Hmm LINQ is storage independant. How about reading a few before posting any nonsense.
That isn't the point. The point is that Java includes transparent persistence (persistence by reachability using class enhancement or proxying) which means that there is no need for extensive use of such query languages. Once you have accesses one or two key objects, the fields of those objects (and objects that they refer to are automatically fetched (either in an optimised fashion, or lazily) without the developer having to do anything. The state of these objects is monitored, so that any changed objects/fields are automatically made persistent. In spite of Microsoft's anti-Java FUD on this matter (they like to think, against all the evidence, that Object Databases are dead), this approach really works well and is used effectively by hundreds of thousands of developers. It is far more advanced that doing everything with an embedded query language (no matter how portable), as it means that the logic of storage and retrieval does not need to be mixed with any other parts of your code. It means that you can transparently apply persistence even to code that is already written.
Sorry, but the embedded query language idea, no matter how portable, is a 'quick fix' messy solution and leads to increased, not reduced, code complexity. This is typical of how Microsoft approach things.
Science hasn't yet begun to get into finding out anything about Love - I look forward to the time when we have more documented understanding about this
If/when scientists can prove the theory evolution, using scientific methods, with full explanation from creation to the world we see around us today, then I will stand up and listen. Until then, evolution is a theory which requires as much faith as religion requires.
This is not a sensible point of view. Eventually, with enough evidence, it is sensible to judge that an theory that takes you most of the way to a complete explanation is leading to the truth. To assume otherwise is just plain silly. After all, we have this reasonable attitude in other areas of science.
We are not yet fully able to explain the nature of matter in it's finest detail - but does this mean we assume that matter doesn't exist? Of course not!
We don't yet have a full understanding of turbulence or drag, which means we are always dealing with approximations when we do calculates of objects moving through the air. Do we therefore doubt that planes can fly? Obviously not!
So... We don't yet have the full details of how life originated and how every species evolved. Does this mean we assume that evolution is only a matter of faith?
Of course not!
(presumably as robots with no free will, as it is scientifically documented)
No scientist holds this view. This is a "straw man" argument.
As a Christian Pastor, I believe the world, and everything in it, was indeed created by God. But I also believe that he is a God of order, and thus there is an order to all things that can be observed and recorded. As science progresses, I would expect that many things that are a mystery to us today would be explained and understood. The fact that there is a scientific explanation for these things does not disprove the existence of God. Sure, for many in the creationist camp, science and God have no business mixing. But there are also those who believe as I do. Why do God and Science have to be mutually exclusive?
They are not mutually exclusive; the problem is almost as soon as you try and define what God is and what He/She does, and if you then proclaim these definitions as either part of your faith, or as dogma, then you are closing down areas for scientific investigation.
Let me take as an example something you have written - you have said you believe that God is a deity of order and thus there is an order to all things that can be observed and recorded. This means you already have a biased opinion that may prevent you from seeing what is really going on when you are presented with scientific result - you seem to have already decided the interpretation you are going to put on it. Science seems to be revealing that there is very little order in things - there is chaos and uncertainty at the base of everything. Quantum Theory seems to imply that the future is fundamentally unpredictable, and Evolutionary Theory seems to suggest that life evolves from random mutations (if you re-wound evolution, it would not 'play back' anything like the same way).
So, you see - even what seems like an innocuous and reasonable belief regarding the nature of God apparently puts you at odds with mainstream science.
Unless religious are prepared to be pragmatic, and accept that they must adapt their beliefs to cope with the findings of Science, there will always be conflict at some level.
One of the most exciting is LINQ which will change the way we interact with SQL databases, and data in general from within our code.
The idea of embedding query languages was tried and abandoned in Java years ago (remember SQL/J). Now Java uses truly portable query languages (such as JDOQL) and persistence mechanisms like JDO 2.0 that allow storage to far more than just relational stores. This uses something called 'transparent persistence' where objects and fields are retrieved and persisted automatically without the need for a query language most of the time.
