With the push of a new campaign, Tools to Compete, Microsoft has effectively admitted that they see Linux as a legitimate threat. Even with the massive reduction in the cost of Windows in big deals, Microsoft has still lost customers. It almost causes one to ask – a passing trend, or the beginning of the end?
Microsoft Ups Competition With Linux
About The Author
Adam Scheinberg
Technology Executive • Web Developer • Father • Foodie • Music Snob • OS enthusiast
Follow me on Mastodon @[email protected]
>>Hardware support is just not there for Linux and may never be.
In my experience over the last year, I have found much better hadware support from Linux than I have with Windows. I just install a distro, and bam, drivers for everything and everything works. Done. But when I install Windows, I always have to install the video driver afterwards, I always have to install the printer driver afterwards, and I often have to install the ethernet and the sound drivers. While it may be easier to go to a manufacturer’s website and download Windows drivers for your hardware, there is no need to do this with Linux.
Your *router* crashed after 3 months? A router does very little. It should run indefinately.
I assume the routing software had a memory leak of some kind, I think the box just ran out of memory and swap space. It was running with win98 after all, not the best memory management to begin with
“Well, one thing from Outlook as already stated is the communication between the exchange server and the client. Mainly the ability to do the scheduling. This can be made available if running Exchange server 2000 or 2003, but unfortunately that has been cost prohibitive for us to update from 5.5. So until we can do that dead in the water.”
Things are evolving rapidly on the groupware side.
With respect to replacing exchange, there are a few solutions at this moment.
A friend of mine is currently using a solution that provides all kinds of extra’s to outlook, including the possibility of sharing schedules using a simple database:
http://www.outlookinside.nl/FAQ.htm
He used a simple .mdb database stored on a samba server for calendaring.
As far as I know, there are 2 OSS groupware solutions for the server side:
– http://opengroupware.org/
– http://kolab.kde.org/
Both need a plugin to be useful with outlook ( http://kolab.kde.org/kolab-plugins.html ):
* Bynari v1 (www.bynari.net)
License: proprietary
Tested by Kroupware project (Reference platform
NT5SP6, OL2000)
might not be available from Company anymore
* Bynari v2 (www.bynari.net)
License: proprietary
Untested, users reported problems
* Konsec Konnector (www.konsec.com)
License: proprietary
In development, alpha version was sighted at Cebit
* Toltec Connector (no website available yet)
License: proprietary
In development, no website, proposed release end of Q3
* otlkcon (otlkcon.sourceforge.net)
License: Free Software
In pre-alpha stage
That is true for most cases, but what about brand new hardware that does not how Linux support yet. What about installing new hardware? Are you gonna re-install the distro? And not all hardware has linux support. I had a DLink 650+ wireless card that had no support. Windows has support for everything and easy to install too. At least in most cases, if there is no support in Windows it most likely doesnt work anywhere.
“what about brand new hardware that does not how Linux support yet.”
Simply don’t buy it yet or use Windows.
It’s really not such a big deal. It’s just a matter of checking _before_ buying.
If I want a wireless card, I check for available cards first, then check which are supported best and only then I go out and buy one.
If a DLink 650+ doesn’t work, then I just buy something else. I don’t care for having the latest and the greatest as long as it works and I can do my job.
It’s just like using a vacuum cleaner. If it needs a new bag, buy a bag that fits!
When I buy something new, I want the best (taking into account price and application of course.) By taking Linux into the equation, I am now limiting my self and no longer buying the best for my need.
My point is that Linux limits the hardware you can use. Windows does not. End of story.
The vacuum example is just wrong.
It is like saying that I need to buy a car, and only able to pick a Ford. I don’t want to live in a world of only Fords.
So does Windows using your analogy, since it doesn’t run on a Mac. Linux does, so perhaps you are the one living in the a world of only fords. Linux supports more hardware than almost any other OS. Just because it may not support tomorrow’s technology, don’t believe for a second that your OS supports more than Linux. 😉
“OSS is a nice idea in theory but is limited to due to the fact that there is just no money in it. Why should a compny pour money into OSS development to get very little out of it.”
There are many reasons why companies might do that:
– market penetration.
