Re: Java and .NET (no Flames Pls)
From: Mike Newton (MNewton_at_discussions.microsoft.com)
Date: 09/17/04
- Next message: Girish Bharadwaj: "Re: ?RE: Could events ever ?overrun? ?or otherwise be missed by a hand?ler?"
- Previous message: Drebin: "Re: How to retrieve user's domain correctly?"
- In reply to: Joe Lentrichia: "Re: Java and .NET (no Flames Pls)"
- Next in thread: Joe Lentrichia: "Re: Java and .NET (no Flames Pls)"
- Reply: Joe Lentrichia: "Re: Java and .NET (no Flames Pls)"
- Messages sorted by: [ date ] [ thread ]
Date: Fri, 17 Sep 2004 08:08:53 -0500
It's not a matter of sides. You can be proficient in C#, but also
understand that it isn't always the best tool for the job. You can also
be good at C#, and be a bigger fan of java.
I think this bit here is the biggest poison that infects MS loonies:
"I think the biggest obstacle for .NET and MS as a whole to overcome is
the ingrained concept in some developers mind that all MS products suck
(not true), and that MS has lots of security issues (somewhat true), and
most of all the general hatred because they have a monopoly on the
desktop/office market. They also get a bad rap for being "proprietary"
when in actuality they are pushing the envelope and just not waiting for
a panel of "experts" to "standardize" the API they happen to be working
on. And for this reason Visual Studio has superior support for Web
Services."
Open standards are what drive technology. Take TCP/IP for example, or
HTTP, HTML, CSS, PNG, etc. These are things that have a set format or
API, and are *not going to change.* This basically means that we can
stop worrying about how such and such is supposed to be implemented at
this point in time, and start worrying about making the existing
implementation more efficient, or start worrying about more important
things.
MS is not pushing the envelope by any means when it comes to open
standards. They've submitted the C# language and got it ISO certified,
but now C# 2.0 is out. They didn't standardize the framework libraries,
many of which contain patented material (I'm not going to even approach
the validity of software patents). When you patent something, you're
trying to make sure that you have a legal monopoly on selling that item
for a set amount of time. It is a version of "proprietary."
The panel of "experts" that you refer to is what you rely on every day.
If it weren't for these "experts," then you'd have a number of
different servers that speak in different languages, and may or may not
understand each other. You'd be able to talk to IIS, for instance, but
maybe not access Hotmail (which, I believe, is still based on BSD).
Then again, these "experts" were around when the industry started, or
they built the infrastructure for the industry themselves. Let's just
call them "experts," because they don't deserve any more respect than
that. </sarcasm>
Visual Studio is a good tool, no doubt about it. Does it, however, work
in the same way with mono? Apache? The problem of creating an end-all
tool like Visual Studio, as easy to use as it is, is that you've got
platform lock. Someone else is making the decision for you, as to what
web server you're going to use, or what operating system you are going
to run. That is, if you really really want to use visual studio. For
those who don't want to be confined to an extremely small set of
platforms, they choose to go with a tool that promises to work with many.
Eclipse, for instance, is open sourced. It has a board-standardized
plugin API, and a number of plugins exist for it. Eclipse, in effect,
has much more flexibility and extensibility than Visual Studio does. It
doesn't have an embedded WYSIWYG HTML editor, nor does it seamlessly
integrate with web servers by default. It does have a java debugger,
though. Everything else is handled by plugins, and I'm certain that
there are HTML plugins, web service plugins, visual form editing tools,
UML, etc.
One last thing, just an FYI, to tell the truth. It isn't the web server
that has debugging built in. When you're debugging a web service in
.NET, or ASP.NET, you're interacting with the .NET runtime, not the web
server. When the web server sees that you're requesting a .NET type
page, it passes the request off to the runtime, which processes the
request itself. In this respect, yes, you can debug java web servlets
and services, since you can debug through the virtual machine.
Sorry... this is a bit of a rant, but astroturfing without knowing both
sides of the equation is rather frustrating.
Joe Lentrichia wrote:
> Jon, it seems you are much in favor of java. Why then, I ask, are you a C#
> MVP?