Other language enhancements will mean that Java has a lot of catching up to do
Even Java 1.5 (sorry 5.0!)?
Looking at matters like object persistence, Microsoft and.NET have a lot of catching up to do...
I'm not going to dicker with you over the liberties you're taking with the term "binary".
Well, I am not shipping source code...
But I do thank you for calling Java a "platform", which it is (i.e. it is not "cross-platform", it is "cross-OS^H^Hseveral-OSes").
You are just playing with words here. If you are being that strict then C binaries are not cross-platform as you need libraries or Kernel services on the OS you are deploying on.
As for 'several OSes' - this is nonsense. You would find it hard to name a major OS now that does not have a Java VM.
You'd be widely viewed as a fool if you don't anyways. "Write once, debug everywhere" may not be as much the case nowadays (holy shit was it back in the JDK 1.0 and 1.1 days!), but only a fool doesn't test before deploying. (Or is that what your customers are for?;-)
Well, then I am a fool. Actually, I am not, because you really, honestly, do not need to test like this, especially if you are deploying on an app server. The JVM implementor and App server vendor has done the work for you - that is why these products are put through so many rigorous tests before they are allowed to be called 'Java' or 'J2EE'. There may be 'edge-case' issues if you are doing, say, very high-traffic networking or graphical game development, but otherwise, you really can safely assume portability.
Interestingly, with Java, instead of having the burden that you refer to here of conforming to your customers' platform(s), you code to one platform and your customers have the burden (and often the expense) of acquiring the (correct version of) the platform you've chosen, with the accompanying performance loss of one level of indirection/not running directly on the native platform. Definitely an advantage of Java, to the developer.
Sorry, but what burden? What cost? To run even a major J2EE application all the client needs is a JRE (free from Sun, or IBM, or HP), and an app server (free and open source from Sun, or Apache, or JBoss, or Geronimo).
Setting up most of these requires no more than a download, and unpack and setting a few environment variables (e.g. JAVA_HOME).
If you can find any cost or bother in this, go ahead and explain. I can't.
And, of course, the idea that there is any performance cost because of a level of indirection is further nonsense. All Java byte codes, be they in your application or the app server, are compiled to highly optimised native code. There is no more performance cost for Java to use these native services than for C to use, say a threading library or sockets library. The JVM implements its services using the native platform in the same way that POSIX-compatible libraries provide access to the native platform. Both are compatibility layers.
I'm sorry, but you are setting up an endless series of non-existent problems with Java to try and justify continued use of C++.
That is not a project but a description of an arguably possible project like those used by the framework developers to "demonstrate" the need for the framework.
No, it was an actual real-life project I have been involved with, not a demo.
Just because something is possible this does not mean it has to be done that way.
True, but it was the sensible way to do things. XML and XSLT made this easy.
All our projects have one thing in common - the code is compiled by the Java compiler. That is the common determinator and I want this to cover most if not all the code.
Fine, but then it is your choice to not use the tools that allow things like XML files to be checked by the IDE. This does not invalidate the point that these things need not be checked at run time.
You're right, I do assume that, so I guess there's something we agree on.
No - we both agree that you assume it, not that it is true:)
IT is mostly run by PHB's, who are largely unable to resist marketing, and definitely cannot make judgments for themselves based on years of experience with computer languages (because they have none).
I disagree. There is a phenomenal amount of IT that is based around small consultants and small developer groups. These are the people who are willing to try new technologies, and have to work hard to keep up-to-date with developments to keep themselves competitive. I am one of these.
We are the people have have followed Java over many years, waiting until it really is capable of taking over from C++, as we can really make use of the faster and safer development, and the ability to really provide platform-independent binaries. As an example, I often write web applications. I can supply my customer with a.WAR archive in the full confidence that they can deploy this binary on any J2EE 1.4 platform - HP/UX, Windows, Solaris, Linux, z/OS etc.
