Ted Leung on the air
Ted Leung on the air: Open Source, Java, Python, and ...
Ted Leung on the air: Open Source, Java, Python, and ...
Sat, 08 May 2004
Will IT matter more after it's commoditized?
Techdirt cites Hal Varian (author of Information Rules)'s reply to Nicholas Carr's article and book questioning the relevance of IT. Here are some quotes that struck me:
[23:37] |
[computers/open_source] |
# |
TB |
F |
G |
7 Comments |
So Mr. Carr's main thesis is right. It is not information technology itself that matters, but how you use it.
Standardization and commoditization of a technology don't always mean that innovation stops. Once products become commodities, they can serve as components for further innovation.
In the 19th century the real innovations came after the basic building blocks were commoditized. Perhaps information technology is like those standardized parts. Desktop PC's, Web servers, databases and scripting languages have become components in larger, more complex systems. As these components have become more standardized, the opportunities to create innovations have multiplied.It's hard to disagree that open source software is forcing the commoditization of entire classes of software. The question now becomes, will that commoditization result in innovation? The question is valid for open source software itself, since many large open source systems, Linux, GNOME, etc, are themselves composed of commoditized components.
I agree totally with Varian's point. But I must stress that Open Source software (or anyone else) is not situated as well to work towards componetization/commoditization of IT as Microsoft.
Why? Because, with dotnet, Microsoft has a coherent, single, strategy for componetizing software. Furthermore they have nearly fifteen years experience in developing componetized systems; with multiple iterations from which they learned something at each step (OLE, COM, COM+, DCOM, SOAP, dotnet).
Even those non-MS environments with a decent componetization strategy (Java Desktop, Gnome, Mozilla, etc.) don't stand much chance because they don't have the market penetration and kinetic energy of Microsoft in this area. Only Mono (which is playing catchup on dotnet) is capable of taking a ride on the juggernaut instead of being steam-rolled beneath it.
And there is no gaurentee the monkey will not end up crushed as well.
Microsoft has not been secretive about their long-term strategy in this area. And it is clear to me that they have been working actively towards this goal for this very reason for some time now. They won't be the only game in town, certainly, but they obviously intend to be the biggest player. And they have always understood the power of the network effect.
I'm not trying to be discouraging here, only realistic. And I am a big proponent of componetization and other build-block software strategies...
Posted by Jack William Bell at Sun May 9 12:12:24 2004
Why? Because, with dotnet, Microsoft has a coherent, single, strategy for componetizing software. Furthermore they have nearly fifteen years experience in developing componetized systems; with multiple iterations from which they learned something at each step (OLE, COM, COM+, DCOM, SOAP, dotnet).
Even those non-MS environments with a decent componetization strategy (Java Desktop, Gnome, Mozilla, etc.) don't stand much chance because they don't have the market penetration and kinetic energy of Microsoft in this area. Only Mono (which is playing catchup on dotnet) is capable of taking a ride on the juggernaut instead of being steam-rolled beneath it.
And there is no gaurentee the monkey will not end up crushed as well.
Microsoft has not been secretive about their long-term strategy in this area. And it is clear to me that they have been working actively towards this goal for this very reason for some time now. They won't be the only game in town, certainly, but they obviously intend to be the biggest player. And they have always understood the power of the network effect.
I'm not trying to be discouraging here, only realistic. And I am a big proponent of componetization and other build-block software strategies...
Posted by Jack William Bell at Sun May 9 12:12:24 2004
I agree that the .NET component story is very good. But remember that this is also about commoditization, which means the transformation of the software business from a high-margin business into a low-margin business, which will definitely have an impact on Microsoft.
Posted by Ted Leung at Sun May 9 16:17:24 2004
Posted by Ted Leung at Sun May 9 16:17:24 2004
Oh yeah, the move to a commodity market is going to hurt MS because they are going to have to cut margins. But they won't have to cut the margins as much as anyone else with a margin to cut. (This is where Open Source is better off.) And they have a lot of room for cutting if they have to.
