for thin servers, IMHO.
Drake42
First of all, Linux places NO threat to Intel. Does Intel REALLY care what kind of OS runs on their chips? No. This is just another "initiative" announcement that means nothing. We will never see a "thin" server from this pair. Besides, the cost of ownership doesn't justify it. Maybe if it was about $500, then..... But that'll never happen either.
OLE2
ActiveX
ADO (or is it DAO?)
Yup, nothing like well thought out technology to make your job easier...
--
--
This signature left intentionally blank.
Linux is fine for thin servers, just ask the folks at Cobalt. Regarding embedded systems, the requirements of the system determine which OSes are appropriate. For instance, Linux would be a reasonable choice for non-realtime applications that boot from a flash device.
--
--
This signature left intentionally blank.
Never spent an hour hunting through every Pep Boys/AutoZone/Wal-Mart in town looking for a set of wipers, eh? Auto manufacturers are just about the worst example of "standards" you could come up with. Aside from the fluids you put in them, there's about 0% "standard parts" in the average automobile... ;-)
Not that that is so much better than CE, but the article indicated embedded NT 4.0
Seán C. McCord
Linux is not an appropriate environment for an embedded system. VxWorks is, QNX is, Linux is not.
That Intel doesn't really care for things to be portable. I mean, sure they're in favor of anything that'll run on their hardware, but they're less interested in seeing that same stuff run on other hardware. You can't really blame them, I guess; but still, it could *potentially* be harmful to the portability of open source stuff. I guess it just depends on how "nice" they decide to be...and how worried they are. If I were them, I wholdn't be *too* worried about it. I mean, most people are going to buy Intel, and it will fit their needs. They're a lot less "guilty" of ripping off their customers than MS is. MS *knows* their stuff doesn't meet anybody's needs!
You gave me something to think about. I thought maybe they were going to try to actually come up with a decent product to compete instead of FUD tactics, but perhaps this *is* just another tactic!
The other thing that is an issue is having an easy set up appliace that will basically do all your networking tasks. Yes linux can do this a lot better I think but can it do it with no or little configuration? The target audience for this equipment is for buisnesses that need networking between resources. One would have to make sure that any one can run this equipment when it is done. I haven't seen anyone ask this question yet (I didn't read all the posts I have to admit) but no vendor will put in linux unless they know that the customers will be able to use it and like it. It shouldn't take that long to make some scripts to make the system easy to use but does the vendor know this?
P.S. I can't spel
A product with this name actually came out a few years back. No idea (or interest) how good it actually was.
All opinions expressed herein are my own, and not those of my employers, who are appalled.
We keep flipping back and forth between single-user machines and timesharing machines - batch, timesharing, apollo/perq/pc's, 68k/x86 UNIX, client/server, applets, applicances, etc. It's all single-user or timesharing, and geniuses in suits keep changing their mind about which they think is the good one. (Clue: your OS should be able to solve both of these problems at once.)
I certainly won't trust microsoft or intel to come up with a new OS/systems design until they can show me the first shred of evidence that they have ever designed a credible one before.
So far, these products haven't really taken off (at least in what i have seen). However, the concept of a "storage area network" or a "zero maintenance" system has an appeal for small offices, home offices, and people that just want turnkey solutions.
Right now, NT sucks for turnkey work: stability, remote configurability, price, resource requirements, etc. all fail the model.
Linux, on the other hand, is dominating these products. I haven't seen one that is Great yet, but things like the Whistle Interjet (?) are a good concept.
MS knows where they need to expand the market. Linux hasn't really gone too far with the handheld market (yet), or with the "enterprise" market (yet).
I am just curious how they can make money in it; the business model seems off. The MS name doesn't add any value to the product.
So, has anyone heard anything about the NetWinder lately? It really seems (from the marketing anyway) that this is the dream product. Good size, good options, small processor, low power consumption... And, in quantities, it might even make the $500 price point...
It's not; that's why Intel has invested in Red Hat.
Intel is pursuing a sensible business strategy by supporting any and all efforts by OS vendors to use their chips instead of rivals AMD and Cyrix. If "Embedded NT" can sell a few thousand Celerons, so be it.
The Red Hat investment ensures that any customer who wants Linux will run reliably on PIII, PII, Merced, etc.
Hmmm. Runs Windows NT variant. Closed box so you had better be able to get timely support from the single vendor you are locked into. Used by small organizations so it better have every single application you might want built into it from the start.
I think I'm missing something.
For a small buisness, sounds like cheap hardware + linux would be the optimum solution in terms of price.
Why pay $1500-$2000 for that? I picked up a no-frills p2 400 256mb ram for about $800 (no monitor of course) and put linux on it at no cost Used a cheap monitor from an old 386 i had around.
Much better than a sealed NT Embedded box IMHO.
This sig is false.
Linux is indeed a Threat to MickeyShaft, but not in the way they seem to think. It is not a Threat because it is cheaper, or because it is open source, but the biggest reason Linux is a Threat to MS Dominion is that IT FREAKING WORKS !!!! People are -mostly- interested in Linux, IMO, because FIRST OF ALL it is more stable, is more scalable, and more powewrful. It does the job better. I think even if it weren't 'free', (distros can cost), and even if it weren't open source, that it would still be very attractive to most because of its PROVEN ability to do the job. The other aspects just make it even -more- attractive.
Unfortunately for M$, this is the only thing they can not face head on. Quality.
