Advice for Future Development

[lboucher]

Hi All

So I am sure most everyone has seen the android development I have done. Well now that this year is finally over from a development standpoint, its time to start looking forward to next year.

I am hoping to take it nice and slow and do things right over the course of 2011.

So I have many ideas (lots of which come from you guys) to add to my app, but need advice on a critical piece first.

I absolutely must create a web/mobile front end for this thing and include a real database. Thinking mySQL but I really don't know what I am talking about.

I am about to start researching on the web, but nothing is better than direct advice from others I can talk with and that aren't selling anything.

So here is what I need:
A user account system where people can sign up, add info, add a pic etc...
I would really like it to support both free and paid accounts. (Don't worry everything is free for DIY'ers, I was just going to charge for commercial light shows that want to get added to the light show finder. If they are charging for there show, then i feel i should get paid for advertising them.)
The ability to have different fields for different account types (free vs paid)
I need to be able to access the database from the app.
I am already using wordpress, not sure how that impacts things??
The ability to resize/adjust/compression/resave images so they are always properly formated.
A mobile front end to create accounts??
The ability to add another entry to the database without an account. (What I am thinking here is the everyday joe smo driving down the street sees an awesome show, it would be cool for him to snap a pic and add it to the database. Obviously he won't want to create an accont, just submit the house.)

Sideline things:
How does one programatically get an address from a lat long??

Ohh and free also rocks, (Or dam near close to free.)

I am sure I am missing some design considerations.
Any help would be awesome.

Thanks

[lboucher]

Address to lat long,

http://code.google.com/apis/maps/documentation/geocoding/

[ptone]

What you are looking at is more web app, then mobile app. For this you should pick a framework. Either way you are in for a bit of a learning curve - but that is not a bad thing. As a Python developer, I'm quite partial to Django, another popular one based on the ruby language is Rails. If you are really wedded to Java - there are choices there too:

http://en.wikipedia.org/wiki/Comparison_of_web_application_frameworks#Java

also see: http://olex.openlogic.com/wazi/2010/choosing-the-right-java-web-development-framework/

But like I say, I'm partial to Django - it also has probably the best geographic support via the GeoDjango module

It would be cool to develop a database of light shows - and then tie your android app to that, but still keep them as independent projects.

-Preston

[lboucher]

but still keep them as independent projects.

-Preston

Why do you suggest keeping them as independent projects?
Its funny I am sitting here looking at this big diagram with almost a dozen different ideas/thoughts that can all be integrated together into this massively awesome thing. I work in systems so I know grand schemes like this can easily devolve into a product that instead of doing everything awesomely, does everything poorly.

Since before I even got my droid I have been really amazed at all the possibilities. Some really awesome things can be accomplished.

[Hunter310]

One reason to keep them independent is the integration of other smartphones. As an iPhone user, I would love to access the same data you have in your Light Finder app, but on my iPhone.

I am sure other smartphone users (Blackberry, Windows Phone, etc) would agree.

Steve E.

[budude]

Can it use celluar location to get you located in addition/instead of normal GPS location? The reason I ask is that I don't normally have my GPS enabled since it sucks up the battery even more than it does now (which is a LOT).

[lboucher]

Right, so they would be independent in that sense. Not that there is any chance I will ever take the time to learn how to develop on multiple mobile setups. But others can build apps for other systems, aka the iPhone version is already being worked.

Blackberry people may have to wait awhile, i mean, thats kinda a dieing breed isn't it?

[lboucher]

Can it use celluar location to get you located in addition/instead of normal GPS location? The reason I ask is that I don't normally have my GPS enabled since it sucks up the battery even more than it does now (which is a LOT).

Just because you have it enabled doesn't mean its using battery. An app has to request it to turn on.

If the little gps icon isn't showing up in the title bar then it shouldn't be sucking power.

All that said, i orginally forced gps due to a logistic issue with the light controller app. AKA cell location can be 5 km off, which means two people in the same town couldn't use the light controller app.