C++ can never match this - I would have to test my application on all those platforms to be confident. Even if I didn't, I would have to provide separate binaries and libraries - it is a mess!
Java is a real benefit to developers like me - it removes so much of the burden of development and portability testing that took up so much of a project.
Sorry, but you are simply wrong about the state of the IT market. We aren't all stupid!
For example, Dell's default "office" suite that they bundle is Corel. Dell bundles MS Office only if you explicitly select it and pay extra.
No. Maybe it is just because I am in the European market, but all I see from Dell is the option for Microsoft software - Works, Office 2003.
When I check for HP home PCs I get offered 'Microsoft Media Centre' PCs, with Microsoft Works, Microsoft AutoRoute, Microsoft Encarta, Microsoft..., Microsoft..., Microsoft.... (oh - iTunes sneaks in there somewhere at the end!). Talk about bundling!
I'm afraid you are contradicted by your own examples.
Microsoft isn't preventing OEMs from installing any 3rd party apps that they want to.
Also, many of these apps aren't bundled equivalently. For example, many MS apps are pre-installed, which gives them a head start over competitors.
The situation I was specifying was for Microsoft to allow competitors products to be packaged equivalently on the same PC. As it is, software like browsers and media players are by default pre-installed. Microsoft also has control over what appears on the desktop and menu items of pre-installed machines.
I was refering to ability, not specific programs... how hard is it to make a windows installer for a generic program work in 99%+ windows installs, compared to doing the same in linux?
From personal experience, it can be very hard, as you get conflicts with existing programs, and differing versions of system libraries. On Linux it is easy for different versions of libraries to co-exist without conflicts. This is not the case on Windows.
how many apps can you directly copy from redhat 6, to ubunto breazy 5.1x?
With the right libraries, plenty.
how many can you do that from win95 to xp?
All I can say is that I have not had much success doing that!
there's a *MUCH* larger amount of app support accrossed disparate versions of windows than linux
I had windows working on the machine in question, I was trying ubuntu to see if I could recover anything from the HD.. wireless didn't work, and I had to do other things, but saying that the installer does it all simply isn't true.
I didn't say it does it all; just that I found it did a lot more than XP.
And doing a post-install of drivers in *nix isn't as easy as download, click->next->next->reboot->done.
And it is often far more complex than this under Windows.
Don't get me wrong, I like linux, love PC-BSD, and really like my Mac... but to say that linux is easier for general use than windows isn't so...
And from lots of experience, I disagree.
Are there disks out there that will install on 99.999% of windows boxes (various versions, 9x or nt based) without issue? (answer: yes)...
answer: no. I have had to scrap many programs that worked fine on earlier versions of Windows that now no longer work under Win 2000 or XP.
Re:C++ has its place
on
Demise of C++?
·
· Score: 2, Interesting
"Python, Ruby, etc. - Often considered too slow."
Only in urban myths.
No, in practical use. Try doing something like image processing in those languages; or (perhaps more realistic) parsing a large XML file with native Python or Ruby code. Now try it in a C++ or Java parser. The difference in speed is phenomenal.
Microsoft don't distribute other people's software with their software, simply because they're a commercial entity and don't want to put themselves out of business. Combined with the competitiveness pressure, meaning they can't distribute much of their software with their OS, you get Windows as more or less a blank slate.
I think you are missing the point. It is not that Microsoft don't distribute other people's software - it is that they come to arrangements with PC vendors to prevent the vendor distributing other people's software, or even alternative versions of Windows. Here is an example:
You could just as easily download the applications and install them in that case. Either way, it's a blank slate when the OS finishes installing, as per the antitrust situation (which is what I prefer, anyway).
Have you tried downloading Open Office over a 56k modem?
You'd actually like to have three different office suites, three different media players, three different browsers, and three different email clients installed from the word "go"? Do you work for Dell or something?