Since most companies have legacy harware in a stockroom somewhere, a server similar to what MS is proposing could be built for free.
1500-2000$? gimme a break.
M.
It seems to me there are a bunch of products that do this sort of thing. I think Cisco makes some boxes that are intended solely as web servers. And if you're looking for an easy to set up server to drop in your network, doesn't Cobalt make a line of thse? Of course, none of these use NT. Who the heck had the bright idea to make NT embedded? Aren't there a great many companies that make embedded OSes that have far more experience than Microsoft? I expect this initiative to fall flat on its face except for the inevitable "let's buy Microsoft because their stock just split again" factor.
I want one!
And BTW, does anyone know if or when Cygnus might release their simulator/development system for Playstation 2? I talked to someone from Cygnus yesterday who said you had to get it from Sony. She didn't really know but thought they might release it to anyone later this year.
"(the kernel and all the tools required to develop and run applications on top of it) is amazingly cross-platform--look at how little platform-specific kernel code there is."
I think I've read both from Linus and AC that the kernel is coded entirely in assembly, and if I'm not mistaken that is very platform specific... Think of how long it too to release the Alpha version of the kernel.
We're running a small network (15 machines) and we have a mail server, news server, internet gateway and 2 printers all hanging off of one machine, for less than Microsoft are charging for their thin server.
The machine is, admittedly, a Windows 98 box, which is regularly taken down (once a week) and rebooted. But it works, it's fast and it didn't cost $1500 (it's a pentium 90 with 64MB of RAM).
I can't see where Microsoft's Market for their thin server is.
My Journal
I am afraid of adding "alien" software to our NT box. For good reason, too... it's an NT box.
Hmmm. That thought is just disgusting enough to actually be worrysome. Judging from past experience, they are VERY likely to do something like that. Funk dat. :(
--
Paranoid
Paranoid
Bwaahahahahaa.
(drop one if this is repeated... wtf is 'Cat got your tongue'?)
Call me a weirdo, but I fail to see the advantage of doing this. Microsoft is attempting to combat Linux by cutting down hardware costs? Excuse me, but one of the major beauties of Linux is its ability to make supported hardware (e.g. less expensive older hardware) run at its full potential. Stripping off a little fat and making it run (stride? mope?) on a smaller processor isn't going to help. Its going to create another area in which Linux shines. If I'm not mistaken, this box is basically a PC minus video and keyboard.
Hmmm.
/me looks at the Linux 386 mail/dns-cache/IRC server sitting over in his closet with *AHEM* no video or keyboard *AHEM*...
To me it seems rediculously easy to install linux and apache on one of these suckers... Even with no video and keyboard, its gotta be possible. And it could be a major selling point of Linux... I mean, they're going to put gobs and gobs of RAM in the thing to ensure that Microsoft doesn't constantly swap. The box will have pretty much all the hardware a small-to-medium-sized webserver needs.
So, is it just me or will this completely and utterly fail to yield an advantage to Microsoft? And, btw, what does Intel care? The majority (afaik) of linux users use their architecture anyway...
--
Paranoid
Paranoid
Bwaahahahahaa.
In addition, NT doesn't have a tradition of remote administration or remote access, while Linux/UNIX has been used that way for a long time. Furthermore, a lot of server software, even for Windows, is written to POSIX APIs, and Microsoft doesn't enjoy the API advantage that they enjoy on the client. And the "high-end features" of NT that give it appeal in the business marketplace (Windows GUI based admin tools, fancy file system, etc.) matter much less on a thin server.
One thing in Microsoft's favor is that they do have all the software necessary to make this work in-house (the OS, SQL Server, IIS), and the incremental cost to them of putting that onto a thin server is small.
If they come out with a high performance, interoperable, thin database server at a good price, I'd actually be interested. I suspect, though, that that will be hard for them; with Linux, Oracle, and others, they are up against some good and established competition. But one can't fault them for trying.
From The Register full article:
Posted 08/04/99 9:07am by John Lettice
MS, Intel demo mutant thin server appliance
Microsoft's first showing of NT Embedded yesterday took the form of the first
demonstration of an alleged 'thin server appliance' co-developed with Intel.
But the implementation seems strangely changed from the thin server appliances Intel
has been bashing on about since last summer, and it seems inevitable that the device's
appearance does not signal a renewal of the formerly close relationship between the
gruesome twosome.
Intel's thin server concept is for a cheap, closed down box that's easy to install, and
performs simple, specific tasks on the network. (Intel network scheme means war with
MS) Something you plug into a small business network and then magically find your
print, email and so forth problems are sorted fits the bill perfectly.
Intel also doesn't want any nonsense about huge multi-purpose operating systems, it
wants single or limited task ones. So traditional embedded operating systems fit the
bill here, rather than bigger, multi-purpose ones like NT Embedded. Intel also insists
that per user licensing is out of the window - if a thin server network is having to pay
a couple of hundred dollars to Microsoft for each person connected, then it's not low
cost at all, is it?
We remarked on how obviously Microsoft NT didn't fit the bill as the operating system
at the time of Intel's announcement, and we remain right.
Yesterday's demo did make a small breakthrough on licensing. It would appear that
Microsoft is willing to let the box host an unlimited number of users, but while this might
appear to be a massive breakthrough on licensing, check out the catches. The thin
server is intended to handle file and print sharing, not application hosting.
