> Simplicity and a clean syntax is what makes Java a > success, turning it into a mix of C++, Ruby, Python, and > Visual Basic (which it exactly what's happening with all > the nonsense being "proposed") destroys that.
What simplicity??
The simplicity and clean syntax in terms of primitve operations, like 5 * 4 * 3 * 2 * 1 ? Or the simplicity and clean syntax in terms of expression, like 5! ?
If mathematicians had thought your way, all calculation would be done using + as only operation instead of creating a highly expressable symbol language with division bar, square root, exponent, sum, integral, a.s.o.
And in many places where productivity of Java is mentioned the answer is always "but you have so good IDEs! you don't have to write so much code by hand ..." .
Also regarding that code is more often read than written, a concise construct whose semantic can be easily grokked is always better than a complex boilerplate pattern recognition process necessary to get the simple ideas behind it.
And when trying to train a newby programming, I don't detect any simplicity in trying to explain why all the brain killing stuff around File I/O and handling must be done this complicate way.
> We don't need function pointers (a.k.a. "closures" to use > the policitally correct euphemism),
Calling closures "function pointers" seems slightly ignorant to me.
Not to mention this "we don't need" ... who are "we"? Seems to be not an including "we", but an excluding one, at least for guys like Sean Landis and me.
> "properties", built-in > database and web servers, and all the other crud that's > being added and has been sine 1.6 (and to a degree since > 1.4 even, when some light at Sun decided that a built-in > XML library would be a good thing despite the entire world > already using another one).
Dumping it now all (syntax, language features, libraries, tools, the JVM itsself) into one bucket does not develop any progress in the current topic, I think.
BTW: @jean-pierre milin: > closures are just a disguised way to get rid of object > orient programming and coming back to pointers
Aren't all references stored in variables "pointers" in the end? So where is the dichotomy between object-oriented and pointers?
And beside that: object orientation is a paradigm, not a syntax based around the keywords 'class' and 'new' . Even a 'closure' is an object to me!
(Would like to adapt a quote -I think- Georg Christoph Lichtenberg once made about chemistry: Who only understands Java, doesn't even understand that)
Flat View: This topic has 15 replies
on 2 pages
[
«
|
12
]
Previous Topic
