[BuzzStarField]

This normally optimistic developer was feeling a little down this morning and found myself muttering the question "Why bother?". After thinking about it a bit, I feel a little better now. But like many of you, I am still a discouraged by PB's deficiencies and delays in getting them resolved. I have been working hard on some improvements to my app but it's getting hard to justify my efforts with an elephant named Magnetometer in the room.

Without access to the compass, I can't work on the enhancements that matter the most. This means that I can't deliver what I promised to my customers - and that bothers me a lot. I am itching to get started implementing Augmented Reality features and I am asking for some assistance from some of the many Android Gurus lurking in these precincts.

Below is a request that I posted in the official RIM developers' forums. I have cross-posted it here in the hope that someone can help with fast-tracking my app's "point-and-view" capability.

---------------------------------------
Magnetometer workaround for AIR app

Android developer's are fortunate in that the PB's magnetometer is available and working in the player. Unfortunately my native AIR app won't have access until NDK 2.0 is released and some kind C++ developer provides an ANE. Meanwhile I am exploring possible ways to fast-track testing features that require access to the compass. The augmented reality features that I want to add to my astronomy app are quite complex and I am looking for a way to get a head start on development. I think that a temporary work-around would be useful for other AIR developers who are wanting to include AR features in their apps but are being held up by SDK limitations..

Woiuld it be possiblee to build an Android daemon app to provide manetometer readings via a socket connection?Could a native AIR app communicate with such a service running in the Android player? Does such an app already exist and if so, how can I get my hands on it?

[Chrisy]

I hope that you hang in there and keep developing for BlackBerry. The PlayBook can only get better from where it is.

BBX is coming next too.

[BuzzStarField]

I hope that you hang in there and keep developing for BlackBerry. The PlayBook can only get better from where it is.

BBX is coming next too.

I'm just looking for ways to speed up the development process so that I can keep my promises. I like to work on the important things first and I was felling frustrated. I didn't mean to give the impression that I was thinking of quitting. No way!

PlayBook does indeed have a bright future.

On the other hand, sales of my app are very poor and there has not been a new review since DevCon. I will be submitting an update with significant enhancements shortly but it won't include the "big one" that everyone has been requesting. I am worried about the reception that it will receive.

[KermEd]

Just remember Half of developing is the adventure! It wouldn't be any fun if it was too easy for us!

[dcburke789]

On the other hand, sales of my app are very poor and there has not been a new review since DevCon. I will be submitting an update with significant enhancements shortly but it won't include the "big one" that everyone has been requesting. I am worried about the reception that it will receive.

Buzz.

I've had your app since it came out. Good job.

But as far as waiting on that one big update ... If this 2.0 debacle has showed us anything is that small incremental updates that fix small things and open new features and options is FAR better than holding out until its all done.



Sent from my BlackBerry 9850 using Tapatalk

[BuzzStarField]

Just remember Half of developing is the adventure! It wouldn't be any fun if it was too easy for us!

You're absolutely right of course. I don't think I would be happy without some challenges. The operative word here is "some". It should be noted that RIM could never, ever be accused of being too easy on BB developers. On the other hand, I heard a lot of great news at DevCon and things WILL get better.

Any comments on my proposal for a fast-track - is this possible?

[pinkert11]

I have your app as well, and it is great. Just used it last night as we took a small bench break while trick-n-treating. I think that we, the customers, understand that the delay of your major features /upgrades are there due to you waiting for updates from RIM.

I build a lot of system tools and work-flow databases etc and the advise that I can provide is wait and do it correctly. Your question above seems to be a round-about way to get the job done, not necessarily the correct or proper method. I often find that although round-about methods get the job done, they can also cause more issues then the wait would have. I am simply thinking about those on 2.0 beta vs those not. Will this cause a divide in your customer base of have and have-nots? With all of the issues that people have on a daily basis with 2.0 would you not want to wait for a solid system release to customers in Feb.

I think your hands are tied, and yes this is frustrating but some inter-mental updates and clear communication on your intend to upgrade on the release of tools and system upgrades is a good going forward strategy.

Again, you have made a great app, that is very enjoyable and useful in it present state, please hold tight and keep the app awesome.

[BuzzStarField]

@pinkert11, thanks for your good advice. Too many devs try to overcome SDK deficiencies by implementing kludges and workarounds in the final product. I'm just looking for a temporary means to get compass readings in order to test my AR code. The app would not be released to App World until there is actual APIs available.

BTW, GPS can give readings for compass heading, but because the device has to be in motion, this method is not appropriate for my app and is also useless for testing. The temporary work-around that I mentioned would give readings from the PBs magnetometer chip and would allow me to test my app's AR functions.

[Spaceheater]

Keep up the good work Buzz! We understand your frustrations.

[PanaSama]

Hi BuzzStarField, when you get access to the magnetometer will your app be something like Star Walk for the ipad?

