Except that the code generated from the class diagrams just contains the method stubs. You still need to open the source for each class in a code editor and type the implementation. Gates is talking about visual programming (i.e. drag and drop and you're done) instead of visual design.
Actually, Average Joe User doesn't even know that Linux is an available option.
The context of this post is that Average Joe Linux User is overwhelmed by all the seemingly duplicitous applications that come bundled with the popular Linux distros.
In response to your post, it just depends on what criteria you evaluate your options. If price is a factor, then perhaps Linux wins.
I'm confused. The last time I looked at mono (which was before the aquisition), it was aimed more at web forms than win forms. Have they changed direction and are now going after win forms?
Your reaction to being overwhelmed by choice is actually quite typical. People seem to have a problem when presented with a set of choices without a clear winner. I think we have an algorithm that works like this.
repeat { __foreach option available { ____evaluate option __} } until there is a clear winner
Which is fine unless, of course, there are two or more options that have the same value. Then you have an eternal loop. Wouldn't this algorithm be better?
foreach option available { __evaluate option __if good enough { ____choose this option __} } choose the best option available
So, you're presented with mozilla and konquerer. You try mozilla first and find that it is good enough. You stop looking. Don't click on that konquerer icon. You're done. Simple, really.
MS-Office has bugs. OO has bugs. So what? If it doesn't have bugs, then it isn't software.
Now, let's see. I can shell out $400 and have an office productivity suite that has some bugs in it but generally I can work with it or I can shell out $0 and have an office productivity suite that has some bugs in it but generally I can work with it. Hmmm, I wonder which one is the better solution?
One more thing. Every dollar that MS-Office sales bring in goes to support their expansion in some other market. MS-Office is the cash cow that underwrites such forays into the server market as SQL Server, IIS, and Exchange.
I'm surprised that I didn't see another post along these lines. Moderators, feel free to mark this as redundant if I missed it.
I was a coder before M$FT gained its power. Back then, IBM was the 800 pound gorilla. I hated IBM because their tools were so primitive and expensive. I prayed for some upstart company to transform the market. Be careful what you ask for.
Unix was very expensive too. I paid over $1K for a port of Sys V to the PC of that day.
My take on the market at that time was that the other vendors were very greedy and elitist. They wanted software development to be so difficult that only the smartest and the best could ever do it. They charged as if they thought that only a very few people would ever write software. Certainly not the millions that write code today.
M$FT changed all that. Their take was to make software development easier so that more people could do it. They could sell more licenses and make it up on volumn. Also, they would leverage all that development since it locked the employers into their technology. Did it cause a lot of lame code to be written? Yes, but from a business perspective, it made a lot more sense than the other, elitist, approach.
Of course, open source would have eventually changed all that anyway. M$FT got there first but, in the end, software would become commoditized with or without Bill.
M$FT also was very aggressive on their competition to the point where there really is no place in the horizontal tool space for new vendors without deep pockets or backing from an already established player.
Would this have happened anyway? Probably so. M$FT did it in a way that was very high profile but other companies stifle this kind of innovation that comes from competition too.
Why should Linux go after the Macintosh market? That doesn't make sense. Mac users don't mind paying top dollar for software so why would OSS be appealing to them? Linux should tarket those markets that are very focused on TCO such as business, non-profits, and schools.
P.S. I take umbrage at your derogatory reference to GIMP which I find to be much superior to PS if what you are doing is web graphics.
To assume that a printer that is marketed as easier for women to use treats women as if they are more stupid than men is an uncritical inference.
After all, if men are willing to buy printers that are hard to use, then perhaps it is the men who are more stupid than women. When was the last time you bought something because it was hard to use?
It has been years since I used Lilypond but, back then, I used a MIDI editor to output a MIDI file and used a Lilypond command line tool to convert the mdi file to an ly file.
Improvisors don't have to but everyone else who call themselves a musician will have to be able to read music. Unless they are able to play something perfectly just by hearing it and then be able to remember it indefinitely. This begs two questions.
