Hmm, this sounds familiar.. as if I had heard the same sentence many times, but never when speaking about openstep. Could it be that they never heard about Java and C#?
Let the the thing die as everybody knows it is already dying...
I don't see why some people have such a hard time accepting it. If some people are ready to develop software for free, it naturally reduces the chances for other people to get paid for it. sure, there still is the need to customize and support it further but the jobs for software product development are gone. Just think about all the jobs involved in developing the commercial products for which open source alternatives exist. If the open source alternatives ever really gained widespread adoption, there would be MUCH less IT jobs around.
Most people seem to agree with the rather heavy emphasis of mathematics on CS degree, but I don't. And yes, I do have Msc degree in CS and almost 10 years of work experience on software development. While I agree that mathematics are very important in certain fields of software development, I would say that for most software development jobs the only benefit of the math classes is the supposedly improved logical thinking.
Nowadays, and even more so in the future, development tools take care of most of the work that previously have been done by software developers. They optimize your code, and they provide the algorithms and data structures. The people who still need math are either doing low-level programming or work in some special field such as games development. I would estimate that those jobs are less than 30% of all programming jobs, and that it will decrease in the future.
In the software projects I have been involed with, the key shortcomings have typically been in requirements gathering, project management, version control etc. I.e. while we did good software we got it done later than we estimated and it did not meet the real requirements of the customers even though they of course had approved the requirements documents. Thus, increasing the understanding of software development processe and project management at the expense of mathematics is not necessarily a bad idea in my opinion. Personally, I remember very little if nothing of all the math courses I had to take at my university even though it is not THAT long ago.
As an other MS developer, I would say that the differences between C# and Java really are not an issue. But the IDE (vs.net) really is. It makes coding, debugging, version control, deployment, db integration etc. so easy that I really don't want to go back to developing with linux. The truth is that in Linux/OSS development you have to use several different tools with different UIs that work only adequately together. With vs.net you have one single consistent IDE for everything and everything is also well document. While some people may get their kicks for tracking down a bug in a 3-tier architecture with primitive tools, I get the kicks for being able to be productive.
I think Avalon/XAML is second in utility to WinFS. WinFS is supposed to bring capabilities formerly found only in document management systems to your OS, and that will be very attractive to big enterprises and CIOs. Finding the right document, and the latest version of it, is a huge problem in enterprises. It is something the end users care about, and something the CIO cares about.
Introducing avalon is a good move from Microsoft, but I don't see it being the reason anyone would buy Longhorn. Not until it is in wide use, and that will take few years. Avalon simply offers a nicer version of HTML that is useful in developing GUIs for your custom solutions. For most enterprises, that is definitely not enough for a justifying a major expense like upgrading the OS in every computer in the enterprise.
While Avalon is nice for the developers and offers also richer user experience, I think developers and IT people will be won over Indigo, as it will (supposedly) make writing distributed systems much much easier than before.
WinFS makes the end users and CIOs want to upgrade to Longhorn, and Indigo wins over the IT staff. Avalon's role is to make sure that after people start to use longhorn, they cannot switch back to any OS that does not support Avalon, ie. for anything non-windows like Linux.
And the comparison is imo clearly wrong in several points, for example regarding c# and arrays. Arraylists and hashtables are part of the.NET base class library, and as such I consider them part of C# also. So how can he say that C# does not support resizable arrays and associative arrays?
UN/CEFACT had important role in the standardization of EDI but the rules have changed since then. I doubt Microsoft would be stupid enough to pay much for influencing the work of UN/CEFACT.
Standards for business-to-business data exchange are not useful unless fairly popular. Network effect is very strong in this area. ebXML has been around for sometime and it also had the contributions of OASIS behind it. Yet there are few (none?) implementations of ebXML in production use. Why would BCF be any more succesful? The successor of EDI will most likely be RosettaNet. Microsoft is also a member of RosettaNet, but I doubt it has much to say in the development of that standard. The cost and significance of software is irrelevant in the total implementation costs of business-to-business data exchange.