It will deal with Internet connectivity, but the no application hosting aspect means no
Web server hosting, so -- phew -- we're not going to have that problem of people
trying to host Web servers on NT Workstation rearing its ugly head again.
In its "thin server appliance" incarnation, NT Embedded is therefore crippled. How
badly crippled remains to be seen, but as you begin to compare features with
projected features for the next generation Windows 2000 NT variant, you'll no doubt
note it is very badly crippled indeed.
Microsoft is obviously trying to preserve revenue streams. It doesn't want to lose high
margin business on NT server sales, and it doesn't want to lose all those lovely client
licences. But it does have to do something about simplified, stripped-down boxes and
network operating systems. And here it's shooting itself in the foot with this particular
project.
It's quoting a price for an OEM-built thin server appliance of $1,000-$2,000, which is
of course basically a standard PC price. No surprises there, as there's going to be a
standard PC in there with the screen and keyboard chopped out. The price leaves
space for MS to charge something in the region of the usual amount for its software,
so again no nasty precedents created here.
But Intel's view of a thin server appliance lies more around the $399-$499 mark. You
can do file and print perfectly adequately at this level, so what is it about the MS
variant that's worth an extra $1,500? Users will vote with their wallets. ®
I'm afraid. I'm afraid, Dave. Dave, my mind is going. I can feel it. I can feel it. My mind is going.
We have a classroom at my college that is comprised of 45 thin clients and 2 prolient servers running a winNT server terminal edition. They are the biggest pain in the a** and NT crashes like crazy. Stick to *NIX for term servers.
Stupidity is its own punishment
No, they're going to be running "embedded NT", which needs "only" 12-16MB to run.
It's worth noting that Intel has its own thin server products, which are based on [34]86s and VxWorks, which sell for ~$400-600. Since this is non-MS, cheap, and no per-user licenses, MS hates it.
This new thing looks like a truce between Intel and MS, but it's unlikely to work well. Embedded NT will be too expensive, and $1000-2000 for a box is just a PC without a monitor and keyboard - not really in the running compared to better tuned custom embedded solutions running a more appropriate OS - Linux for example.
Several things jump out; anyone know anything more?
/. and knows people who can, in less than a day or two.
What's going to be inside one of these things for $1,500-$2,000?
I'd expect a current $70 Celeron300A without floppy, CD-ROM, Linux installed, $40 SVGA video card, $90 motherboard with onboard sound, $40 NIC, $900 for a 18GB SCSI HD(for kicks), $200 for a SCSI card, and $500 for 256mb of SDRAM... $1800 machine, with totally awesome HD and more than adequate CPU for a small business, no? With only 30-100 people, right? Scrape the memory and HD down to say, 9GB SCSI and 64mb memory, and the price goes down to $1000 dollars or something =)
If you cared about reliability(for a small office of only 30-100?) you could set up 2 of these machines mirroring each other, for $2000 dollars, right?
I don't know that one could argue ease of use either, because in the Intel/M$ case, because it's sealed, if something goes wrong you aren't give the option of fixing it. In a Linux box, worse case, no one in the office knows how to fix it. Best case, one of the office workers originally set it up and can probably get it up to speed, or hangs out at
If this is such a hot market, why does Intel and M$ think they can beat out a $1000 Linux box?
Why do they think they will dominate that market?
Am I missing something?
AS
-AS
*Pikachu*
I must admit I am confused, I fail to see how a stripped down closed NT box can compete with Linux. Aside from the Cobalt Qube, ease of use and simplicity are not Linux strong points. So what sense would it make to make a 'competitor' with these features. Plus given M$ problems with shipping and maintaining their products, can they handle another one.
They fail to address Linux's strengths; stability, features, cost etc. So unless they are going after the Qube, I hear BOB.
Woe be on to them, all who rise against poor people, shall perish in a the end. Buju Banton
Intel is setting itself up to win from all ends. If Intel wasn't doing this with MS, MS'd do it with AMD.
Highly cool.
Mike
--
Mike
--
"Wi nøt trei a høliday in Sweden this yër?"
Seperating the engine from the storyline really makes sense for certain types of games. Games similar to the old style Sierra text/graphics adventures would really lend themselves to this kind of distribution. The game engine is developed and released for free. The artwork, music, and story are composed and then made available for purchase as a separate module.
Since the source for the game engine will be available, anyone could make a game. An easy to use game development kit might then be produced to facilitate this. The quality of the game that you make would determine its value in the marketplace. An weekend of hacking a game together might produce something good enough to download. A couple of months of careful design by skilled writers, artists, and musicians would be worthy of purchase. Realisitically, the game would have to be killer to be able to survive in this type of market. Yet isnt this what we want? Tying the game to something with a strong brand-image would also help make it commercially successful. Movie tie-ins, background music by famous artists, and recreations of great storylines could set a profitable game apart from a free download.
On the other side of the spectrum the game engine would be a great teaching tool. The intracacies of plot development could be studied by doing. Effects of visual images and music on a storyline would be very easy to play with. The creative process of game develoment is made available to everyone. Not only that, but lessons learned from the process could be applied to other more academic areas such as writing.
Now if MS had a clue, they could make a thin server for games of this type, to allow multiplayer functionality. Actually, that might not be a bad use for these lil boxen. Any moron could setup a game server for a LAN party, or online multiplayer game. Shoot, that would be the only thing I'd really trust them with, and it could take advantage of the large base of game apps. Hmmmm MS Game Server 2001, almost sounds plausible.
