Complete.Org: Mailing Lists: Archives: freeciv-dev: December 2001:
[Freeciv-Dev] Re: Development Strategies [Was Documentation, Usability a
Home

[Freeciv-Dev] Re: Development Strategies [Was Documentation, Usability a

[Top] [All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index] [Thread Index]
To: freeciv development list <freeciv-dev@xxxxxxxxxxx>
Subject: [Freeciv-Dev] Re: Development Strategies [Was Documentation, Usability and Development]
From: Gregor Zeitlinger <zeitling@xxxxxxxxxxxxxxxxxxxxxxx>
Date: Sun, 2 Dec 2001 12:41:15 +0100 (CET)
Reply-to: gregor@xxxxxxxxxxxxx

On Sat, 1 Dec 2001, Andrew Sutton wrote:
> let the old version stay as it is... there's no sense changing things 
> drastically now.
I agree

> i'm REALLY not a fan of java for performance critical applications. it's fine 
> for clients and web apps, but i just don't think a virtual machine can 
> guarantee the performance of a native application. 
I meant Java as a native application. a Just in Time Compiler can do that
and GCJ can compile to native code in the first place.

> rewriting the server in 
> java would mean that all would be developers would have to learn java and its 
> subtelties in addition to the actual game.
true. How many are those? And what would be the benefit from learning java
compared to the costs?

> i think c++ is probably the best choice above c. yes, c hackers will have to 
> learn c++, but since the syntax is similar, it shouldn't be too much of an 
> issue, 
but so is the java one. For a C hacker I think its the same cost to learn
C++ or Java

> and advanced developers can always answer questions about code. it 
> also builds to native binaries and can take full advantage of a) other c++ 
> toolkits b) the stl c) dynamically loaded libraries.
same for java

> if there were a *very* good reason to rewrite the server in java, 
the reasons I mentioned. 
1) Standardized (not true for C++)
 a) in syntax
 b) in GUI
2) ALOT of tools available for
 a) XML Parsing
 b) Scripting
 c) CORBA or other communication (RMI is faster but Java only)
 d) SVG (don't know if we need that)
 e) probably more I forgot

> some other 
> vm language (c#), or an uncommon programming language (smalltack, ocamel, 
> ada, fortran, basic, etc), than it would we would have valid considerations. 
2 problems about those:
 1) not very many people know those (or can easily learn them (except for
C# maybe)
 2) they don't have as many tools for XML, CORBA...
 3) other than that I like Ada and Eiffel from what I've seen but it
doesn't really matter conisering 1 and 2

> besides... you can write c++ to be very, very portable - without too much 
> effort.
ok, but as far as i understood you cannot use a couple of nice features
then

to summarize: Java is similar to C++ but has some critical advantages. The
drawback is that more people might know C++. (but those should have an
easy time learning Java, especially when they see some code)
Can those who it might concern comment on this?

-- 
Gregor Zeitlinger      
gregor@xxxxxxxxxxxxx



[Prev in Thread] Current Thread [Next in Thread]