Re:More than two options...
on
JSF vs ASP.net
·
· Score: 1
Clustered transactions aren't always a requirement (in fact, they're *rarely* a requirement), and certainly weren't listed as such by the article author.
As for maintainability... good luck getting maintainability with *any* language when you have dozens of coders on a project. PHP is not inherently unmaintainable, it's a language, capable of some pretty sweet things. So are Java, C#, ruby, and python, just to name a few (note: I avoided perl, because *shudder*;)). Whether or not a program is maintainable is dependent upon how the designers design, implement, and document the code, not the language itself.
Re:I haven't worked with both, but....
on
JSF vs ASP.net
·
· Score: 1
I'll give you the better refactoring on Eclipse -- VS only just got a very limited version in 2005.... *finally*, but Eclipse's is definitely much better. The debugger is debatable, I think they're about on par, although VS has only just (finally!) gotten edit and continue. As for the rest, I think it's more about personal taste -- I prefer VS's project management and configuration management to Eclipse's.
However, there *is* one thing that Eclipse has over VS... totally free. You can't beat that very often!
Although MS did for me just last month -- I got a free copy of VS 2005 Standard Edition when I attended one of their VS 2005 launch events. Sure, it's not Professional, and I wouldn't mind Team Suite with the automated testing and architectural tools, but it's pretty damn good for free!
I've enjoyed developing PHP applications, so you might also want to consider some form of LAMP-oriented system. If you do, consider using Zend Platform, the new PHP server management system from Zend; it's supposed to make administration of PHP webservers much easier.
That said, I have a radical suggestion: use a rich client rather than web application. Even with AJAX, the web is still a very limited user experience platform -- I've always preferred manly GUIs to sissified pansy-web-pages.;) If you go that route, consider using either Java WebStart (if you prefer Java) or Microsoft's ClickOnce (if you prefer.NET).
If you prefer the.NET route, which I would personally recommend over Java, you might also consider developing the application in the new Windows Presentation Foundation. Although it's still in BETA right now, WPF applications deployed with ClickOnce have the option of running inside the browser -- the user gets a much richer experience than web applications can grant even with AJAX, but are still on the web. Think of the browser option as a cross between a Java applet (which.NET has been sorely lacking) and Mozilla's XUL.
Of course, there's a trade-off... if you go the rich-client route and use.NET, you cut off non-Windows platforms. If you run WPF, you are cutting edge (it's not even finished yet!!), but you have a small client-base until Vista takes off.
Any given method should have more than one test case; any given class could have dozens. Are you honestly suggesting that it's better to read and analyze all of that test code to determine what a method is and how it works -- i.e. the general context surrounding it -- rather than a simple paragraph or two in english? Do you like to chew tinfoil? Do you like to roll around on broken glass?
As for drift, you forget that refactoring is a part of test-driven development; test case code drifts just as much as API documentation can.
Good context for modular and reusable code include things like the the purpose for class. Is the class a custom GUI control, a data object, a control/worker object? What does the data represent? What does the worker object do? What does your control allow users to do?
Of course, that's modular-level, but the same must be done on the method level and lower. What does this method do? Does it animate a portion of the control in some way? Does it maniplate the data to a new state? What do the arguments represent? Is x in screen coordinates, container/control coordinates, 3d coordinates?
None of this information can be readily obtained from source code or test cases without careful analysis, and even then a deep understanding of the underlying APIs is necessary, which of course requires, you guessed it, the same context.
Doxygen is nice because it standardizes a particular commentation style over multiple languages, so that whatever you use for a project (or within a project), you comment in roughly the same way, using the same commands, etc.
Personally, however, I very much prefer the xml-oriented way of doing it found in Javadoc and.NET's comment/documentation scheme.
Even if you're the best normalizer/refactorer (is "refactorer" a word?) in the world, you had better explain what a function or method or class or aspect or (etc..) does, how it's used, and what the expected results are, or you have made code that is a nightmare to maintain.
In truth, function-point comments of a well-normalized/refactored program represent easily 99% of the necessary commentation for a program. The only time you ever really need comments otherwise is to explain a particularly tricky algorithm that, say for performance reasons, must resist typical normalization/refactoring. Even then, you can easily place an overview of the algorithm in the documentation (or reference a whitepaper), and again remove most of your inline comments.