Here is the most succinctly-put set of reasons I've seen yet about why Windows-NT won't work, and is not likely to start working right in the near to mid term future.
/ncw-06-1998/ncw-06-lastten.html
An excellent analysis of the real costs NT will incur from IT in the coming months.
Also, a few really good insights into the problems Microsoft faces in retooling NT5 for network portable client transparency and user context passing and preservation.
The previous 3 articles in the series are worth reading, too.
http://www.ncworldmag.com
Indeed, I do not run a server, nor own a 486. It was just a freakin' example of what Linux is capable of.
Though if I had any decent web connection at home, I can see getting some crappy lower-end machine to take over the serving duties. Maybe my mom's old P-75. =)
--
Okay, I got Linux installed. So where's the free beer everyone keeps talking about??
"Oh! You don't need to buy cheap Linux servers! Soon good ol' Microsoft will give you all the cheap servers you need! Friendly! Easy! Microsoft! You can click and drool through all your intranet needs! No scary text from Microsoft!"
Six months later. . .
"Oh, we found the implementation of thin servers to be too limiting. Our small business customers demand a more complete, robust, crappy, expensive solution. Just get NT 4* and a few P-III 600s."
*Y...er, W2K will have been pushed back to a March '00 release by then.
Okay, I got Linux installed. So where's the free beer everyone keeps talking about??
Because it runs so damn well on old hardware, I'd guess. Why upgrade now to Intel's latest and greatest overhyped, extra-cached money sponge when you can serve your needs (and your webpages!) with that old 486 you had in the closet?
Linux does NOT encourage the "upgrade cycle" that mainstream computing has fallen into. That could very well make Intel nervous.
--
Okay, I got Linux installed. So where's the free beer everyone keeps talking about??
Microsoft has to do this. Systems like Linux are very compelling on the server front, and this is about the only way that they can really compete with Linux. After all, they cannot give their bread and butter away for free now... can they?
Moreover, Intel probably is not terribly serious about this market, as they will sell the server running Linux or running Windows. So this is at best a distraction, and a low profit (e.g. low attention span) direction. In Microsoft's case it is important, as the economics of Linux as a small workgroup server are really hard to beat, especially in direct comparison to NT server.
I suspect that the company that is more concerned about this is Sun. Linux has already munched its low end business and this thin server idea basically nails the coffin on any realistic strategy that they might try as a comeback in this market. I have to admit that this is all getting rather amusing.
Actually this could prove extremely healthy for Linux. As far as a MS based developer is concerned (and there are loads and loads of them), the main problem with Linux is that we don't know how to develop on it. How, for instance, am I to make a DCOM server under Visual C if I'm using Linux (I know there is a unix dcom library, but it costs almost as much as NT server, and I still can't use ATL). Likewise an ASP page (without resorting to chillisoft's $900 odd ASP for Apache thing).
:)
NT has a completely different set of problems: Stability and Price (or to be more accurate, absurd licensing). Especially if you want to connect more than, say, five users. So the question is how to make the best of both worlds? How to make use of the inherent stability of Linux to run the database, net connection, etc. And host a middle tier on NT, where it's easy to develop?
It seems to me that Embedded NT could help out here. It should be possible to make a diskless board boot from a Linux box and just host the middle tier. Bingo, we have an easy to develop for application layer. The Win95 clients can merilly connect via DCOM, share their files (and be logged on via) Samba, and everyone's happy.
Have yourselves a URL.
http://www.microsoft.com/embedded/winnt.htm
So, exactly how far out am I?
Dave
I write a blog now, you should be afraid.
Intel needs to sell chips and motherboards - lots and lots of them. Especially the pricy, "high-powered" pairs. Now then, along comes Linux and it runs just fine on a 486. "486?!?" shouts Andy Grove. Those are dirt cheap! Windows 9x are pigs on anything smaller than a Pentium 90 and to really appreciate the OS bloat you need a PII.
So, Yes! Linux is a threat because it is more efficient on the same hardware. Does this logic boggle the mind or what?
They are creating another proprietary, non-scaleable solution. Unlike NT, You don't need a stripped-down thin version of Linux to create a thin server. With Linux, you can expand the system as your needs expand.
What they'll miss with their approach is the entire segment of the market who are taking their old hardware and giving it new life as thin servers using Linux or freeBSD for essentially zero cost. In our company, we already have 13 old pentium machines functioning in this capacity. This would have cost between $13,000 and $26,000 using these new machines.
How does Microsoft compete in the server market?
Take a look at Office 2000 and youll know the answer. They dont compete (as if they ever had).
A little example?
Until know the company I work for used NS products: Navigator, Enterprise Server, Proxy Server etc.
Now they are in Microsofts Office2000 Early-Adopter-Program (Im not sure of the name but you know what I mean). So what happens?
"Hmm, when we install Office2000 IE5 will be out there, so why not use it?"
"Well, there are a lot of great new functions in Office2000, but we need IIS for that, so lets install some"
"Oh, we have IIS running? What do we need NS Enterprise Server for?"
Its like theyre saying: "We know our server is crap but hey, we posses the desktop (which is NOT the REAL problem, that is they posses THE STANDARD OFFICE APP). So lets do what we do best, produce some REAL GREAT NEW FEATURES which only work with our servers."
And BANG, there goes competition.
Im not sure how to compete whith that.
We need something to compete with MS Office which is able to make the transition extremly simple.