Same trick as MS did with IE. Give something away for free to prevent competitors from taking over the market. In economic theory, this is called a “loss-leader”
– cost savings.
A lot of software is developed exclusively for in-house use. This software is not part of the core business of the company; they don’t make money by selling it. It is just something they need to make money some other way. In this case, they have the development costs anyway, so giving away the source code doesn’t hurt, but doing so does deliver somthing in return: development costs can suddenly be shared with the community and more money is saved because the community helps fixing bugs and might add new functionality for free.
– head start
If you want to develop software, sometimes you can gain a giant head start by building on an existing OSS solution that already does 80 or 90% of what you need.
Then, you can save 80-90% of your development costs for the same result.
This might not make sense if you want to make money selling your software, but it certainly does if you just need the software to make money some other way.
You know, selling software is not the only way you can make money writing it. There are other ways that sometimes make sense and sometimes they don’t.
You are right when when it comes to not selling the software and making money in other ways. I am a developer for a consulting firm, lots of software written and not actually sold. But there are limits.
My point was not that OSS is not useful. It is useful and it serves a purpose.
I just do not believe it will completely take over due to the limitiions of the lack of profit.
Lets take the market penetration example. Lets pretend that IE does not exist and Microsoft improved on Mozilla and included it free with Windows to gain market share (like a few years back.) Well, Windows is proprietery(sp?) and sold and Mozilla gains. So far so good.
But I argue this idea only works with some proprietery software and this proprietery software is of more valuable then the OSS software, at least in the mind of the company.
Obviously there has to be limits to this arguement. I am only gonna stick with x86 hardware.
So, on x86 hardware, Linux limits my choices of hardware while Windows does not.
This is not the place for a x86 vs. the world argument
“I just do not believe it will completely take over due to the limitiions of the lack of profit.”
I go with you with that. I also believe proprietary software has it’s place and OSS will not take over everything.
I think OSS will take over a lot of commodity softwarem though. Things everybody needs, like an OS, a browser and Office software.
These things have cristalled out in many ways and current OSS products are rapidly approaching the point were OSS products equal the proprietary ones in functionality.
In those cases OSS is very competitive and should kill the proprietary competitors if products were selected on pure economic reasons.
My theory is that more and more software that has reached a mature point will be surpassed by OSS competitors in due time.
Therefore, I think proprietary software will continue to be competitive in those markets that still need a lot of innovation, while OSS takes over yesterdays innovations.
That doesn’t mean that OSS can not be innovative; it is. It’s just not the only innovative way of developing software efficiently, so it has competitors.
“But I argue this idea only works with some proprietery software and this proprietery software is of more valuable then the OSS software, at least in the mind of the company.”
I’m not sure what you mean exactly.
What I understand is that you mean to say that it would have worked for MS because they sell it together with something else. The OS in this case.
But then, what?
“Yes thats marketshare not user base. If you use user base Windows probably has around 80%. Either way I have a hard time figuring out how people can’t decide wether they have a monopoly with 80-90% of a market. Marketshare is what percentage of each OS was sold/installed during a certain (few months ) time frame.”
I don’t agree with your point here.
I do agree that counting sells/installs is not accurate. But google does not count sells of installs. It counts what OSes are used to search on Google.
Now, I do agree that this is also not an 100% accurate measurement if you want to measure user base, since not all users would use Google equally often. Technical savvy people are relatively more likely to work on Linux and those are the people that use internet relatively much.
On the other hand, most browsers report themselves as MS browsers, which might influence the count, although I might hope that Google found a way around that.
But, is there anything better?
How on earth would you measure user base? With all the legal/illegal copying there is no way to backtrace.
So, you’d have to use something on the internet. In that sense, I think the Google measurement is the most accurate measurement we have.
Yikes!! I wasn’t making much sense. Sorry. I am tired and sick (which is why I am not at work, hehehe:)
I think we have lost focus and it is my fault. I love some OSS. I use mozilla, I am acutally typing this using vim through a w3m browser.
OSS works for some things: Browser, Text Editor, even something like Apachee.
I agree with you and the idea of OSS taking over past inovations.
I have always said that OSS has its place. I feel that this place is close to where it is now.