As for function-point comments/documentation, I suggest something like Javadoc, C#'s Intellidoc, or some similar platform like doxygen, although if you're manly or insane enough, you can always maintain the documentation separate from the code.
Damn good trick, then. 'Cause C# and.NET are heads and shoulders better than Java, speaking as someone who programs regularly in both languages.
At the moment, the only advantage to Java is Swing's cross-platform compatibility, but that is, as has already been mentioned, easily overcome with Mono and cross-platform GUI libraries.
Only better: full application as well as intra-browser (XUL is intra-browser only), plus a *much* better rendering engine, just to name a few. Plus, Microsoft is throwing its full support behind it as a key technology and will constantly upgrade and tweak the system, whereas Mozilla has quit bothering with XUL entirely a while ago.
As another poster mentioned, WPF is the Windows Presentation Framework, formerly Avalon.
It's the newest GUI platform from Microsoft, shipping with Vista, and it's getting tons of hype. Well deserved hype, IMO, as it's a very powerful framework. I've only just begun to dabble with the framework beta, but what I've seen and been able to do just so far has been amazing.
Because the next big thing, which you could learn while it's still cutting edge technology and known by only a select few, is Windows Vista and the Windows Presentation Framework. Powerful GUI it is... Powerful GUI
It's not really that wierd. Vertical bars are a popular list separation item, and I've used them in cookies for web applications I've designed many times. In your example, you have a twelve-item list, with the first two items equal to "CA" and "NA" respectively, and the remaining equal to "".
What they're doing with the list is anybody's guess.:)
Slashdot Mirror
Clustered transactions aren't always a requirement (in fact, they're *rarely* a requirement), and certainly weren't listed as such by the article author.
;)). Whether or not a program is maintainable is dependent upon how the designers design, implement, and document the code, not the language itself.
As for maintainability... good luck getting maintainability with *any* language when you have dozens of coders on a project. PHP is not inherently unmaintainable, it's a language, capable of some pretty sweet things. So are Java, C#, ruby, and python, just to name a few (note: I avoided perl, because *shudder*
I'll give you the better refactoring on Eclipse -- VS only just got a very limited version in 2005.... *finally*, but Eclipse's is definitely much better. The debugger is debatable, I think they're about on par, although VS has only just (finally!) gotten edit and continue. As for the rest, I think it's more about personal taste -- I prefer VS's project management and configuration management to Eclipse's.
However, there *is* one thing that Eclipse has over VS... totally free. You can't beat that very often!
Although MS did for me just last month -- I got a free copy of VS 2005 Standard Edition when I attended one of their VS 2005 launch events. Sure, it's not Professional, and I wouldn't mind Team Suite with the automated testing and architectural tools, but it's pretty damn good for free!
I've enjoyed developing PHP applications, so you might also want to consider some form of LAMP-oriented system. If you do, consider using Zend Platform, the new PHP server management system from Zend; it's supposed to make administration of PHP webservers much easier.
;) If you go that route, consider using either Java WebStart (if you prefer Java) or Microsoft's ClickOnce (if you prefer .NET).
.NET route, which I would personally recommend over Java, you might also consider developing the application in the new Windows Presentation Foundation. Although it's still in BETA right now, WPF applications deployed with ClickOnce have the option of running inside the browser -- the user gets a much richer experience than web applications can grant even with AJAX, but are still on the web. Think of the browser option as a cross between a Java applet (which .NET has been sorely lacking) and Mozilla's XUL.
.NET, you cut off non-Windows platforms. If you run WPF, you are cutting edge (it's not even finished yet!!), but you have a small client-base until Vista takes off.
That said, I have a radical suggestion: use a rich client rather than web application. Even with AJAX, the web is still a very limited user experience platform -- I've always preferred manly GUIs to sissified pansy-web-pages.
If you prefer the
Of course, there's a trade-off... if you go the rich-client route and use
After setting Yoda down, look looks around apprehensively.
LUKE: I feel cold.... death......
YODA: That file is strong with the dark side. A domain of perl it is. In you must go.
Perl vs. Python flame war commencing in 5...... 4....... 3....... 2....... 1......