People in companies are NOT using Windows, they are using Word, Excel or Outlook. And companies are NOT going to change that if they have to send 25.000 employees back to "Linux-Office Training-Camp".
But most important might be the "Openess of protocols" that is reading and writing of MS Office file formats. Give me an office application for Linux which reads and writes MS Office documents perfectly and I get rid of about 85% of my problems (not talking about Outlook, Exchange and that Mapi crap).
I'm sure that people will actually buy these things and Microsoft will probably make money, for a while. I think that what's beginning to sink in (for some, anyway), though, is that fighting free software is kind of like fighting the undead -- there's not much you can really do to stop it. I wonder if it's sunk in with Microsoft that with Linux, Gnu, etc. (and even the whole internet, in a way) they're not really up against a product per se, but a whole new paradigm -- where people are making money not from owning information but by actually performing services and doing work -- and we all know that you can't make billions that way.
These are certainly interesting times.
-- It only takes 20 minutes for a liberal to become a conservative thanks to our new outpatient surgical procedure!
Better parallel would be Chevy a threat to Shell Oil. If Chevrolet (Linux) comes out with a car that uses 1/10 as much gas, and has higher performance, you better belive that the oil company (Intel) is going to be right nervous.
Why yes, I AM a rocket scientist!
...becuase Microsoft server software never works reliably.
The reason Microsoft can't relase reliable server apps (IMHO) is that they hide them under layers of GUIs and abstraction and crud.
More complex = less relaible.
Adding a "web administration" layer on top of this can only make it worse.
If you want a cheap web server, try Linux running on an old Pentium box with Apache.
Do you even know anything about perl? -- AC Replying to Tom Christiansen post.
...who want to go back to the mainframe / host days of computing? What's next, Visual COBOL? I mean, at least with client / server I can do work if the network goes down.
DrLunch.com The site that tells you what's for lunch!
This is one of the original arguements that IT gurus used to open the door for NT.
Now Microsoft is fragmenting their operation system offerings all by themselves.
LOL!
Contrary to earlier MS press releases they are increasing the number of Windows variants not decreasing. Soon we'll see Win95, Win98, Win98 version 2, WinNT, WinCE, and Windows 2000.
I wouldn't bet on this thin server version of NT being out before 2010.
I'm not convinced what they're planning will be able to run
Linux. They've been talking about this "WinCPU" thing -
think WinModem and the trouble those are causing Linux at
the moment.
If M$ pay Intel a lot of money to develop a proprietary CPU that only
M$ have the API for, I'm sure Intel would happily take their money.
Then we end up with very cheap hardware that can only run Windoze...
That's exactly right. The only reason a proprietary software vendor would want to go open source is if it would boost the sale of a complimentary product. As an example, Apple makes their money on hardware sales. The OS is really a loss of revenue. Another example would be Netscape. They were already giving the browser away. They made their money on server sales. Doing work on the browser was a loss of revenue, but it did help support their server. They certainly didn't lose any money by giving away mozilla; In fact their was a big potential that giving away the source to the browser would increase their server sales. Unfortunately for Netscape, another open source project called Apache steam rolled right over all of the proprietary web servers.
The companies that are going open source aren't software vendors. That would be corporate suicide.
--- A Jesus Fish eating a Darwin Fish only proves Darwin's point.
How can you except _any_ company under _any_ business model to compete with the Linus Torvalds of the world who give away their products and don't have to pay their "employees?"
:)
Uhm, I don't. They can't compete. End of story
The notion that software should be free because it costs next to nothing to distribute it is as plausible as the notion of paying on artist for the cost of his canvas and a can of point! Of course people should except to get paid for software they write. Not everyone is an artist capable of turning a bucket of paint and a blank canvas into a sight to behold!
I'm not saying you don't pay the artist. His work is worth more than just the cost of the canvas and paint. But how much should prints of that painting cost?
Programmers should still get paid the same salary for the act of programming, but the net worth of the software should not be how high it is currently.
Freeing the source code just takes out the middlemen (software vendors) of the industry. They are the only ones that need to fear free software. Companies will always need, and pay, developers.
Example: Netscape releases the source code for Navigator. Let's say you are a developer for Navigator. Suddenly you just lost your job because of this free software thing. It's ok though. There is a big ISP near by that uses netscape as it's default browser. Now that it has the source code for netscape, it hires you to hack on it and customize it for that ISP. So, you are still hacking on navigator like you were before it went free, still getting the same pay maybe even more, but instead of working at Netscape Inc. you are working for the ISP instead. You don't lose any jobs by going open, they just transfer from software vendors to companies that use the vendor's software.
--- A Jesus Fish eating a Darwin Fish only proves Darwin's point.
Ummm, you expect Microsoft to target a market defined by the fact that it spends no money? Open a window, you need some air (no not that kind of window).
Isn't that what WinCE was for? Remember "Windows" everywhere?
And what about NT 5? The Big Operating System that couldn't. That was for servers, I thought.
Where's the logic?
Windows is smart, it can multiply very rapidly: Win3.1 Win95 Win98 WinNT4 Win2000 etc etc.
--
Win CE?? That's a recipe for disaster.
-Sapo
There are some interesting issues with going open source. For a product that has stiff compition, going open source would allow their competitors to see how they designed their product and copy ideas more easily. Sure, going open source would gain them a huge user base because it is free, but what makes more money, 4.6*10^19 users * $0, or 20,000 * $100? If there is very little need for services on the software (the software is easy to use and performs a specific task with no customization), there is no way to make money.