The first question is what is a musician? If I can play an album of music, am I a musician? If I play multiple albums simultaneously in a musically interesting way, am I a musician? If I play a modern keyboard that can play arpeggios automatically, am I a musician? Do I have to play in realtime to be called a musician? Do I have to be able to play a musical instrument at all to be called a musician? Are composers muscians?
The second question is what does it mean to read music? Can I read music if all I can do is just read the bar chord icons? Can I read music if some of the subtleties of a particularly technical score are lost on me? How hard a piece does it have to be before I can read music? If I can follow the notes to Twinkle, Twinkle Little Star, then can I read music?
From a developer's standpoint, I wish that you were correct. It is much easier to write CLI tools than it is to write windows or web based ones.
You propose that the barrier to CLI is lack of an easy search capability. Building some front end to man and/or apropo would be very easy and was proposed in other posts. My own experience leads me to believe that searching for the desired command is not the only barrier to entry.
I used to be on the adjunct faculty to the USF. I taught a class on New Software Paradigms. In that class, I would show some J2EE stuff on a Linux laptop. I ran an X session but sometimes I launched a bash window and typed in some commands to show them what had to be started in order for certain tiers to work (i.e. postgresql, rmiregistry). Whenever I did, I noticed that the students looked nauseated.
These were not developers but they were not ludites either. This reaction did not come from a single class but from years of teaching this class. It was almost as if it was beneath them to type in commands on order to get things done.
I think that the post about Oracle looking at acquiring BEA is more a response to their threat to Oracle in the tool space.
Oracle constantly wants to own the entire app server, not just the database tier. They come up with all these technologies to do this but these technologies all fail miserably. Their core competancy is clearly the database. If they want to own the app tier, then they should just purchase one. BEA has demonstrated success there.
Also, the very nature of EJB's entity beans precludes the need for a high performance database. Why does the database need to be fast when the corresponding objects will live in the memory of the app server. Want better response? Just buy more memory.
This is hair splitting. You could also argue that compiled C code is "interpreted" by the CPU's nanocode. It may be academically true but no one in the field would agree and you would be labeled as an idiot were you to casually refer to C as an interpreted language.
The same is so for Java and.NET languages such as C#. They are not in the same category as interpreted perl or python. You still have to take the extra step to compile programs written in Java or C# before they can be run. Also, Java and C# have support for compile time checking of static typing (which, IMHO, is a good thing).
You are correct. It is the person. The cost of interviewing would be very high if you had a qualified person interview everyone who applied for the position as that person would be interviewing full time. So, you use the resume to weed out candidates who don't seem likely to qualify.
When looking at a resume, positive experience is the most likely and compelling determinant for getting that interview. How does someone new to the profession get a chance? Certification exams were invented to address that need.
It is both an art and a science to develop a test that truly and accurately evaluates the tester's capabilities and knowledge. It is not easy. IMHO, MSFT's attempts to develop such exams are a complete and dismal failure.
In the interest of full disclosure, I feel compelled to reveal that it has been seven years since I have taken (and passed) a Microsoft certification program. At that time, my employer needed a certain number of developers to be certified in order to improve the company's relationship with MSFT. I already had a BSCE and eight years of professional experience. I could tell right away that the tests were pitiful at evaluating knowledge or experience.
My experience with more recent MSCEs seem to indicate that nothing has changed. Some are good. Some are bad. The MSCE status indicates only that you had enough initiative and discipline to stop playing video games long enough to do something about your career.
The ubiquitousness of OWL/RDF on the web would, IMHO, be tied to two things; adoption by a large search engine such as google and the availability of a cool tool that would make the authoring process easy.
Imagine a mind mapping tool like freemind that could export to OWL. That would work for static content. Content publishers would do it for the same reason that they include keyword meta tags in their pages now.
Slashdot Mirror
Except that the code generated from the class diagrams just contains the method stubs. You still need to open the source for each class in a code editor and type the implementation. Gates is talking about visual programming (i.e. drag and drop and you're done) instead of visual design.
Actually, Average Joe User doesn't even know that Linux is an available option.
The context of this post is that Average Joe Linux User is overwhelmed by all the seemingly duplicitous applications that come bundled with the popular Linux distros.
