Are Java geeks welcome here?

[mikester]

First of all, I love the concept of LinuxMCE, and props to all of you who have been putting in so much work to get this project off the ground.  There's still a long way to go, but I think this idea is worth pushing forward.  In fact, I've been envisioning writing my own HTPC software for the last 6 months because I haven't been happy with anything out there.  Then I found LMCE and guess what - there's about an 80% overlap between your vision and mine.  So count me in for helping out with this project after I get my system up and running (currently in progress).

There is just one small detail.  Even though I've been a professional developer for almost 20 years, it has been about 8 years since I used C/C++ regularly (although that is all I used the previous 12 years).  In fact, I loathe the idea of having to use C again, and the few times I have to use it at my job, it makes me wish I was getting a root canal instead. (just kidding  ).  I am a Java developer through and through, and specifically I have a lot of experience with Eclipse plugins, RCP applications, and the SWT toolkit.  Before you C guys roll your eyes and start throwing out all the same old arguments - I defy you to look at a good SWT application and tell the difference between a native Windows/OSx/GTK application.  It's that good, both in performance and host OS integration.

OK, so that being said, I really want to be able to help out with this project, and if push came to shove I would probably be willing to dust off my C++ skills for short and specific tasks.  But I feel I have a lot to offer with my Java/RCP/SWT skills, particularly in the 2 areas where I see that LMCE needs the most help - media management and UI customization.  Now maybe work is already underway in these areas, and I certainly don't want to step on anyone's toes, but if not, I would love to look into either rewriting the GUI editor (forgot the name, but I know everyone complains about it) or adding a nice media library frontend with IMDB and TVDB integration, etc.  I promise you I could write nice apps using Java that would make even the Windoze MCE people drool.

I just wanted to get some feedback on this idea before I get my hopes up.

[hari]

All contributions are welcome!!

Some Java things that come up my mind:
* HAdesigner
* Improving the JavaMO
* Doing a full orbiter for Android
* integrating the cidero upnp software
* Doing a java RA/dce library
...
profit

best regards,
Hari

[mikester]

Good to hear!  As I get my system up and running, I'll start looking at the code and learning the architecture better so I can jump in.  I've already spent some time reading over the development docs, although I'm not sure how much of that is up to date.  But there's nothing like getting your hands dirty digging into source code to learn a system.  I figure building a development core and being able to compile the C code should be my first step, then I'll go from there.

BTW - an Android orbiter is a great idea.  I looked at the SDK a few months back when they first released it, and thought it would be fun to write an Android app, but never came up with a good idea.  It seems like the mobile phone app space has boiled down to half a dozen compelling apps, and a bunch of "me-too" apps, so there was nothing I saw that I could exploit.  But that was before LMCE...

[kir]

Indeed, every piece of code is welcomed And there is no (hopefully!) language-wars people here. Please feel free ask questions about the arch - this is always a good start. If you haven't read the wiki - read it, as there is a lot of information on the system arch (some pages can be outdated, so don't trust it 100% - LMCE is a live system, so things could change).

By the way, if I recall correctly, the 0710 betas does install java vm by default

[elspic]

mikester:

Glad to have you on board. Hari is probably right on track with tasks that need to be worked on; the only thing that I would ad is the Floorplan Designer in the setup. I'm going to try to get a small developer conference together soon-ly to see just what we, collectively know about the code-base.

As far as the "profit" portion Hari, I've got full faith in a LMCE orbiter being good enough for an award in the Android Developer's Challenge. If I knew java, I'd help, but I'd really like to see an android version to submit for the deadline so if anyone wants to help Hari run with what he's done so far, we could probably get something together before the April 14th deadline. A little money in a development pool certainly couldn't hurt. You could also submit it yourself and just take the money and run
Just kicking around ideas.
-Ruben

[aaron.b]

I vaguely remember a Java DCE a while ago to allow Java devices.  We're trying to see if we can find where that code went.