... with only $45 shipping and handling") :-)
Like I said, OSS works great for areas of software where there is huge amounts of customization (operating systems, custom software, etc), but for anything less, it most likely will fail miserably. I'm not putting down OSS (I like it), just pointing out the limitations.
Commercial propietary software is not all bad. A lot of it provides products to users that would never survive under OSS. 99% of computer users can't program, therefore it is impossible for them to help maintain their desired applications. Thats why most OSS projects today are tools made for coders by coders. Linux is a great example. Linus didn't want to pay for the commercial UNIX varients, so he set out to make his own. A tool for a coder, by a coder.
Thus you can give away your game, and sell the artwork, msuic, story, etc. like any other copyrighted material such as books, paintings, etc.
So, what's the difference between charging for the art or the entire package? Allocate your money however you want, you are still "selling" it. ("Call now and will give you this advanced digital watch FREE!
- Darrick
Programmers should still get paid the same salary for the act of programming, but the net worth of the software should not be how high it is currently.
So if you are not selling the product, where does the money to pay the salaries come from? Remember, it is the company trying to make a profit, not necessarily the coder. The developer on Netscape is a good example on how the *coder* makes money by going to the ISP to customize navigator, but where is the company that developed it in the first place? They don't exist anymore. This is why most companies are not going open source. Sure, the individual engineer could still make money, but not the company.
- Darrick
The economics of proprietary software are flawed. No other market sells a good that is as ridiculously over priced compared to it's creation costs as software
Wrong. Creating software is more than just printing manuals and burning CDs. Most software takes teams of coders and quality assurance engineers years to produce. During this time, they are all drawing salaries and benefits just like any other profession.
Lets say you have a team of 10 coders and 10 QA, each earning an average of $50,000/year. Assume the total cost to the company is 1.6 times their salary to account for benefits. If it takes one year write the software, it costs the company $1.6 million ($80,000*20 employess) before they can even sell it! And this is only counting the salaries of the coders and QA. This does not account for salaries of management and sales and marketing, or even paying the electricity bills. If you throw all that in you are probably getting close $3 million.
Now assume they can sell the application for $100 profit per copy. At a production cost of $3 million, they would have to sell 30,000 copies (which is a lot tougher than you think) just to break even, let alone make a profit.
Most software companies won't even try to make a piece of software unless they feel they can get five times the return on investment. This isn't because they're greedy, but rather a for saftey reasons. If they project the market wrong, or overestimate demand, they can lose their shirt really quickly.
Despite the recent hype of open source software, the OSS model does not work in all areas of software. OSS works well in the operating system and software standards arena, but fails in the niche markets and small applications and games.
Don't get mad at the "evil" software houses that charge for their software. Trust me, most of them are making about the same amount of profit as any other type of business. Its only the big ones with dominant market shares that rake it in because they sell millions of copies. And the only reason they sell millions of copies is because it is a standard necessity for people. And when something becomes a standard, this is when OSS takes over (Linux, GIMP, KOffice, etc).
- Darrick
Keep in mind that some of those legacy Intel PCs have various components that Linux might not be directly anticipating. Learning of indirect sources of drivers, etc. may take more time than cost of a Cobalt machine.
Colbalt was featured in major story in current Wired. That seems to be the ticket to me. FWIW, we run 3 NTS machines and 5 NTWS machines using a wide variety of apps and don't incur the lack of NT stability issues many allude to on /.
And I have business clients with smaller and larger networks.
what the hell are you saying? That you don't have the problems slashdot posters say, or you do?
We have experienced solid stability with Xenix (in the '80s), Novell (early '90s), and NT (present) in systems supported in small businesses. If NT were as unstable as I read about here we would have to abandon it --- we have to have high % of uptime. Granted we don't have 100s banging on NT, but those we do bang a lot and it is important that they have effective uptime.
When suitable app software is available for client(s) environment(s) we will consider installing Linux. Nothing elitist about that, just that systems need to serve people effectively. For some people, Linux is best option today. For others, MS cannot be excluded today.
Yep, and as that becomes more pervasive lots of folks will file using web-based tax software. In terms of total number of 1040s the potential is for a large majority to file that way. But tax professionals with clients with extensive complications (we spend 20-75 hours of competent professional time on some of them). For that kind of work, calc speed (frequent iterations) is king. So I'm skeptical that we'll have enough fulltime Internet bandwidth to compete with FastEthernet LAN and fast HD. I expect as things move to XML that many more apps will become web-based. I'm ready for the next best way. But I also have to deal with the present and short term future ... and in doing so, cannot abandon MS.
I agree that COM and DCOM are very attractive aspects of Windows for programmers. Since the COM standard is supposedly open, I've often considered starting a "Free COM" project.
I've looked over CORBA (perhaps with Java) as a possible alternative to COM on Linux, but I haven't done anything with it and can't really comment on it.
As far as re-using components on a local machine, COM becomes a little less necessary with open-source software. COM provides a way to reuse binary code, but open-source software lets you grab the original code and use it. This is NOT the only advantage of COM, of course, and copy/pasting code for reuse is a quick way to bloat executables (as well as cause problems for people developing non-open software).
