Content deleted Content added
Line 121:
:If you want a POV tag in these sections ''then you must state what the problem is with those sections''. I recognize that there are some who question the "weight" of the entire article. That is still an unresolved dispute and I encourage you to join the discussion below. But these sections are already covered by the article POV tage and you have not stated anything POV specific to these sections, except for misunderstandings about how the C# feature works. [[User:Useerup|Useerup]] ([[User talk:Useerup|talk]]) 11:38, 23 October 2010 (UTC)
:Response: You are just dismissing the arguments,and you follow them up with logical fallacies. Finally this being said , there many comments in saying this article is POV bias, but someone how you alone are dismissing them. The overall article has POV bias , I bringing up certian points and you pretty much dismissing all of them, without any real logic/or reasoning behind you statements. You haven't quoted any articles nor have you linked to any independent articles outside of wikipedia to support your conclusions. I will not put the tags in the article because you are correct it is the wrong tag, but I do not find it constructive if the length of the article is very long, you miss the POV tag at the beginning.
Why is this article a comparison of features when there have been several points cited that the architectures were tackled completely differently? The two languages are superficially similar, both in syntax and in architecture, and their conceptualizations are identical(Pure OO with single base class/multi interface inheritance, ad-hoc polymorphism, Garbage Collection and compilation to an intermediate byte code that is constructed to a VM that jits the byte code to native code, decoupling implementation and design, allowing for code-once debug anywhere methedologies). The identical structure is nice since it gives you the ability to code an abstract model and then implement the model in either language, but doing anything particularly advanced in the language is highly-non trivial and I as a C# dev would be lost in the advanced idioms in Java, and if a Java programmer can just read a complicated Linq query, then they are by definition as much a C# dev as me. To summarize, I feel we need validation of why a comparison is reasonable, and then we need a validation of why comparison of features is reasonable.
Side note on the table, that table read very biased to me as a C# dev in favor of C#. In particular it seemed that typed function pointers(delegates in C#), one of the most advanced features of C#, was really milked for all it's worth. Java has an idiom to emulate delegates, but it neither mechanically nor semantically is quite the same(in particular delegates in C# are not hugely idiomatic, the idiom in Java is so idiomatic I actually know it, and I don't code in Java). The addition of delegates opens up completely new design patterns(functional ones, usually), and as a result is a huge deal that can be milked. But that table seems to represent a half dozen features that are all really just "C# has function pointers". A number of the rest of the features could be summarized as "C# has a true top type for all types in safe code", which Java doesn't(and I am not saying that is a bad thing, but from a type-theoretic standpoint it is true. The safe code is a caveat that has to be added since the formal semantics of certain things pointers do, afaik, is not defined which is why they are made to be completely outside the type system, pointers should from a type theoretic view should not be assignment compatible to object so they aren't. A better rephrasing might be "all items that should, from a type theoretic stand point, be assignment compatible to object, are." Although this would imply that there aren't any flaws with the type system, which isn't true.) This also summarizes the issue with struct types, whether or not pass by value types in C# are or are not in the realm of OO design I don't care to get into, but they are assignment compatible object in C#, and they are not in Java(although the new auto-promotion feature will give them some syntatic sugar to help with that.) Further on this, struct types are in the CLR, and do NOT get auto-boxed, they work, under the hood, quite differently(dig through the IL if you don't believe me. I thought this was how it worked as well, but it isn't(it would lead to bizarre performance characteristics if it did.)
[[User:LCpdx|LCpdx]] ([[User talk:LCpdx|talk]]) 22:15, 11 December 2010 (UTC)
== Value types ==
| |||