javacompiler-users

[Sean M. <sm...@gm...>]

vitallis <http://forum.java.sun.com/profile.jspa?userID=285720>
Posts:720
Registered: 8/1/00
	
*Re: Help me on njawin   
Dec 8, 2005 12:35 AM (reply 7 of 7) *
	
 
	
Hi,
trying to run jawin you waste your time! This project is developing for 
a number of years and still have many of bugs. In 2002 I rewrote this 
project and called if NJAWIN (the commercial version is OLEJA). NJAWIN 
is free and available at
http://simtel.net/product.php[id]60701[sekid]0[SiteID]simtel.net
NJAWIN supports only OLE Automation, WIN32 function call support was 
removed. COM specifics are hidden.
NJAWIN looks for its native DLL in the DIR where its JAR is stored. So 
keep NJAWIN's DLL and JAR in the same DIR.


That post is from the Sun Java forums. I'm not comparing the 
functionality of Jawin/NJawin/OLEJA to NativeJ/javaCompiler in any way; 
I'm stating simply that the project started open source and it's now 
commercial (OLEJA). Also, quite revealingly, I have been completely 
unable to find NJawin available anywhere on the Internet. I think the 
author had it taken down! He certainly doesn't support it anymore.

Furthermore, I tried a recent version of OLEJA and I much prefer the 
functionality and improvements of Jawin. I'm not saying your project 
must necessarily be inferior to an open source alternative, but it seems 
to be a trend in my experience; NJawin is merely one example.

I think you're missing some of the point of the commercialization of 
F/OSS software. The point is not to destroy your F/OSS offering and stop 
maintaining it, after "using" the help of the community to get you 
started. The point is to offer a commercial version with value-added 
features that go beyond the core offering, such as priority tech 
support, fancy or thorough documentation, and a license that allows it 
to be used royalty-free in other commercial, closed-source projects (if 
the OSS license used for the base project doesn't already allow this.)

Need a case study? Take your pick: openSUSE/SLED; Fedora Core/RHEL; 
FreeSpire/Linspire; QT; the list goes on. These projects represent what 
I view as the "real" point of F/OSS -- providing the core 
developer/project founder(s) a profit source while still offering the 
open source option to individuals and non-commercial entities. From what 
I understand of your announcement, javaCompiler will die at 0.8, and any 
improvements you make to the core functionality of this product will go 
into, and only into, the commercial product. The project thus ceases to 
be of use to the open source community as soon as an incompatible or 
newer version of SWT, AWT, or the Java Specification becomes needed.

Just because other developers can choose to pick up the gauntlet you 
dropped and continue javaCompiler development doesn't seem to justify 
essentially abandoning an OSS project. You and all F/OSS developers are 
of course free to abandon projects when you have other 
responsibilities/needs/interests, but this is a special case where 
development and interest on your part continues, but under another 
banner. If the majority of F/OSS projects take the route of 
javaCompiler, well, there wouldn't be very much of an Open Source 
movement at all, would there? In fact, you'd have to license GCJ if you 
wanted to take advantage of any new features they happen to develop for 
you over the course of the next n years. You wouldn't like that very 
much, would you???

I'm not trying to give you a hard time. I actually used javaCompiler on 
a small non-commercial project for my mom and she really liked it (MY 
code, bundled nicely with YOUR program into an EXE rather than a messy 
JAR). I appreciate your work and the amount of effort you have invested 
into this project. I just wish it didn't have to end this way.

I have been considering using the Java 1.5 & JFace platform for a new 
desktop application requested at work. The standard Windows tools used 
by our in-house developers are dated and cumbersome, and are not 
resilient to changes in the upcoming Vista. So Java will go a long way 
to making the platform leap safely. Due to licensing issues (and product 
support concerns, for that matter) I will probably go with Excelsior JET 
rather than javaCompiler. If javaCompiler were continued, and NativeJ 
were a commercial version with added benefits, I would more seriously 
consider buying NativeJ.

As for the OSS programs compiled into EXE/bins, my project website is in 
the works - I'll have to host it without any fancy server-side 
processing, since my web space is a 100mb chunk on a server that only 
runs Apache on our university network. I'll start by using 
Cygwin/MinGW/classpath with programs that don't compile with 
JavaCompiler.. my first one will probably be PasswordManager. I'll look 
at the CVS version of SwingWT to see if it has the implementations I 
need. :)

Regards,

Sean

Marco Trudel wrote:
> Sean McNamara wrote:
>   
>> I can still download javaCompiler 0.8, including source, online.. if the 
>> license for that permits, we may be able to fork it under another name 
>> and continue to improve it F/OSS style.
>>     
>
> Until JavaCompiler 0.8, it was GPL. So feel free to continue development 
> of it under GPL terms. But also understand that the GUI frontend is only 
> 10% of the whole work. Building, maintaining and testing the compilers 
> is the real work.
>
>
>   
>> Sounds like it's gone the way of 
>> NJawin, but imho Jawin is way better anyway, and it's Open Source, so...
>>     
>
> Do you have a link to NJawin? I never heard of it and a quick google 
> search didn't bring anything up...
>
>
>   
>> I'm disappointed, but not altogether surprised... In Marco's defense, 
>>     
>
> There is nothing to defense. If you think there is, then you don't 
> understand the concept of free software. Although I'm starting to sound 
> like a broken idiotic robot: please read the FAQ.
>
>
>   
>> Be that as it may, I am saddened to see another OSS project dry up,
>>     
>
> Why do you think it will dry up? I will maintain it as I always did...
>
>
>   
>> and 
>> I hope we can make the best of it by using Marco's work as a basis for 
>> other projects. I may begin learning the internals of javaCompiler and, 
>> while not necessarily using the JavaCompiler GUI,
>>     
>
> In this case, just compile GCJ yourself. That's the whole idea behind 
> free software ;-)
> But keep in mind that, depending on your background, you'll need a month 
> (if you've my experience in this field) until years (if you're able to 
> do it at all) to reach the same quality that NativeJ provides right now 
> (and big improvements are coming).
> So, if we calculate one month for the task, you would have hourly wage 
> of 0.25$ (you spare one NativeJ licence). Ah yeah, you wouldn't have 
> your usual wage from your job and probably need longer for it.
>
>
>   
>> I'll take the time to 
>> learn what goes wrong in the native compilation of several popular Java 
>> programs. My goal will be to try and get some popular OSS Java programs 
>> compiled into native ELF32 and Win32 binaries for ease of distribution.
>>     
>
> Cool! Is there a project website?
>
>
>   
>> I have several SWT programs in mind that should be easier to get working 
>> than the Swing/AWT ones.
>>     
>
> Yes, SWT completely works with GCJ. Now that Java is under GPL with the 
> GPL exception, Swing and AWT support can be taken from Sun. So that will 
> be supported very soon as well...
>
>
> Marco
>
> -------------------------------------------------------------------------
> Using Tomcat but need to do more? Need to support web services, security?
> Get stuff done quickly with pre-integrated technology to make your job easier
> Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> _______________________________________________
> javaCompiler-users mailing list
> jav...@li...
> https://lists.sourceforge.net/lists/listinfo/javacompiler-users
>
>