DCOM is a little harder. One aspect helping Linux here is the fact that processes under Linux are less cumbersome to start than under Windows. Windows developers are used to calling up an in-process object everytime they need to access a piece of code quickly. Under Linux, starting a seperate process remotely and reading its output from stdout is much faster than doing the same thing in Windows. This, BTW, is why MS recommends using ISAPI and ASP instead of CGI on IIS -- starting a new process is just too darned slow.
I'm sort of taking both sides here -- a lot of things in Linux make COM and DCOM unnecessary, but it will still be REAL cool when I can remotely access and re-use a component on Linux.
Save the whales. Feed the hungry. Free the mallocs.
Isn't Corel already producing a series of Linux-based servers based on the StrongArm? I seem to recall you could get a rack-mounted web server from them for just over $1,000. I don't hear much about them these days, though.
Too bad. I once worked at an ISP that would have been better off with multiple overlapping cheap servers instead of a few monsters. It seems like a worthwhile idea.
Save the whales. Feed the hungry. Free the mallocs.
My college loves COBOL (don't ask me why) & they converted to running Visual COBOL about a year ago. It's ease of use is lower than the older DOS based ide they used.
hell, the GUI _is_ the OS. all of win32 is emulated on top of the NT executive.
If someone was indeed crazy enough to take on one of these "thin" machines, I feel it would be a case of caveat emptor - let the buyer beware!
You may just end up with a rebadged DOS 3.3 - Thats the only compact and stable codebase Microbloat ever had, and they have after all proven themselves to be an environmentally conscious company before - look how many times they've recycled the WinBloats95 engine!!! Woo-Hoo -Windows 2000 Personal Edition, or is it Win95 osr5? It's hard to tell from this distance!
I'll be happy with Linux, thanks all the same fella's, at the rate it's evolving M$ may just have to go into the (X)Window Manager business, or start pinching the code from Gnome and KDE, just to keep Bill in pies...
>:-)
Many find the software package they want to use to run thier business, then buy the hardware (plus OS) that runs it. They know that you can't run billing, point of sales, accounting or inventory tracking with generic OS-anything. But there are a lot of business that are running Xenix, SCO, etc. because UNIX is te preferred platform of the vendors that they like and trust.
Find the software vendors that make the stuff that the business world likes and turn them on the Linux. Then it won't matter what new toy MS makes.
-- I browse at +5 with stripped sigs
The $2000 price point is not intended to undercut the cost of linux+enough of a box to run it (say, a brand new emachines system? that's essentially cheaper and more reliable than anything they've got gathering dust in a closet...)
The $2000 price point is intended to undercut the cost of the linux dude the small biz will call to set up the linux+emachines system.
M$ is not trying to kill the virus (linux), its trying to choke off the vector (hobbyists and consultants).
- Why are you putting on tennis shoes? you don't think you can outrun the bear, do you?
- I don't have to outrun the bear, I just have to outrun you
With Intel's investment in Red Hat, it seems that they are effectively playing both sides.
I'm sure this "thin union" idea came from within M$, possibly just for PR value.
Hasn't the media heard about the Cobalt Cube? I guess Cobalt needs to advertise a little more. :)
If I had to suggest a drop-in server for a turnkey operation, a Cobalt Cube would be on my list of suggestions. Depends upon the customer's needs though.
Codifex Maximus ~ In search of... a shorter sig.
And I'll take you up on that bet. I have an Alpha sitting right here that I've never had a problem with; it runs all the software I've tried to compile. How many non-Intel machines do you own, and can you name an operating system component that isn't usable on it? I'm expecting a reply.
when they rever to "n alleged 'thin server appliance' "
."
It's obviously not going to be positive at that point . . .
I wish I could remember the old tale about the cub reporter getting the lecture on "alleged," "purportedly," etc. Before being sent to a society function.
"Mrs. Johsnon is allegedly the wife of Jack Johnson, and claims that Paul Smith is her father . .
It also contains the amusing quote:
We remarked on how obviously Microsoft NT didn't fit the bill as the operating system at the time of Intel's announcement, and we remain right.
Since when is Linux a threat to Intel?
I don't know if the public has gotten smarter - but I do think it's gotten more angry at malfunctioning computers.
...
At least when IBM ruled the roost, their systems stayed up
D
----
I think that what's beginning to sink in (for some, anyway), though, is that fighting free software is kind of like fighting the undead -- there's not much you can really do to stop it. Thank you for stating a thought that's been crawling through my subconscious for a long time! I guess we can now officially identify ourselves with the Zombie Hordes of the B-Class sci-fi horrors. You're absolutely right, of course. Open code is The Thing That Would Not Die. You can fork it, you can FUD it into submission, but you absolutely cannot kill it. Of course, you have now put a nasty vision in my head. I'll have to tell Illiad over at UF...
--The basis of all love is respect
MS just doesn't get it. They think Linux is being sucessful because it is stripped-down and featureless. They couldn't be any further from the truth. It is sucessful because it is exactly as featureful as you *want* it to be, and no more. Yes, it can be a stripped down file/print/network-bridge machine, but it can be easily expanded later when need be.
They just don't get it.
Don't label something "offtopic" unless you know the topic well enough to tell what's on topic.
Linux is hardly processor-agnostic. It strives toward that goal, and does fairly well, but just take a look at the verious programs out there. I'll bet you that for every ten you pick at random from the list (not counting the kernel and tools required for an absolute bare-bones distribution), all ten will compile out of the box on an Intel-based system, but at least one of those ten will compile out of the box only on an intel-based system, and it's more likely to be two or three.