Who says they have to be installed? They could be provided as install CDs. This is the way things were done years ago, before MS got too powerful.
Key statement: I set up Linux workstations
Let your novice users try to set up and use both and see which is more successful.
Answer: Linux. The Ubuntu install is far simpler than Windows, and was the first install of any OS on PC hardware I have ever seen that needed no prompting or additional drivers to deal with hardware.
Give a beginner a Windows XP CD and an Ubuntu CD and I have no doubts they would find Ubuntu easier to install in most cases.
I'd have to include Apache Tomcat - the open source Java application server that is the reference standard for JSP and Servlet implementations, but is of such high quality that it can be used for high-performance websites, and commercial app servers often bundle it as part of their product.
And what is Microsoft supposed to do about applications? If it bundled Microsoft Office in with Windows, the anti-competition people would be on their backs the day it hit the shelves. They have no choice but keep the OS relatively free of apps - too many partners they don't want to piss off and the anti-competition people just waiting with multi-million dollar fines! Look at the shit they are having to go through here in Europe with Windows Media Player for example!
They could do what they used to do years ago - allow the bundling of MS applications and alternatives on the same PC - perhaps as CDs. They you could chose MS Works or Corel Office or Open Office...
How about doing a review from the perspective of someone who has never used a computer before - then lets see which one is easier to use (hint: the answer will be Windows XP by a massive margin).
No. I set up Linux workstations in a company where here are often novice users. They have no difficulty using Linux. Why should they? A modern Linux desktop and office applications work in pretty much the same way as a Windows desktop - apart from the lack of regular virus warnings and the reduced ability to play games (in our company, we consider these to be good things).
The flaw in your thinking is that you forgot that "most popular" does not always equal "best". Much more often it equals "most marketed", as in the case of the two redundant (to C++) languages mentioned.
The flaw in your thinking is that you assume that everyone except a few wise C++ supporters are unable to resist marketing, and can't make judgements for themselves based on years of experience of comparing C++, Java and C#.
Generating XML and transforming it via XSLT on the server is plain stupid. I have never seen a project which required this extra flexibility and justified the huge overhead of using the ugly XSLT. Piping the XML to the client and doing the XSLT transformations there might be ok in cases where bandwith is of importance.
Here is an example, so you can say that you have 'seen' such a project. Sometimes it is useful to generate XML because it can be selectively transformed by XSLT to a variety of report formats as the user requires: XML+XSLT can output RTF, or it can output XSL:FO which can be rendered as postscript or PDF, which can be returned to the user. The same XML output can be transformed into these different formats simply by supplying a different XSL stylesheet. This is simple to use.
XSLT processing can be very efficient, and to say XSLT is 'ugly' is simply a personal aesthetic judgment, and not of any value in judging how useful or fast it is.
All these frameworks limit testing capabilities and move things away from the compiler to the runtime. EL expressions, JSP taglibs, Beans using Reflection, huge XML configuration, XSLT and so on are all evaluated at runtime and none of this can be checked by a compiler.
This is simply not true. A good IDE (like Eclipse) has plugins that can check the validity of most of these before runtime.
The most insightful post here yet. These compromise languages are, by design, almost never "the best tool for the job". They only exist for fodder in the pissing contest between Sun and MS, and the PHB's and fresh college grads caught up in it.
And as Java becomes not only the most popular development language for general purpose work, but also on Sourceforge, it is quite amusing to see yet another 'everyone is wrong but me' post. Because obviously almost all of the IT industry has got it wrong, and consists only of PHBs and fresh college grads.
Slashdot Mirror
I wouldn't use dice.com as a tool to measure Python usage.
Then what would you use?
Hmm LINQ is storage independant. How about reading a few before posting any nonsense.