When i switched over to the light show finder app I just kept it that way.
(If you have the latest version you might notice that I do use cell location for initial fix if needed, but still require GPS. Figured people would need that to get navigation anyways, so i left it in the code. All that said, you are right that it isn't needed for the show finder.)

So all that said, this is something for me to work on next year.

[lboucher]

Thanks, after looking around I think i will go with Django, learning some python is a plus and the latest book has good reviews.

[ptone]

Great - I think you will be happy with both python and django. I'm chronically overextended - but I'd be happy to help best I could on a Django based show database.

I also have an idea to make it easier to manage a mobile device app. I'm working on iPhone - but principle would be the same.

The idea would be to have the server running on the show computer (or on another server on the network). The mobile device then just connects over an open local wifi (or one could open a port on their firewall and just use dyndns).

I'm aiming to keep it all HTML and websocket based, so that people don't even need to download an app - just go to a URL.

For example, I just figured out how to integrate the accelerometer data with websockets from a web page - no app involved in this demo:

http://www.youtube.com/watch?v=KL9cdQxsNRM

This can also be combined with multi-touch events (also via JS and websockets) to come up with a pretty cool show controller.

-Preston

[dpitts]

I would like to see a plug-in with the ability to test the channels with an Android. A simple check box for every channel would suffice. An IP address and port of computer running Vixen would be all that is needed for input. The plug-in could even connect to a web service and the testing app could be a web application. I do love the public interaction of your app. I would be willing to develop the testing app if you are not interested. Could help me get started with a simple input plug-in. The details of input plug-ins are not easily found. Do you know where I can find documentation or example of input plug-in.

[kychristmas]

I would like to see a plug-in with the ability to test the channels with an Android. A simple check box for every channel would suffice. An IP address and port of computer running Vixen would be all that is needed for input. The plug-in could even connect to a web service and the testing app could be a web application. I do love the public interaction of your app. I would be willing to develop the testing app if you are not interested. Could help me get started with a simple input plug-in. The details of input plug-ins are not easily found. Do you know where I can find documentation or example of input plug-in.

There's some source code around somewhere. Use the google search to search for "Vixen Plugin Source code" That's a more advanced plugin you are looking for because it requires a some in-between code.

There's also a Vixen Server that can process commands from another application/server. I never got it working, but many people have. You could setup a little Web-page for the control. There's a guy on here that had a webcam and site that allowed you to start sequences and turn channels on and off on his demo display.

If you are just looking for a simple testing tool. I used my IPOD Touch/iPad to run test. I did it via a VNC Iphone app that allowed me to remote to my show computer. Very easy on the iPad, a little tougher on the Ipod Touch with the smaller screen. But in both cases its very useful and literally a 10 minutes setup.

[boyelroy11]

Something that I found myself wanting was contact info. There was another show near my house, but it was never on when I went by. The house was usually dark, so I didn't want to knock. If email info could be included, I could have made a contact within the diy community. Obviously, not everyone will want info out there, so an opt out could be offered, or just a generic "contact us" link with hidden email address.

Other than that, I think I looked at every listing on my phone, even if not in person... :) Thanks for your dedication to this.

Bruce

[lboucher]

Hi All

I do still plan on working on this, kinda distracted with another project and the rest of life right now. There is sample code out there, and i could send ya some, but it might be worth waiting for Vixen 3.0. If i do make this I will build a test app right into it.

Thanks for the contact idea, thats a good thought. Will have to find a way to keep the email address secret though.

[DynamoBen]

Hi All

I do still plan on working on this, kinda distracted with another project and the rest of life right now. There is sample code out there, and i could send ya some, but it might be worth waiting for Vixen 3.0. If i do make this I will build a test app right into it.

Thanks for the contact idea, thats a good thought. Will have to find a way to keep the email address secret though.

v3.0 is only in the concept stages, not sure when it will release. I would suggest you get this going on the current version of Vixen and port it if necessary later.

[dirknerkle]

Saw this article in a trade rag I get every week... thought it was interesting since it focused on Android and we have some development going on with it...

http://www.crn.com/news/applications-os/229402026/is-the-android-operating-system-too-risky-for-enterprise-business.htm