[BuzzStarField]

Hi BuzzStarField, when you get access to the magnetometer will your app be something like Star Walk for the ipad?

You'll be able to point your PB at any point in the sky and see a fully annotated representation of that part of the sky on the device's screen. Searches will also be guided by the accelerometer and the magnetometer. Say you want to find where Jupiter is - you'll just have to point the device at the sky. An on-screen pointer will indicate which way you need to move the device in order put Jupiter in the cross hairs.

The following note may hurt my sales but so what: For the next few weeks you do not need any kind of app to help you find Jupiter. You can't miss it in the sky these days, even under bright city lights it's just so bright and beautiful!

[HaTaX]

Buzz, I'm probably not exactly the person you want to hear back from but I do have good news on your idea, it's very possible!

Here's just an example that shows you can pass data between the AP and applications running natively on the PB, take a look at the attached screenshot. On the left you see Android Player running "Remote Web Desktop" in the Android player, and on the right you can see my simple browser app open to "http://127.0.0.1:8999". Here's an installable BAR file for OS 2.0's AP if you would like to test or look into this more: Remote_Web_Desktop_v5.6.0.bar

Soooo, you need an Android daemon running on the PB that basically just streams GPS info over HTTP/whatever to 127.0.0.1 only. Not too bad from a programming stance, but I'm not sure how familiar you are with the Android SDK.

Hope this helps!

[BuzzStarField]

Buzz, I'm probably not exactly the person you want to hear back from but I do have good news on your idea, it's very possible!

Here's just an example that shows you can pass data between the AP and applications running natively on the PB, take a look at the attached screenshot. On the left you see Android Player running "Remote Web Desktop" in the Android player, and on the right you can see my simple browser app open to "http://127.0.0.1:8999". Here's an installable BAR file for OS 2.0's AP if you would like to test or look into this more: Remote_Web_Desktop_v5.6.0.bar

Soooo, you need an Android daemon running on the PB that basically just streams GPS info over HTTP/whatever to 127.0.0.1 only. Not too bad from a programming stance, but I'm not sure how familiar you are with the Android SDK.

Hope this helps!

Thanks HaTaX. It's very nice to hear from you - we have our differences but we both want good things to happen wrt the PlayBook. I am not at all familiar with the Android SDK so I would appreciate any help I can get. What I need is a simple daemon that listens for magnetometer events and updates local variables. My app would run a timer loop that makes periodic requests for current data - the daemon would pass a formatted string to my app. Doesn't have to be fancy or particularly robust - I just need something that's reasonably reliable for testing. I understand that Android AIR apps need a similar daemon for this same purpose. If so, that would probably be a good place to start.

I don't have any information about what data a magnetometer provides beyond a simple value for heading. NDK docs are not yet available for the PB. I would need to know which direction the device is pointed regardless of its angle/orientation in 3 dimensions. I definitely need a better understanding of how the sensor works before designing my implementation. Can you point me to an online link for Android documentation regarding usage of the compass in real AR situations?

Thanks fagain or responding. If you can point me in the right direction, it would be greatly appreciated.

[PanaSama]

You'll be able to point your PB at any point in the sky and see a fully annotated representation of that part of the sky on the device's screen. Searches will also be guided by the accelerometer and the magnetometer. Say you want to find where Jupiter is - you'll just have to point the device at the sky. An on-screen pointer will indicate which way you need to move the device in order put Jupiter in the cross hairs.

The following note may hurt my sales but so what: For the next few weeks you do not need any kind of app to help you find Jupiter. You can't miss it in the sky these days, even under bright city lights it's just so bright and beautiful!

Thanks BuzzStarField i've just bought your app, i look forward to the updates and i hope RIM gives solution to your magnetometer problem soon.

[Wolfgan]

But like many of you, I am still a discouraged by PB's deficiencies and delays in getting them resolved. I have been working hard on some improvements to my app but it's getting hard to justify my efforts with an elephant named Magnetometer in the room.

Without access to the compass, I can't work on the enhancements that matter the most. This means that I can't deliver what I promised to my customers - and that bothers me a lot.

I also got your app from a while ago and enjoying it at night from time to time. Did you try emailing/twitting Alec Saunders?
He seems to be very open, and tt least may be able to give you some direction or hope on when the magnetometer be exposed for AIR apps.
Wolf.

[elle-ade]

I'm just looking for ways to speed up the development process so that I can keep my promises. I like to work on the important things first and I was felling frustrated. I didn't mean to give the impression that I was thinking of quitting. No way!

PlayBook does indeed have a bright future.

On the other hand, sales of my app are very poor and there has not been a new review since DevCon. I will be submitting an update with significant enhancements shortly but it won't include the "big one" that everyone has been requesting. I am worried about the reception that it will receive.

Well, why don't you consider other ways of marketing your app? You can post on other blogs, post a link whenever you reply to an article, and so on.