In response to your post, it just depends on what criteria you evaluate your options. If price is a factor, then perhaps Linux wins.
I'm confused. The last time I looked at mono (which was before the aquisition), it was aimed more at web forms than win forms. Have they changed direction and are now going after win forms?
Your reaction to being overwhelmed by choice is actually quite typical. People seem to have a problem when presented with a set of choices without a clear winner. I think we have an algorithm that works like this.
Which is fine unless, of course, there are two or more options that have the same value. Then you have an eternal loop. Wouldn't this algorithm be better?
So, you're presented with mozilla and konquerer. You try mozilla first and find that it is good enough. You stop looking. Don't click on that konquerer icon. You're done. Simple, really.
MS-Office has bugs. OO has bugs. So what? If it doesn't have bugs, then it isn't software.
Now, let's see. I can shell out $400 and have an office productivity suite that has some bugs in it but generally I can work with it or I can shell out $0 and have an office productivity suite that has some bugs in it but generally I can work with it. Hmmm, I wonder which one is the better solution?
One more thing. Every dollar that MS-Office sales bring in goes to support their expansion in some other market. MS-Office is the cash cow that underwrites such forays into the server market as SQL Server, IIS, and Exchange.
He stood on the shoulders of giants just as G. Spencer Brown stood on his shoulders.
I'm surprised that I didn't see another post along these lines. Moderators, feel free to mark this as redundant if I missed it.
I was a coder before M$FT gained its power. Back then, IBM was the 800 pound gorilla. I hated IBM because their tools were so primitive and expensive. I prayed for some upstart company to transform the market. Be careful what you ask for.
Unix was very expensive too. I paid over $1K for a port of Sys V to the PC of that day.
My take on the market at that time was that the other vendors were very greedy and elitist. They wanted software development to be so difficult that only the smartest and the best could ever do it. They charged as if they thought that only a very few people would ever write software. Certainly not the millions that write code today.
M$FT changed all that. Their take was to make software development easier so that more people could do it. They could sell more licenses and make it up on volumn. Also, they would leverage all that development since it locked the employers into their technology. Did it cause a lot of lame code to be written? Yes, but from a business perspective, it made a lot more sense than the other, elitist, approach.
Of course, open source would have eventually changed all that anyway. M$FT got there first but, in the end, software would become commoditized with or without Bill.
M$FT also was very aggressive on their competition to the point where there really is no place in the horizontal tool space for new vendors without deep pockets or backing from an already established player.
Would this have happened anyway? Probably so. M$FT did it in a way that was very high profile but other companies stifle this kind of innovation that comes from competition too.
Cygwin is really great! It is free to use unless you want to include it in a non-GPL app. If you do, then it is very, very expensive.
Exactly. It's not that Linux is hard. It's just that Linux is perceived as hard. Unfortunately in this case, perception is everything.
Why should Linux go after the Macintosh market? That doesn't make sense. Mac users don't mind paying top dollar for software so why would OSS be appealing to them? Linux should tarket those markets that are very focused on TCO such as business, non-profits, and schools.
P.S. I take umbrage at your derogatory reference to GIMP which I find to be much superior to PS if what you are doing is web graphics.
Take a look at dia as your Linux visio replacement.
To assume that a printer that is marketed as easier for women to use treats women as if they are more stupid than men is an uncritical inference.
After all, if men are willing to buy printers that are hard to use, then perhaps it is the men who are more stupid than women. When was the last time you bought something because it was hard to use?
It has been years since I used Lilypond but, back then, I used a MIDI editor to output a MIDI file and used a Lilypond command line tool to convert the mdi file to an ly file.
Improvisors don't have to but everyone else who call themselves a musician will have to be able to read music. Unless they are able to play something perfectly just by hearing it and then be able to remember it indefinitely. This begs two questions.
The first question is what is a musician? If I can play an album of music, am I a musician? If I play multiple albums simultaneously in a musically interesting way, am I a musician? If I play a modern keyboard that can play arpeggios automatically, am I a musician? Do I have to play in realtime to be called a musician? Do I have to be able to play a musical instrument at all to be called a musician? Are composers muscians?