Linux is not a cross-platform operating system. It is an operating system for Intel-based PCs, which happens to run on some other platforms as well. It strives to be cross-platform; someday it might even achieve it. But it's not there yet.
Never spent an hour hunting through every Pep Boys/AutoZone/Wal-Mart in town looking for a set of wipers, eh? Auto manufacturers are just about the worst example of "standards" you could come up with. Aside from the fluids you put in them, there's about 0% "standard parts" in the average automobile... ;-)
The wipers were an attempt at irony.. :-D
There may not be a "standard" wiper, but the general design of them is not patented by Intel, unlike say SLOT ONE (or Two...).
Socket7 was turning the CPU into a commoddity, so Intel's response was not to turn up the technology but to create a new socket and deny access to it, then make the techies take a back seat to Marketing. What a joke.
At least Intel pay well, so they can steal employees from Motorola. Motorola is too busy being anal to their golden employee, and allowing a gestapo-like IT department to "force standardize" on INTEL computers, at the expense of working Motorola-based Apple (or even Windows NT/PowerPC).
THAT must be demoralizing to employees who care about their company. They must have the same pointy-haired bosses middle-management "fat" found in places like IBM and Digital oops I mean Compaq.
It IS a goal for Linux to be platform independant, and it does a fine job at it. You can run Linux on LOTS of old computers so it's not necessary to buy a new computer as often. For example, all the Windows users will be buying new PC's within a year of Windows2000 for the same reasons people bought new machines when going from Win3.1 to Win95: your old software MAY work just fine but if it's not supported you WILL cough up money for a new system.
HOW MANY users of PhotoShop 3.0 still run on Windows 3.1? By contrast, how capable is the same machine running Linux. EVEN IF Linux were not faster than Windows, the very fact that you can OBTAIN a software upgrade for the computer may be enough to stick with it; some people are "cheap". How many users of 68040 MacOS 7.5 through 8.1 are still out there? LOTS. The systems are slow by today's standards, but you're not kicked down a flight of stairs like Microsoft users are.
All things being the same: price, performance and software availability, would you rule out non-Intel CPU's for a dedicated Linux-only box? You would be foolish to do so. Intel has every reason to fear Linux... after the Microsoft takedown THEY are next!
I'd *love* to see that Russian "Merced killer" become a reality, with a Linux port. CPU's are too expensive. The weakness of any product is allowing the customer to evaluate the competition; this is what software upgrades are all about. Intel's strategy is going to be to encourage BINARY file distribution, discourage commercial software from distributing the source (it's rare but it COULD catch on...), and attempt to introduce Intel-specific bits into libraries and kernels.
Intel is not a standard. REAL Commodities like SVGA monitors, the size of a soda can, analog clocks, and even winshield wipers ARE standards. CPU manufacturers sell magical sealed black boxes that can't be peered into and are just as capable of screwing you over as the Microsoft monopoly.
It's my understanding that Microsoft still sells a few copies of OS/2 1.x (at the original not-cheap price of $500) for embedded applications. Why not use IBM OS/2? I don't know.
NT-Embedded is designed to be a replacement for this product.
--
Business. Numbers. Money. People. Computer World.
I doubt it. Microsoft has been working on Embedded NT for some time now.
But it's good to be paranoid, because everyone's out to get you.
--
Business. Numbers. Money. People. Computer World.
Actually, you can go open source and charge for your products. Niche markets where you are going to have five clients are a bit tricky, you probably can't go free redistribution in those markets.
Games, on the other hand, are in a wonderful position to go open source. In all but a few cases, the code that runs the game isn't worth all that much, it's the story, artwork, and music that makes the game. Without the artwork, story, and music, the game is next to worthless. Thus you can give away your game, and sell the artwork, msuic, story, etc. like any other copyrighted material such as books, paintings, etc.
They laughed at Einstein. They laughed at the Wright Brothers. But they also laughed at Bozo the Clown. -- C. Sagan
The article states that Microsoft can't compete with linux in terms of price, but they believe that the ease in which you can port apps to these "thin servers" is what will give them an edge. I thought the main selling point of these servers is that they are thin and only serve one function. Why would you want to port an app to it?
It's always nice to see Microsoft making bad business decisions. We'll see a lot more of them as they try and combat Linux.
What can Microsoft do to stop the Free Software juggernaut? Their loss is inevitable; all they can do is slow it down. The economics of proprietary software are flawed. No other market sells a good that is as ridiculously over priced compared to it's creation costs as software. It's this flaw that made Gates the richest man in the history of the world, and it's the same flaw that will destroy almost all of proprietary software. If free software didn't come along to destroy the industry, piracy would have.
I think it's kind of funny watching the industry topple. It's not as apparent with Microsoft (because we don't have the end user apps to topple them yet), but just look at proprietary unix vendors. Most of them are dead and the rest will be shortly. The only ones that are really making money are the ones that run on the Big Iron that linux can't do, yet.
Anyway, to tie all this together and keep it on topic; The problem is how does Microsoft compete with Linux in the server market? The answer is that it doesn't. We know that, but they don't, or at least they are too arrogant to admit it. So instead they come up with a dumb answer to the "Competing with Linux" problem: "Thin" servers running WinCE. Fortunately this is flawed too. They propose a $1500-$2000 solution to a problem that Linux already solves for about $150 worth of hardware.
--- A Jesus Fish eating a Darwin Fish only proves Darwin's point.