The real point is that they are going to continue selling Office and their OS; always the two biggest moneymakers. And they are able to do this because those represent the cornerstones of IT in the future. Between existing product lock-in and the ease of componetization in those areas, new IT apps built on Office components will be everywhere. And because of the network effect of all your partners and competitors using this, your company is going to have to go along.
Somehow I think MS realizes the glory days of high margins are over. But I also think they have a strategy for moving forward. Remember, this is a company that acted like they were scrabbling for their lives when they had a 90% market share. They are always working hard to keep it going, using every tool at hand (even when it gets them in trouble). Right now they may be one of the few companies that is planning more than five years into the future.
Meanwhile, in that future, the real money isn't going to be in selling commoditized software. It is going to be in consulting and administration services. And that is an area where MS has always sucked. People just don't trust them enough to turn over the keys. So, in that case, their reputation works against them.
But it helps people like me...
Posted by Jack William Bell at Sun May 9 17:29:21 2004
The real point is that they are going to continue selling Office and their OS; always the two biggest moneymakers. And they are able to do this because those represent the cornerstones of IT in the future. Between existing product lock-in and the ease of componetization in those areas, new IT apps built on Office components will be everywhere. And because of the network effect of all your partners and competitors using this, your company is going to have to go along.
Somehow I think MS realizes the glory days of high margins are over. But I also think they have a strategy for moving forward. Remember, this is a company that acted like they were scrabbling for their lives when they had a 90% market share. They are always working hard to keep it going, using every tool at hand (even when it gets them in trouble). Right now they may be one of the few companies that is planning more than five years into the future.
Meanwhile, in that future, the real money isn't going to be in selling commoditized software. It is going to be in consulting and administration services. And that is an area where MS has always sucked. People just don't trust them enough to turn over the keys. So, in that case, their reputation works against them.
But it helps people like me...
Posted by Jack William Bell at Sun May 9 17:29:21 2004
Here's a related piece:
http://www.manageability.org/blog/stuff/modularity-and-conservation-of-profits
Which talks about how high margins can actually propagate from commodities to integration and back again. In short, just because something becomes a commodity, doesn't mean that eventually you can again reap the benefits of mastering that technology.
Now JWB comments "Open Source software (or anyone else) is not situated as well to work towards componetization/commoditization of IT as Microsoft" is completely ludicrous. Java systems are already componentized by default and if you want to find more advanced forms of componentization you'll find it in the java world (JMX, Eclipse plugins, Dependecy injection etc.)
Carlos
Posted by Carlos E. Perez at Mon May 10 04:07:20 2004
http://www.manageability.org/blog/stuff/modularity-and-conservation-of-profits
Which talks about how high margins can actually propagate from commodities to integration and back again. In short, just because something becomes a commodity, doesn't mean that eventually you can again reap the benefits of mastering that technology.
Now JWB comments "Open Source software (or anyone else) is not situated as well to work towards componetization/commoditization of IT as Microsoft" is completely ludicrous. Java systems are already componentized by default and if you want to find more advanced forms of componentization you'll find it in the java world (JMX, Eclipse plugins, Dependecy injection etc.)
Carlos
Posted by Carlos E. Perez at Mon May 10 04:07:20 2004
Two points to Carlos:
* Object-Orientation does not a component model make. Java does provide higher-level componetization features, but they don't have the market penetration that MS has. Which was the main thrust of my argument: That this kind of thing is extremely sensitive to the network effect.
* Java isn't exactly 'Open Source'...
Posted by Jack William Bell at Mon May 10 08:08:45 2004
* Object-Orientation does not a component model make. Java does provide higher-level componetization features, but they don't have the market penetration that MS has. Which was the main thrust of my argument: That this kind of thing is extremely sensitive to the network effect.
* Java isn't exactly 'Open Source'...
Posted by Jack William Bell at Mon May 10 08:08:45 2004
JWB,
I agree that there's more to componentization than OO, that why I gave concrete examples like Eclipse and Dependency Injection as examples that go beyond a standard OO model.