The second question is what does it mean to read music? Can I read music if all I can do is just read the bar chord icons? Can I read music if some of the subtleties of a particularly technical score are lost on me? How hard a piece does it have to be before I can read music? If I can follow the notes to Twinkle, Twinkle Little Star, then can I read music?
The only hope here is that the people who made the trailer cut it in such a way as to misrepresent the actual content of the movie.
That's not true. I submitted a story about Lowe's migrating their Point Of Purchase machines to Linux and it was rejected.
I got mine for $10 in the discount bin before Loki officially went under. $148, you say? Perhaps it's time to put mine up on ebay.
From a developer's standpoint, I wish that you were correct. It is much easier to write CLI tools than it is to write windows or web based ones.
You propose that the barrier to CLI is lack of an easy search capability. Building some front end to man and/or apropo would be very easy and was proposed in other posts. My own experience leads me to believe that searching for the desired command is not the only barrier to entry.
I used to be on the adjunct faculty to the USF. I taught a class on New Software Paradigms. In that class, I would show some J2EE stuff on a Linux laptop. I ran an X session but sometimes I launched a bash window and typed in some commands to show them what had to be started in order for certain tiers to work (i.e. postgresql, rmiregistry). Whenever I did, I noticed that the students looked nauseated.
These were not developers but they were not ludites either. This reaction did not come from a single class but from years of teaching this class. It was almost as if it was beneath them to type in commands on order to get things done.
They are still pulling the same trick only now they call their COM version J++ with WFC and their .NET version J#.
Those savvy enough to tell the difference in the branding are also savvy enough not to swallow the poison pill to begin with.
I think that the post about Oracle looking at acquiring BEA is more a response to their threat to Oracle in the tool space.
Oracle constantly wants to own the entire app server, not just the database tier. They come up with all these technologies to do this but these technologies all fail miserably. Their core competancy is clearly the database. If they want to own the app tier, then they should just purchase one. BEA has demonstrated success there.
Also, the very nature of EJB's entity beans precludes the need for a high performance database. Why does the database need to be fast when the corresponding objects will live in the memory of the app server. Want better response? Just buy more memory.
This is hair splitting. You could also argue that compiled C code is "interpreted" by the CPU's nanocode. It may be academically true but no one in the field would agree and you would be labeled as an idiot were you to casually refer to C as an interpreted language.
The same is so for Java and .NET languages such as C#. They are not in the same category as interpreted perl or python. You still have to take the extra step to compile programs written in Java or C# before they can be run. Also, Java and C# have support for compile time checking of static typing (which, IMHO, is a good thing).
You are correct. It is the person. The cost of interviewing would be very high if you had a qualified person interview everyone who applied for the position as that person would be interviewing full time. So, you use the resume to weed out candidates who don't seem likely to qualify.
When looking at a resume, positive experience is the most likely and compelling determinant for getting that interview. How does someone new to the profession get a chance? Certification exams were invented to address that need.
It is both an art and a science to develop a test that truly and accurately evaluates the tester's capabilities and knowledge. It is not easy. IMHO, MSFT's attempts to develop such exams are a complete and dismal failure.
In the interest of full disclosure, I feel compelled to reveal that it has been seven years since I have taken (and passed) a Microsoft certification program. At that time, my employer needed a certain number of developers to be certified in order to improve the company's relationship with MSFT. I already had a BSCE and eight years of professional experience. I could tell right away that the tests were pitiful at evaluating knowledge or experience.
My experience with more recent MSCEs seem to indicate that nothing has changed. Some are good. Some are bad. The MSCE status indicates only that you had enough initiative and discipline to stop playing video games long enough to do something about your career.
The ubiquitousness of OWL/RDF on the web would, IMHO, be tied to two things; adoption by a large search engine such as google and the availability of a cool tool that would make the authoring process easy.
Imagine a mind mapping tool like freemind that could export to OWL. That would work for static content. Content publishers would do it for the same reason that they include keyword meta tags in their pages now.
Regarding OpenOffice vs MS-Office, the question isn't "is MS-Office better than OO?" The question is "is MS-Office better than OO by $400?"