[BuzzStarField]

I also got your app from a while ago and enjoying it at night from time to time. Did you try emailing/twitting Alec Saunders?
He seems to be very open, and tt least may be able to give you some direction or hope on when the magnetometer be exposed for AIR apps.
Wolf.

Thank you for the question and I apologize for the length of my answer. We have actually heard quite a bit of news. The following is the latest status as confirmed by RIM in various threads in the official developers' forums:

1. We do not have to wait for AIR 3.0 for ActionScript Native Extensions (ANEs). The PB implementation of current version of AIR (2.7) already supports ANEs so AIR developers now have access to NDK 1.0 APIs.This ability is not available on any other platform. A tutorial was recently posted so that we can get started using ANEs. This is all great news but it does not solve my magnetometer problem anyway (see next point 2).

2. Native APIs exposing the magnetometer will not be available until NDK 2.0 is released. The release date has not been announced (unless you accept "soon" as a firm date)

3. RIM has not announced that they will provide an ANE for the magnetometer nor the ActionScript classes that would handle events generated by the sensor. Therefore I am resigned to the fact that I will either have to learn C++ myself (not trivial) or I will have to convince a friendly NDK developer to lend a hand (i.e. do the work for me). Neither of these "solutions" is particularly attractive to the average AIR developer. One problem is that each and every developer would be re-inventing the wheel for each API function. Not very efficient!

4. There has been some talk of asking RIM to set up a formal mechanism so that developers with NDK experience can share common NDKs with devs like myself. There would be a central repository from which we could retrieve required libraries. This sounds like something that RIM could and should run with but so far nothing has happened. So I do not know if and when RIM will take ownership of this worthwhile project.

[BuzzStarField]

Well, why don't you consider other ways of marketing your app? You can post on other blogs, post a link whenever you reply to an article, and so on.

Actually the vast majority of my sales are driven from my presence here - and I owe a debt of gratitude to all of you for your support.

In addition I have recently started a Facebook page and will consider posting ads there when resources permit. I am also highly visible on the RIM support site and have articles in the CB App news section, PlayBookDaily.com and BlackBerrySync.com. If you do a Google search on the words "playbook" and "astronomy" you will get several references to my app - even a few tweets. I am constantly trying to get noticed but my need to do some actual coding (and and my "real" life and livelihood) keep getting in the way.

With all this, my numbers do add up to very much - however I don't complain much because they are currently above average for a niche PB app like mine. For me, this is a long-term project and I am prepared to be patient. My financial situation will improve if and when RIM sells some PlayBooks. Where there is volume, return on investment follows. February cannot come soon enough!

[HaTaX]

Thanks again or responding. If you can point me in the right direction, it would be greatly appreciated.

Here's a good starting point with an Android application that will read the magnetometer and accelerometer from the sensors and should give you a good basis to work from to see what the values are and what you can expect out of them.

Coding for Android: Using orientation sensors: Simple Compass sample

And here's some sample code showing a simple TCP server from the Android perspective:

Incorporating Socket Programming into your Applications � Think Android
&
Simple connection example part II - TCP communication | Hello Android

Hopefully between those links you can at least get started. I'd go grab the Android 2.3.3 SDK (API10), you can get it here: Android SDK | Android Developers

Once you get it installed and such, if you'd like some more help I'd be willing to see what I can help out with. Let me know...

[BuzzStarField]

@HaTaX... Thanks!

Fantastic. Just what I was looking for.

[HaTaX]

You're welcome! Also, I tried to choose the 3 different sites strategically so you would get exposure to some really helpful Android coding sites. That's one thing that's different from the BB platform, TONS of shared code out there to help people get started quickly with the platform. Those 3 sites should provide you with examples for most coding scenarios, good luck!

[JDukeOSBB]

I cannot personally contribute much to this thread, but I have an ongoing email conversation with A. Saunders at RIM, and I specifically asked about this in my reply to him. Hopefully something helpful will come of it.

[BuzzStarField]

I cannot personally contribute much to this thread, but I have an ongoing email conversation with A. Saunders at RIM, and I specifically asked about this in my reply to him. Hopefully something helpful will come of it.

Your action is much appreciated and this is yet another reason why CB is a great place to be.

[JDukeOSBB]

Your action is much appreciated and this is yet another reason why CB is a great place to be.

Heck, I was so excited when I first saw your app way back when it came out, I couldn't buy it fast enough! I love your app and I want magnetometer support nearly as bad as you do. I have noticed that Google Sky (on my Playbook via AP) is quite quite inaccurate, in fact, it's flat out wrong.

Posted from my CrackBerry at wapforums.crackberry.com

[snoozininsomniac]

i just saw this thread and immediately bought the app. i agree that this app could easily land in top 10 playbook apps. while i cant contribute to your dilemma, i do hope the information you need from rim comes soon.

-a supporting customer