My question is what does this mean to Symbian. The alliance to prevent Microsoft from controlling the operating systems of future mobile phones seems to have vanished now that Symbian is becoming a subsidiary of Nokia. How come the other big mobile phone manufacturers were not interested in buying Psion's Symbian shares? I would think that Nokia wouldn't have bought the shares unless it absolutely had to. Must be hard now to convince other mobile phone manufacturers that Symbian OS does not favour Nokia over other companies. Does this mean that the chances to see Microsoft OS in my next mobile phone have increased?
I think the current market situation is much closer to free market than monopoly. As far as I know it is still legal to develop, market and sell software products almost everywhere in the world. And for almost every software product need, there are alternatives - like linux and open office. The freedom to buy the software product that you want is pretty common as well.
While I agree that the "vendor of product W" is not known for playing fairly, I believe that rules of capitalism still apply for software products. As most people currently choose Product W over product L, there must be reasons for it. If people use product W simply because it is easy to exchange office documents with it, that is a valid reason too. The reasons probably vary from one person to another though.
The rulings of US courts on whether this situation is monopoly or not are in my opinion irrelevant. The rulings are based too much on politics and I am sure that all of us can think of many rulings by US courts that we find simply unbelievable.
I believe that the users of product W have much to say about the evolution of the product. More so, in many cases, than the users of product L.
As the product W is a commercial product, the sole reason for its existence is to bring money to the vendor. Money that comes only if the users decide to buy the product. Thus there is very strong incentive to listen to what the users of the product W want.
The Product L on the other hand is not a commercial product and the reason for its existence is not to please the users but the developers of the product. In some happy occasions the needs of the developers and the users are same, but that is not always the case.
Combine with the stronger incentive to please the users the larger amount of available resourcers, it is not surpising that the Product W is in many cases superior to Product L - from the users point of view.
Slashdot Mirror
Hmm, this sounds familiar.. as if I had heard the same sentence many times, but never when speaking about openstep. Could it be that they never heard about Java and C#? Let the the thing die as everybody knows it is already dying...
I don't see why some people have such a hard time accepting it. If some people are ready to develop software for free, it naturally reduces the chances for other people to get paid for it. sure, there still is the need to customize and support it further but the jobs for software product development are gone. Just think about all the jobs involved in developing the commercial products for which open source alternatives exist. If the open source alternatives ever really gained widespread adoption, there would be MUCH less IT jobs around.
Most people seem to agree with the rather heavy emphasis of mathematics on CS degree, but I don't. And yes, I do have Msc degree in CS and almost 10 years of work experience on software development. While I agree that mathematics are very important in certain fields of software development, I would say that for most software development jobs the only benefit of the math classes is the supposedly improved logical thinking.
Nowadays, and even more so in the future, development tools take care of most of the work that previously have been done by software developers. They optimize your code, and they provide the algorithms and data structures. The people who still need math are either doing low-level programming or work in some special field such as games development. I would estimate that those jobs are less than 30% of all programming jobs, and that it will decrease in the future.
In the software projects I have been involed with, the key shortcomings have typically been in requirements gathering, project management, version control etc. I.e. while we did good software we got it done later than we estimated and it did not meet the real requirements of the customers even though they of course had approved the requirements documents. Thus, increasing the understanding of software development processe and project management at the expense of mathematics is not necessarily a bad idea in my opinion. Personally, I remember very little if nothing of all the math courses I had to take at my university even though it is not THAT long ago.
As an other MS developer, I would say that the differences between C# and Java really are not an issue. But the IDE (vs.net) really is. It makes coding, debugging, version control, deployment, db integration etc. so easy that I really don't want to go back to developing with linux. The truth is that in Linux/OSS development you have to use several different tools with different UIs that work only adequately together. With vs.net you have one single consistent IDE for everything and everything is also well document. While some people may get their kicks for tracking down a bug in a 3-tier architecture with primitive tools, I get the kicks for being able to be productive.