That isn't the point. The point is that Java includes transparent persistence (persistence by reachability using class enhancement or proxying) which means that there is no need for extensive use of such query languages. Once you have accesses one or two key objects, the fields of those objects (and objects that they refer to are automatically fetched (either in an optimised fashion, or lazily) without the developer having to do anything. The state of these objects is monitored, so that any changed objects/fields are automatically made persistent. In spite of Microsoft's anti-Java FUD on this matter (they like to think, against all the evidence, that Object Databases are dead), this approach really works well and is used effectively by hundreds of thousands of developers. It is far more advanced that doing everything with an embedded query language (no matter how portable), as it means that the logic of storage and retrieval does not need to be mixed with any other parts of your code. It means that you can transparently apply persistence even to code that is already written.
Sorry, but the embedded query language idea, no matter how portable, is a 'quick fix' messy solution and leads to increased, not reduced, code complexity. This is typical of how Microsoft approach things.
Science hasn't yet begun to get into finding out anything about Love - I look forward to the time when we have more documented understanding about this
Wait no longer:
http://www.bbc.co.uk/science/hottopics/love/
If/when scientists can prove the theory evolution, using scientific methods, with full explanation from creation to the world we see around us today, then I will stand up and listen. Until then, evolution is a theory which requires as much faith as religion requires.
This is not a sensible point of view. Eventually, with enough evidence, it is sensible to judge that an theory that takes you most of the way to a complete explanation is leading to the truth. To assume otherwise is just plain silly. After all, we have this reasonable attitude in other areas of science.
We are not yet fully able to explain the nature of matter in it's finest detail - but does this mean we assume that matter doesn't exist? Of course not!
We don't yet have a full understanding of turbulence or drag, which means we are always dealing with approximations when we do calculates of objects moving through the air. Do we therefore doubt that planes can fly? Obviously not!
So...
We don't yet have the full details of how life originated and how every species evolved. Does this mean we assume that evolution is only a matter of faith?
Of course not!
(presumably as robots with no free will, as it is scientifically documented)
No scientist holds this view. This is a "straw man" argument.
As a Christian Pastor, I believe the world, and everything in it, was indeed created by God. But I also believe that he is a God of order, and thus there is an order to all things that can be observed and recorded. As science progresses, I would expect that many things that are a mystery to us today would be explained and understood. The fact that there is a scientific explanation for these things does not disprove the existence of God. Sure, for many in the creationist camp, science and God have no business mixing. But there are also those who believe as I do. Why do God and Science have to be mutually exclusive?
They are not mutually exclusive; the problem is almost as soon as you try and define what God is and what He/She does, and if you then proclaim these definitions as either part of your faith, or as dogma, then you are closing down areas for scientific investigation.
Let me take as an example something you have written - you have said you believe that God is a deity of order and thus there is an order to all things that can be observed and recorded. This means you already have a biased opinion that may prevent you from seeing what is really going on when you are presented with scientific result - you seem to have already decided the interpretation you are going to put on it. Science seems to be revealing that there is very little order in things - there is chaos and uncertainty at the base of everything. Quantum Theory seems to imply that the future is fundamentally unpredictable, and Evolutionary Theory seems to suggest that life evolves from random mutations (if you re-wound evolution, it would not 'play back' anything like the same way).
So, you see - even what seems like an innocuous and reasonable belief regarding the nature of God apparently puts you at odds with mainstream science.
Unless religious are prepared to be pragmatic, and accept that they must adapt their beliefs to cope with the findings of Science, there will always be conflict at some level.
One of the most exciting is LINQ which will change the way we interact with SQL databases, and data in general from within our code.
.NET have a lot of catching up to do...
The idea of embedding query languages was tried and abandoned in Java years ago (remember SQL/J). Now Java uses truly portable query languages (such as JDOQL) and persistence mechanisms like JDO 2.0 that allow storage to far more than just relational stores. This uses something called 'transparent persistence' where objects and fields are retrieved and persisted automatically without the need for a query language most of the time.
Other language enhancements will mean that Java has a lot of catching up to do
Even Java 1.5 (sorry 5.0!)?