Java doesn't have market penetration? At the same time your saying Microsoft and therefore .NET has? C'mon who are you fooling?
Java isn't exactly 'open source', however 'open source' drives java. Witness the movement in EJB 3.0 that's driven by open source (i.e Hibernate and Spring).
Posted by Carlos E. Perez at Mon May 10 13:50:23 2004
I agree that there's more to componentization than OO, that why I gave concrete examples like Eclipse and Dependency Injection as examples that go beyond a standard OO model.
Java doesn't have market penetration? At the same time your saying Microsoft and therefore .NET has? C'mon who are you fooling?
Java isn't exactly 'open source', however 'open source' drives java. Witness the movement in EJB 3.0 that's driven by open source (i.e Hibernate and Spring).
Posted by Carlos E. Perez at Mon May 10 13:50:23 2004
"Once products become commodities, they can serve as components for further innovation."
I agree with this point; you only have to look around to see the effects of building-block software, from the Linux command line with its "
I disagree that Microsoft is in a better position to work from componentised software though. .Net doesn't really componentise anything useful - it's just another big library. Jabber is working to componentise the whole instant-messaging arena, for instance. As Jabber matures, you get the ability to easily link your application into a fairly simple instant messaging environment the likes of which you just don't get with MSN Messenger plugged into Exchange.
Why hasn't computer telephone taken off yet? MS has dialer.exe, TAPI, COM, scheduled tasks... Have you tried to get something to dial your modem and play a sound file? Does your modem support Unimodem V? It just doesn't click together nicely.
Maybe oneday it will be enough of a single block that it can be part of other applications - remote desktop with a voice chat channel, for instance. Or maybe not - after all, why would Microsoft change it?
Open Source software needs a couple of dedicated people to make a change, Microsoft needs... well, where would you start getting Microsoft to add command line options or COM control to the phone dialer mini-app?
Posted by sfb at Tue May 11 08:06:28 2004
I agree with this point; you only have to look around to see the effects of building-block software, from the Linux command line with its "
cat txtfile | grep string | wc" command chaining to using javascript bookmarklets to manipulate the DOM of an arbitrary webpage.
I disagree that Microsoft is in a better position to work from componentised software though. .Net doesn't really componentise anything useful - it's just another big library. Jabber is working to componentise the whole instant-messaging arena, for instance. As Jabber matures, you get the ability to easily link your application into a fairly simple instant messaging environment the likes of which you just don't get with MSN Messenger plugged into Exchange.
Why hasn't computer telephone taken off yet? MS has dialer.exe, TAPI, COM, scheduled tasks... Have you tried to get something to dial your modem and play a sound file? Does your modem support Unimodem V? It just doesn't click together nicely.
Maybe oneday it will be enough of a single block that it can be part of other applications - remote desktop with a voice chat channel, for instance. Or maybe not - after all, why would Microsoft change it?
Open Source software needs a couple of dedicated people to make a change, Microsoft needs... well, where would you start getting Microsoft to add command line options or COM control to the phone dialer mini-app?
Posted by sfb at Tue May 11 08:06:28 2004
You can subscribe to an RSS feed of the comments for this blog:
Add a comment here:
You can use some HTML tags in the comment text:
To insert a URI, just type it -- no need to write an anchor tag.
Allowable html tags are:
You can also use some Wiki style:
URI => [uri title]
<em> => _emphasized text_
<b> => *bold text*
Ordered list => consecutive lines starting spaces and an asterisk
To insert a URI, just type it -- no need to write an anchor tag.
Allowable html tags are:
<a href>
, <em>
, <i>
, <b>
, <blockquote>
, <br/>
, <p>
, <code>
, <pre>
, <cite>
, <sub>
and <sup>
.You can also use some Wiki style:
URI => [uri title]
<em> => _emphasized text_
<b> => *bold text*
Ordered list => consecutive lines starting spaces and an asterisk