Any given method should have more than one test case; any given class could have dozens. Are you honestly suggesting that it's better to read and analyze all of that test code to determine what a method is and how it works -- i.e. the general context surrounding it -- rather than a simple paragraph or two in english? Do you like to chew tinfoil? Do you like to roll around on broken glass?
As for drift, you forget that refactoring is a part of test-driven development; test case code drifts just as much as API documentation can.
Good context for modular and reusable code include things like the the purpose for class. Is the class a custom GUI control, a data object, a control/worker object? What does the data represent? What does the worker object do? What does your control allow users to do?
Of course, that's modular-level, but the same must be done on the method level and lower. What does this method do? Does it animate a portion of the control in some way? Does it maniplate the data to a new state? What do the arguments represent? Is x in screen coordinates, container/control coordinates, 3d coordinates?
None of this information can be readily obtained from source code or test cases without careful analysis, and even then a deep understanding of the underlying APIs is necessary, which of course requires, you guessed it, the same context.
Long live c++!
you are too sharp-witted for me. I will now pound sand.
I could go for some hashbrowns.
Automated tests do not grant context. Only non-structured narrative documentation can fully explain the purpose and use of a block of code.
Doxygen is nice because it standardizes a particular commentation style over multiple languages, so that whatever you use for a project (or within a project), you comment in roughly the same way, using the same commands, etc.
.NET's comment/documentation scheme.
Personally, however, I very much prefer the xml-oriented way of doing it found in Javadoc and
Even if you're the best normalizer/refactorer (is "refactorer" a word?) in the world, you had better explain what a function or method or class or aspect or (etc..) does, how it's used, and what the expected results are, or you have made code that is a nightmare to maintain.
In truth, function-point comments of a well-normalized/refactored program represent easily 99% of the necessary commentation for a program. The only time you ever really need comments otherwise is to explain a particularly tricky algorithm that, say for performance reasons, must resist typical normalization/refactoring. Even then, you can easily place an overview of the algorithm in the documentation (or reference a whitepaper), and again remove most of your inline comments.
As for function-point comments/documentation, I suggest something like Javadoc, C#'s Intellidoc, or some similar platform like doxygen, although if you're manly or insane enough, you can always maintain the documentation separate from the code.
Damn good trick, then. 'Cause C# and .NET are heads and shoulders better than Java, speaking as someone who programs regularly in both languages.
At the moment, the only advantage to Java is Swing's cross-platform compatibility, but that is, as has already been mentioned, easily overcome with Mono and cross-platform GUI libraries.
Yes.
Only better: full application as well as intra-browser (XUL is intra-browser only), plus a *much* better rendering engine, just to name a few. Plus, Microsoft is throwing its full support behind it as a key technology and will constantly upgrade and tweak the system, whereas Mozilla has quit bothering with XUL entirely a while ago.
As another poster mentioned, WPF is the Windows Presentation Framework, formerly Avalon.
It's the newest GUI platform from Microsoft, shipping with Vista, and it's getting tons of hype. Well deserved hype, IMO, as it's a very powerful framework. I've only just begun to dabble with the framework beta, but what I've seen and been able to do just so far has been amazing.
When was Java ever a dynamic web server technology?
JSP != Java.
That's like saying Ruby on Rails == Ruby.
No, it's moved to .NET and WPF.
*ducks*
Because the next big thing, which you could learn while it's still cutting edge technology and known by only a select few, is Windows Vista and the Windows Presentation Framework. Powerful GUI it is... Powerful GUI
XUL is tied too closely to HTML/CSS. Go XAML instead! :D
Here's an example of a web service that as a /. geek you've probably heard about:
.NET per se, but it's definitely a web service.
Google maps API.
Sure, it doesn't use
And let's not forget the Google Ad Words API. And others, like specialized search services.
XML is neither a means of storing data, nor a means of communicating data, but is only a means of *marking up* data.
I've searched the dell site, and found nothing. Care to back that up with a link?
I'd call that a typo. His final math does come out to megabytes per second (MB/s) but he typoed it to Mbs.
someone will have a binary stripper
Will he get a binary lap dance?
It's not really that wierd. Vertical bars are a popular list separation item, and I've used them in cookies for web applications I've designed many times. In your example, you have a twelve-item list, with the first two items equal to "CA" and "NA" respectively, and the remaining equal to "".
:)
What they're doing with the list is anybody's guess.