Looking at matters like object persistence, Microsoft and
I'm not going to dicker with you over the liberties you're taking with the term "binary".
;-)
Well, I am not shipping source code...
But I do thank you for calling Java a "platform", which it is (i.e. it is not "cross-platform", it is "cross-OS^H^Hseveral-OSes").
You are just playing with words here. If you are being that strict then C binaries are not cross-platform as you need libraries or Kernel services on the OS you are deploying on.
As for 'several OSes' - this is nonsense. You would find it hard to name a major OS now that does not have a Java VM.
You'd be widely viewed as a fool if you don't anyways. "Write once, debug everywhere" may not be as much the case nowadays (holy shit was it back in the JDK 1.0 and 1.1 days!), but only a fool doesn't test before deploying. (Or is that what your customers are for?
Well, then I am a fool. Actually, I am not, because you really, honestly, do not need to test like this, especially if you are deploying on an app server. The JVM implementor and App server vendor has done the work for you - that is why these products are put through so many rigorous tests before they are allowed to be called 'Java' or 'J2EE'. There may be 'edge-case' issues if you are doing, say, very high-traffic networking or graphical game development, but otherwise, you really can safely assume portability.
Interestingly, with Java, instead of having the burden that you refer to here of conforming to your customers' platform(s), you code to one platform and your customers have the burden (and often the expense) of acquiring the (correct version of) the platform you've chosen, with the accompanying performance loss of one level of indirection/not running directly on the native platform. Definitely an advantage of Java, to the developer.
Sorry, but what burden? What cost? To run even a major J2EE application all the client needs is a JRE (free from Sun, or IBM, or HP), and an app server (free and open source from Sun, or Apache, or JBoss, or Geronimo).
Setting up most of these requires no more than a download, and unpack and setting a few environment variables (e.g. JAVA_HOME).
If you can find any cost or bother in this, go ahead and explain. I can't.
And, of course, the idea that there is any performance cost because of a level of indirection is further nonsense. All Java byte codes, be they in your application or the app server, are compiled to highly optimised native code. There is no more performance cost for Java to use these native services than for C to use, say a threading library or sockets library. The JVM implements its services using the native platform in the same way that POSIX-compatible libraries provide access to the native platform. Both are compatibility layers.
I'm sorry, but you are setting up an endless series of non-existent problems with Java to try and justify continued use of C++.
That is not a project but a description of an arguably possible project like those used by the framework developers to "demonstrate" the need for the framework.
No, it was an actual real-life project I have been involved with, not a demo.
Just because something is possible this does not mean it has to be done that way.
True, but it was the sensible way to do things. XML and XSLT made this easy.
All our projects have one thing in common - the code is compiled by the Java compiler. That is the common determinator and I want this to cover most if not all the code.
Fine, but then it is your choice to not use the tools that allow things like XML files to be checked by the IDE. This does not invalidate the point that these things need not be checked at run time.
You're right, I do assume that, so I guess there's something we agree on.
:)
.WAR archive in the full confidence that they can deploy this binary on any J2EE 1.4 platform - HP/UX, Windows, Solaris, Linux, z/OS etc.
No - we both agree that you assume it, not that it is true
IT is mostly run by PHB's, who are largely unable to resist marketing, and definitely cannot make judgments for themselves based on years of experience with computer languages (because they have none).
I disagree. There is a phenomenal amount of IT that is based around small consultants and small developer groups. These are the people who are willing to try new technologies, and have to work hard to keep up-to-date with developments to keep themselves competitive. I am one of these.
We are the people have have followed Java over many years, waiting until it really is capable of taking over from C++, as we can really make use of the faster and safer development, and the ability to really provide platform-independent binaries. As an example, I often write web applications. I can supply my customer with a
C++ can never match this - I would have to test my application on all those platforms to be confident. Even if I didn't, I would have to provide separate binaries and libraries - it is a mess!