I think Avalon/XAML is second in utility to WinFS. WinFS is supposed to bring capabilities formerly found only in document management systems to your OS, and that will be very attractive to big enterprises and CIOs. Finding the right document, and the latest version of it, is a huge problem in enterprises. It is something the end users care about, and something the CIO cares about.
Introducing avalon is a good move from Microsoft, but I don't see it being the reason anyone would buy Longhorn. Not until it is in wide use, and that will take few years. Avalon simply offers a nicer version of HTML that is useful in developing GUIs for your custom solutions. For most enterprises, that is definitely not enough for a justifying a major expense like upgrading the OS in every computer in the enterprise.
While Avalon is nice for the developers and offers also richer user experience, I think developers and IT people will be won over Indigo, as it will (supposedly) make writing distributed systems much much easier than before.
WinFS makes the end users and CIOs want to upgrade to Longhorn, and Indigo wins over the IT staff. Avalon's role is to make sure that after people start to use longhorn, they cannot switch back to any OS that does not support Avalon, ie. for anything non-windows like Linux.
And the comparison is imo clearly wrong in several points, for example regarding c# and arrays. Arraylists and hashtables are part of the .NET base class library, and as such I consider them part of C# also. So how can he say that C# does not support resizable arrays and associative arrays?
UN/CEFACT had important role in the standardization of EDI but the rules have changed since then. I doubt Microsoft would be stupid enough to pay much for influencing the work of UN/CEFACT.
Standards for business-to-business data exchange are not useful unless fairly popular. Network effect is very strong in this area. ebXML has been around for sometime and it also had the contributions of OASIS behind it. Yet there are few (none?) implementations of ebXML in production use. Why would BCF be any more succesful? The successor of EDI will most likely be RosettaNet. Microsoft is also a member of RosettaNet, but I doubt it has much to say in the development of that standard. The cost and significance of software is irrelevant in the total implementation costs of business-to-business data exchange.
My question is what does this mean to Symbian. The alliance to prevent Microsoft from controlling the operating systems of future mobile phones seems to have vanished now that Symbian is becoming a subsidiary of Nokia. How come the other big mobile phone manufacturers were not interested in buying Psion's Symbian shares? I would think that Nokia wouldn't have bought the shares unless it absolutely had to. Must be hard now to convince other mobile phone manufacturers that Symbian OS does not favour Nokia over other companies. Does this mean that the chances to see Microsoft OS in my next mobile phone have increased?
I think the current market situation is much closer to free market than monopoly. As far as I know it is still legal to develop, market and sell software products almost everywhere in the world. And for almost every software product need, there are alternatives - like linux and open office. The freedom to buy the software product that you want is pretty common as well.
While I agree that the "vendor of product W" is not known for playing fairly, I believe that rules of capitalism still apply for software products. As most people currently choose Product W over product L, there must be reasons for it. If people use product W simply because it is easy to exchange office documents with it, that is a valid reason too. The reasons probably vary from one person to another though.
The rulings of US courts on whether this situation is monopoly or not are in my opinion irrelevant. The rulings are based too much on politics and I am sure that all of us can think of many rulings by US courts that we find simply unbelievable.I believe that the users of product W have much to say about the evolution of the product. More so, in many cases, than the users of product L.
As the product W is a commercial product, the sole reason for its existence is to bring money to the vendor. Money that comes only if the users decide to buy the product. Thus there is very strong incentive to listen to what the users of the product W want.
The Product L on the other hand is not a commercial product and the reason for its existence is not to please the users but the developers of the product. In some happy occasions the needs of the developers and the users are same, but that is not always the case.
Combine with the stronger incentive to please the users the larger amount of available resourcers, it is not surpising that the Product W is in many cases superior to Product L - from the users point of view.