I never see Linux challenging Windows on the desktop.
“I never see Linux challenging Windows on the desktop.”
I think it depends on 1 thing. And that is on how many companies will switch over to cross platform development tools, like Qt.
If Linux would capture like 5% of the market and Apple keeps it’s 3%, then they would have 8% together. Both can be easily targeted with both Qt and WxWindows.
That means that if you develop with cross platform tools, you get to enter 8% of the market with marginal extra costs. And, the 8% Linux/Mac is like completely open terrain for a lot of markets.
For example, there’s hundreds of small accounting packages out there. On Linux, there may be 10.
There’s, what, 10.000? native games for Windows. For Linux, there might be 500.
There’s all kinds of small shareware stuff for Windows. Virtually nothing for Linux.
What I mean to say is that by the time Linux reaches 3-5%, it becomes an interesting market, _if_ you can produce it for little extra costs.
Because a toolset like Qt can be used for free on Linux, it is very interesting for start-up companies, like the many innovative companies that are started by Engineers that just graduated from Universities.
These companies have a competitive advantage to companies that are locked into MS’s tools. They get to keep 8% of the market for themselves!
And that might just be the edge Linux needs.
There’s one more thing.
If you use open source cross platform tools, you can open up parts of your software.
You have the ability to selectively block your competitors out of your 8% (potentially growing) market.
You release just enough of your product under the GPL so that you have to look twice in your rear-view mirror to see your competitors once.
On other words: cross platform tools have the potential of giving you a virtual monopoly on 8% of the market.
Then add the (small) army of volunteers that help _you_ compete.
That’s starting to sound _very_ compelling to me.
I mean, how much do you think 8% of the Office market will be?
Or the small-business accounting market?
If you were Sun, would you settle for 1% of 8% of the Office market, while on the same time keeping your competitor out of your way?
You make a good point, but I feel that 8% is much too small of a market for companies to start thinking about using cross-platform toolkits. Microsoft supporting their toolkits gives them a huge edge over over cross-platform tool-kits.
This got me thinking. Lets say you have a company that makes an open-source office package equal to office in everyway and it is cross-platform.
How do you distribute it?
Through an apt-get type system? No money there.
Sell in stores or online? People will just get the source and compile themselves.
Sell to people who don’t know how to compile? What stops others from getting your source and distributing it for free?
Not provide source? Linux people won’t buy.
Why make software if no one can get it?
The trouble is the Linux market won’t grow unless there is quality software and there won’t be quality software unless market grows.
Also, if a cross-platform tool-kit is used. Where is the reason to leave Windows? Doesnt change the fact that hardware sucks under linux and multimedia is much better under windows and the other reasons for using windows to begin with.
Coming to think of it, the release of Open Office costed Sun like, what?
100 million dollars?
According to
http://news.com.com/2100-1001-277033.html?legacy=cnet
Sun bought StarOffice for $73.5 million dollars.
Then $100 million might not be too far of.
Let’s find out what MS earns on Office. According to http://money.cnn.com/2001/07/19/technology/earns_microsoft/
:
“Connors also highlighted the company’s desktop applications business, through which it derives the bulk of its revenue. Sales from that unit rose 8.7 percent to $2.5 billion from $2.3 billion.”
So, MS earns about $2.5 billion on it’s desktop applications, mainly word.
That means: Sun would minimally earn 1% of 8% of $1-2.5 billion, which is $800.000 – $2.000.000 a year.
Not too bad. And that’s with only 1% of 8% of the currently _paying_ market. It excludes everything currently pirated!
This means, my estimate of Sun’s earnings is probably a very pessimistic estimate.
I think I would also have taken the gamble, if I could afford it financially.
Where is the reason to leave Windows?
Money.
Linux costs like $100 for a cd that you can use in your _whole_ company for as long as _you_ want.
It means you have a consistent desktop across your whole company, since everybody is using the same version of the OS.
What windows users call “Roaming Profiles” would actually work across you company, no matter wether the computer you’re happen to use is 1 or 5 years old, because everybody is using the same version of the OS.
Rolling out software in your company is easier, because you’ve got to deal with only one version of the operating system.