>
> I agree with all your statements. Except the last one where a consumer of a
> .NET webservice might not be able to. I haven't seen this, unless perhaps
> you're talking about returning a DataSet, in which case I have to argue that
> the benefits outweigh the consumption downfalls. Even a DataSet can be
> consumed tho I believe.
>
> I will download Eclipse and give it a try next time I'm doing some Java code
> (likely next month) just on your recommendations alone. Some things I will
> look for: Does it contain a WYSIWYG HTML editor? Does it have an integrated
> WS proxy generator? Interactive debugging, with a built in web server? Good
> debugging features: watches, etc.
>
> One other thing, and this is strictly a preference nothing to do really with
> support (although you could argue perhaps some sense of abstraction) is
> Properties of a class instead of getters and setters. I was having a funny
> conversation with a Java programmer and he mistakenly thought that Properties
> were a violation of encapsulation, until I told him that they weren't
> internal data being accessed directly. (eg: Java: getThisProp{}
> setThisProp{}, C#: ThisProp{ get: set: };
>
>
> "Jon Skeet [C# MVP]" wrote:
>
>
>> <"=?Utf-8?B?Sm9lIExlbnRyaWNoaWE=?=" <jlentrichia(nospam)@hotmail.com>>
>>wrote:
>>
>>>I like .NET over Java for a couple of reasons. The primary thing is the
>>>wealth of libraries that all the languages individually (VB, C++, etc) used
>>>to have that you can now access from any of the .NET languages.
>>
>>Well, Java has a pretty huge wealth of libraries too. (And it's not
>>just Java that targets the JVM, although that's easily the biggest
>>language using it, of course.)
>>
>>
>>>I like C#
>>>because I think the performance is great (primarily as a result of an
>>>excellently well written compiler).
>>
>>The C# compiler has little to do with performance - it's fairly simple,
>>I believe. The JIT and the libraries are what really drive performance
>>in .NET.
>>
>>
>>>Finally I think Visual Studio is still unmatched. I haven't used Eclipse,
>>>but I have used JBuilder and NetBeans and both are really slow (mainly
>>>because they are built on Swing).
>>
>>If you haven't used Eclipse you can't sensibly claim that Visual Studio
>>is unmatched. Consider what your opinion of an Eclipse fan would be if
>>they said that Eclipse is unmatched but that they hadn't used Visual
>>Studio.
>>
>>
>>>Swing was/is a really bad idea, they
>>>finally got it right when they started implementing UI features in the native
>>>API's of the respective OS.
>>
>>Have you used Swing recently? It's come on in leaps and bounds in the
>>last couple of releases, and I gather the new release due soon is even
>>better in this sense. Swing can be very slow when written badly, but
>>can perform well if properly handled.
>>
>>
>>>I think the biggest obstacle for .NET and MS as a whole to overcome is the
>>>ingrained concept in some developers mind that all MS products suck (not
>>>true), and that MS has lots of security issues (somewhat true), and most of
>>>all the general hatred because they have a monopoly on the desktop/office
>>>market. They also get a bad rap for being "proprietary" when in actuality
>>>they are pushing the envelope and just not waiting for a panel of "experts"
>>>to "standardize" the API they happen to be working on. And for this reason
>>>Visual Studio has superior support for Web Services.
>>
>>Interesting - I thought one of the criticisms levelled at Microsoft's
>>version of web services was that it went beyond the standard, thus
>>making it impossible for some standard web service consumers to consume
>>all .NET web services. I'm not sure I consider that superior support.
>>
>>--
>>Jon Skeet - <skeet@pobox.com>
>>http://www.pobox.com/~skeet
>>If replying to the group, please do not mail me too
>>
- Next message: Girish Bharadwaj: "Re: ?RE: Could events ever ?overrun? ?or otherwise be missed by a hand?ler?"
- Previous message: Drebin: "Re: How to retrieve user's domain correctly?"
- In reply to: Joe Lentrichia: "Re: Java and .NET (no Flames Pls)"
- Next in thread: Joe Lentrichia: "Re: Java and .NET (no Flames Pls)"
- Reply: Joe Lentrichia: "Re: Java and .NET (no Flames Pls)"
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|