I agree that Java could make a huge difference in LinuxMCE with the mobile phone apps.  I designed the original 'mobile orbiter' circa 2003 on the Symbian Series 60 platform.  It was considered a great killer app.  Everybody loved being able to control the system both at home and while away on the mobile phone.  And it's become even more important now as oil prices rise and people are thinking more about the need to be 'green', remote control of the home is a really hot topic right now.  Instead of leaving the a/c on all day so you return to a cool home, just turn it on 30 minutes prior to your return, etc...

The problem is Symbian Series 60 is dead, and there is no ubiquitous o/s for mobile phones.  So if members of the FOSS community want to work on a Java orbiter that runs on a variety of mobile phones, Pluto will provide direct access to our devs and a lot of support since it's very important to us as well.

[tschak909]

aaron.b: I don't know if you've been following, but hari basically has one running.

http://forum.linuxmce.org/index.php?topic=2893.msg25973#msg25973

it's a testament to the design of the system that Hari was able to do this so bloody quickly, and I'm nothing short of impressed.

-Thom

[bulek]

Hi,

if I may add to the todo list :
- Web Orbiter has currently quite annoying refreshing...

Regards,

Bulek.

[darrenmason]

As Aaron hinted as there was a sourceforge project that a guy put together to create a Java interface to the DCERouter. http://sourceforge.net/projects/jpluto/
One of the pluto guys actually changed some stuff so that the messages were more friendly to another language.
I think the ability to create devices in java would be a gain in the project. I would prefer it as well, as like the original poster I thought I had gotten away from C++ years back.

A rewrite of HADDesigner sounds like it could be well suited to someone with SWT skills as well.

ps; Just out of interest mikester, what is the 20% (of your vision) that you think can't be done/supported by this project?

[ddamron]

aaron.b: I don't know if you've been following, but hari basically has one running.

http://forum.linuxmce.org/index.php?topic=2893.msg25973#msg25973

it's a testament to the design of the system that Hari was able to do this so bloody quickly, and I'm nothing short of impressed.

-Thom

All I can say is ME TOO!

Glad to see another developer wanting to get their hands dirty!

I never really considered myself a software guy too much, most of my tinkering is with hardware.  The software side was a necessary evil.  (This is why I've been putting off c++ for so long)

Heh, I got tired of waiting for a specific feature, so I decided to attempt it myself.  Next thing I know, I'm being called a 'dev'.. heh, funny how things turn out.
 

[jondecker76]

Bulek - regarding the web orbiter refreshing.. I updated the wiki just a few days ago about how to solve the annoying web orbiter refreshing..  Go to the wiki, search for web orbiter, then look at the bottom of the page for how to change the refresh rate.

[mikester]

ps; Just out of interest mikester, what is the 20% (of your vision) that you think can't be done/supported by this project?

The main thing I would have done differently is use Java from the start, so that the core, orbiters, md's, everything would be cross-platform.  So if you wanted to setup a core on a Mac and have a mix of Windows and Linux media directors - no problem.  The other main difference is that I was focusing more on media handling and treating stuff like home automation as an afterthought, whereas LMCE is practically the opposite.  But that's good - I think Home Automation, Security, and Telecom are the harder problems to solve, whereas media handling has been done over and over.  IMO it would be easier to enhance a system like LMCE with more robust media handling than to retrofit a whole house control solution onto a primarily media-based front end.

Almost everything else LMCE does is eerily close to what I have been thinking of over the last year, which is really exciting to me because I've wanted a system like this for some time (to the point I was ready to start writing it myself) and all of a sudden I find that this system does exist and has a huge headstart over what I had been planning to do.

[Ender]

Hi everybody,

I remember that I've wrote a simple java dce layer to be able to have a tivo orbiter for tivo system few time ago. You can find it here : http://svn.linuxmce.org/pluto/trunk/extra/TiVo_Orbiter/.
You can find the dce classes here : http://svn.linuxmce.org/pluto/trunk/extra/TiVo_Orbiter/src/com/tivo/hme/dce/

Regards,
Ender

[darrenmason]

There you go Mikester - Ender has given a good head start. You could convert his Tivo HME based orbiter to run with SWT. I think that would be a useful item and would familarize you with a lot of the workings?

[danielk]

Just FYI, with Pluto's permission I've attached GPL headers to all the Java files in the TiVo_Orbiter.