Java is a real benefit to developers like me - it removes so much of the burden of development and portability testing that took up so much of a project.
Sorry, but you are simply wrong about the state of the IT market. We aren't all stupid!
For example, Dell's default "office" suite that they bundle is Corel. Dell bundles MS Office only if you explicitly select it and pay extra.
No. Maybe it is just because I am in the European market, but all I see from Dell is the option for Microsoft software - Works, Office 2003.
When I check for HP home PCs I get offered 'Microsoft Media Centre' PCs, with Microsoft Works, Microsoft AutoRoute, Microsoft Encarta, Microsoft..., Microsoft..., Microsoft.... (oh - iTunes sneaks in there somewhere at the end!). Talk about bundling!
I'm afraid you are contradicted by your own examples.
Microsoft isn't preventing OEMs from installing any 3rd party apps that they want to.
Also, many of these apps aren't bundled equivalently. For example, many MS apps are pre-installed, which gives them a head start over competitors.
The situation I was specifying was for Microsoft to allow competitors products to be packaged equivalently on the same PC. As it is, software like browsers and media players are by default pre-installed. Microsoft also has control over what appears on the desktop and menu items of pre-installed machines.
The problem with bundling the alternatives is that MS would be obligated to support them.
It is not Microsoft who would bundle them; it is Microsoft who would change their OEM licenses to allow the PC seller to bundle them.
I was refering to ability, not specific programs... how hard is it to make a windows installer for a generic program work in 99%+ windows installs, compared to doing the same in linux?
From personal experience, it can be very hard, as you get conflicts with existing programs, and differing versions of system libraries. On Linux it is easy for different versions of libraries to co-exist without conflicts. This is not the case on Windows.
how many apps can you directly copy from redhat 6, to ubunto breazy 5.1x?
With the right libraries, plenty.
how many can you do that from win95 to xp?
All I can say is that I have not had much success doing that!
there's a *MUCH* larger amount of app support accrossed disparate versions of windows than linux
No, not true.
I had windows working on the machine in question, I was trying ubuntu to see if I could recover anything from the HD.. wireless didn't work, and I had to do other things, but saying that the installer does it all simply isn't true.
I didn't say it does it all; just that I found it did a lot more than XP.
And doing a post-install of drivers in *nix isn't as easy as download, click->next->next->reboot->done.
And it is often far more complex than this under Windows.
Don't get me wrong, I like linux, love PC-BSD, and really like my Mac... but to say that linux is easier for general use than windows isn't so...
And from lots of experience, I disagree.
Are there disks out there that will install on 99.999% of windows boxes (various versions, 9x or nt based) without issue? (answer: yes)...
answer: no. I have had to scrap many programs that worked fine on earlier versions of Windows that now no longer work under Win 2000 or XP.
"Python, Ruby, etc. - Often considered too slow."
Only in urban myths.
No, in practical use. Try doing something like image processing in those languages; or (perhaps more realistic) parsing a large XML file with native Python or Ruby code. Now try it in a C++ or Java parser. The difference in speed is phenomenal.
You cite an article from 2002 as current? Where's the mod points for "fucking idiotic"?
If you actually read the article you will see that it is current. The OEM license change was initiated in 2002, but it is still active.
Microsoft don't distribute other people's software with their software, simply because they're a commercial entity and don't want to put themselves out of business. Combined with the competitiveness pressure, meaning they can't distribute much of their software with their OS, you get Windows as more or less a blank slate.
, 00.asp
I think you are missing the point. It is not that Microsoft don't distribute other people's software - it is that they come to arrangements with PC vendors to prevent the vendor distributing other people's software, or even alternative versions of Windows. Here is an example:
http://www.extremetech.com/article2/0,3973,760198
It is about Microsoft controlling what the PC vendor can do.
You could just as easily download the applications and install them in that case. Either way, it's a blank slate when the OS finishes installing, as per the antitrust situation (which is what I prefer, anyway).