System maintainance is easier, because everybody is using the same version and Linux is really easy to administer remotely.
Hardware is cheaper, because Linux supports older hardware, if only by hooking older hardware up to newer cpu’s as X terminals.
Read the stuff about Ball, if you want hard evidence:
http://www.osnews.com/story.php?news_id=4303
“I feel that 8% is much too small of a market for companies to start thinking about using cross-platform toolkits.”
How are you going to get into .NET?
Are you going to rewrite and debug your software in c#, or are you going to let your toolkit provider figure out the hard stuff, so you can just recompile and link your current c++ software into .NET as a dll?
“This got me thinking. Lets say you have a company that makes an open-source office package equal to office in everyway and it is cross-platform.
How do you distribute it?”
Ok.
I would take a good look at epm: http://www.easysw.com/epm/
They claim EPM Can:
* Generate portable script-based distribution packages complete with installation and removal scripts.
* Generate vendor distributions in AIX, BSD, Compaq Tru64, Debian, HP-UX, IRIX, MacOS X, Red Hat, and Solaris formats.
* Provide a complete, cross-platform software distribution solution for your applications.
For the record: Except for Windows, AFAIK
“Not provide source? Linux people won’t buy.”
Don’t forget we’re talking about the situation that Linux somehow got 5% of the market. Those 5% is going to contain a substantial portion of business users, who have no problem installing their own software, especially if you deliver it in their OSes package format.
There’s no reason these customers would not be prepared to take a little action on themselves. After all, somehow they managed to get Linux up&running, so there _must_ be someone that can install a package?
And, remember, no one else is interested, so where else would they go if no open source alternative is available?
They have no other choice but running a windows package under an emulator. These things cost money, if only because for a reliable one you still need a Windows license.
And that’s why native packages are compelling, and people _are_ willing to pay, at least up to the cost of a Windows license.
Hmm. Wonder what 1% of 8% of the Windows licensing revenue is worth :-))
“Why make software if no one can get it?”
I think I made my point.
“The trouble is the Linux market won’t grow unless there is quality software and there won’t be quality software unless market grows.”
My point is: somewhere, there’s gonna’ be companies that _do_ develop cross platform. They can have an extra piece of action for virtually nothing.
These companies have a better chance to grow, because they have a larger _and_ opener market. This gives them also a better chance of survival on the big Windows market.
To put it even simpler: they have a competitive advantage.
So, my point is that by the time Linux reaches 5% of the market, and that’s what you gave me if I interpret your comments correnly, there’s no way to stop it.
At some point, the Linux market is going to be so big that Qt will release it’s Windows port under the GPL.
Then you have the option of rapidly penetrating the Windows market by opening up the non-competitive parts of your software.
By the way, with WxWindows you have that option right away. But I doubt if their .net support will come soon enough. My bet is for Qt on that one.
My point is: _if_ Linux reaches 5% of the market, you have a fair chance of making a lot of money on the Linux _and_ Windows market _if_ you make the important decision now.
I would say that if you _can_ switch to a cross platform development toolkit within a reasonable budget: go for it. There’s nothing to lose.
“How are you going to get into .NET?”
Why get into .NET? I would rather use CGI and Java and have the same benefit, as well as be able to pick the OS it runs on. With .NET it ties you into MS once again where you can not change if you wanted to down the road, with no added functionality that is not already out there.
“Why get into .NET?”
I don’t see the point in going to .NET either, but I guess what I meant to say is that if you’re into the MS path, you probably can’t escape .NET forever. MS will do everything they can to keep you into their upgrade mill.
In that case, you’re going to have to invest in porting anyway.
Therefore, I questioned your argument that 8% might not be enough to justify investing in cross-platform development tools.
Again, that doesn’t mean that everybody has to go along, but just that a substantial number of developers will have to invest into porting activities, be it to .NET or to cross platform, which includes Java.
Companies that are in the leading position might just go the way of least resistance following MS, but then they do leave a niche market of in the order of 8% completely open for any competor that might be around.
So, if you’re the underdog, don’t ignore the Non-MS market. StarOffice, for example, survived for years on a small portion of the Windows market and the OS/2 market before Sun bought them.