Have you tried downloading Open Office over a 56k modem?
You'd actually like to have three different office suites, three different media players, three different browsers, and three different email clients installed from the word "go"? Do you work for Dell or something?
Who says they have to be installed? They could be provided as install CDs. This is the way things were done years ago, before MS got too powerful.
IT would be an utter stupidity to do that. You are saying that along with what I have and own, I bundle my neighbor's property too when I sell mine!
Perhaps you should look up the definition of 'monopoly'.
Key statement: I set up Linux workstations
Let your novice users try to set up and use both and see which is more successful.
Answer: Linux. The Ubuntu install is far simpler than Windows, and was the first install of any OS on PC hardware I have ever seen that needed no prompting or additional drivers to deal with hardware.
Give a beginner a Windows XP CD and an Ubuntu CD and I have no doubts they would find Ubuntu easier to install in most cases.
I'd have to include Apache Tomcat - the open source Java application server that is the reference standard for JSP and Servlet implementations, but is of such high quality that it can be used for high-performance websites, and commercial app servers often bundle it as part of their product.
And what is Microsoft supposed to do about applications? If it bundled Microsoft Office in with Windows, the anti-competition people would be on their backs the day it hit the shelves. They have no choice but keep the OS relatively free of apps - too many partners they don't want to piss off and the anti-competition people just waiting with multi-million dollar fines! Look at the shit they are having to go through here in Europe with Windows Media Player for example!
They could do what they used to do years ago - allow the bundling of MS applications and alternatives on the same PC - perhaps as CDs. They you could chose MS Works or Corel Office or Open Office...
How about doing a review from the perspective of someone who has never used a computer before - then lets see which one is easier to use (hint: the answer will be Windows XP by a massive margin).
No. I set up Linux workstations in a company where here are often novice users. They have no difficulty using Linux. Why should they? A modern Linux desktop and office applications work in pretty much the same way as a Windows desktop - apart from the lack of regular virus warnings and the reduced ability to play games (in our company, we consider these to be good things).
The flaw in your thinking is that you forgot that "most popular" does not always equal "best". Much more often it equals "most marketed", as in the case of the two redundant (to C++) languages mentioned.
The flaw in your thinking is that you assume that everyone except a few wise C++ supporters are unable to resist marketing, and can't make judgements for themselves based on years of experience of comparing C++, Java and C#.
Generating XML and transforming it via XSLT on the server is plain stupid. I have never seen a project which required this extra flexibility and justified the huge overhead of using the ugly XSLT. Piping the XML to the client and doing the XSLT transformations there might be ok in cases where bandwith is of importance.
Here is an example, so you can say that you have 'seen' such a project. Sometimes it is useful to generate XML because it can be selectively transformed by XSLT to a variety of report formats as the user requires: XML+XSLT can output RTF, or it can output XSL:FO which can be rendered as postscript or PDF, which can be returned to the user. The same XML output can be transformed into these different formats simply by supplying a different XSL stylesheet. This is simple to use.
XSLT processing can be very efficient, and to say XSLT is 'ugly' is simply a personal aesthetic judgment, and not of any value in judging how useful or fast it is.
All these frameworks limit testing capabilities and move things away from the compiler to the runtime. EL expressions, JSP taglibs, Beans using Reflection, huge XML configuration, XSLT and so on are all evaluated at runtime and none of this can be checked by a compiler.
This is simply not true. A good IDE (like Eclipse) has plugins that can check the validity of most of these before runtime.
The most insightful post here yet. These compromise languages are, by design, almost never "the best tool for the job". They only exist for fodder in the pissing contest between Sun and MS, and the PHB's and fresh college grads caught up in it.
And as Java becomes not only the most popular development language for general purpose work, but also on Sourceforge, it is quite amusing to see yet another 'everyone is wrong but me' post. Because obviously almost all of the IT industry has got it wrong, and consists only of PHBs and fresh college grads.