[Android] NITdroid 0.4.x - Page 4 - maemo.org

EasternPA | # 31 | 2009-01-22, 15:43 | Report

Originally Posted by Andyschism View Post

Could not agree more. I love reading these threads and seeing the progress, but I have no clue whats being said about 75% of the time. Lol.

Some of the "easy install" steps were a little quirky, even for a 15 year Linux guy. It was at times unclear where some of the steps were to be executed (on the Linux host or the tablet, and if on the tablet via the tablet's own terminal emulator or via the adb shell), and instructions to power down and/or disconnect the USB cable were also completely missing in places. Believe me, I read the first thread for 3 good weeks before diving in, and even after that I was pleasantly surprised to find out how easy it was to return to Maemo.

After 0.4.1 comes out, and we see how it behaves, how about in the spirit of "everyone contributing something", maybe you and I can sit down and "newbyize" the instructions? I've had to write simple instructions for people for almost 30 years, but I'm sure there's some stuff that I assume everyone knows or can easily figure out, but they just don't or can't.

Have you tried the install yet or is it still too confusing to even try it out?

The Following User Says Thank You to EasternPA For This Useful Post:
Andyschism

EasternPA | # 32 | 2009-01-22, 15:48 | Report

Don't know what happened, but the Software Directory worked within 90 seconds today. Weird.

Also, wifi is dropping a *LOT* now. I wonder if the MID profile is trying to put the wifi in some kind of sleep state but is killing radio altogether.

Last edited by EasternPA; 2009-01-22 at 16:19.

Andyschism | # 33 | 2009-01-23, 19:09 | Report

Originally Posted by EasternPA View Post

Some of the "easy install" steps were a little quirky, even for a 15 year Linux guy. It was at times unclear where some of the steps were to be executed (on the Linux host or the tablet, and if on the tablet via the tablet's own terminal emulator or via the adb shell), and instructions to power down and/or disconnect the USB cable were also completely missing in places. Believe me, I read the first thread for 3 good weeks before diving in, and even after that I was pleasantly surprised to find out how easy it was to return to Maemo.

After 0.4.1 comes out, and we see how it behaves, how about in the spirit of "everyone contributing something", maybe you and I can sit down and "newbyize" the instructions? I've had to write simple instructions for people for almost 30 years, but I'm sure there's some stuff that I assume everyone knows or can easily figure out, but they just don't or can't.

Have you tried the install yet or is it still too confusing to even try it out?

Awesome idea! I'll try and help as much as I can but like I said I really dont know whats going on. I just see people saying this isnt working then someone says try this and it works or it dosnt. Its like traveling to a foreign country and not knowing the language at all, but I can make out the body language and facial expressions!

jakemaheu | # 34 | 2009-01-23, 19:50 | Report

Have you guys even read my install guide lately?

I wrote a script that does EVERYTHING, assuming you have the packages. All you have to do is select your partition sizes. If you need, get Linux (Ubuntu Live CD would be perfect). Really, it isn't that difficult. There are also steps in it that tell you exactly what to do.

jakemaheu | # 35 | 2009-01-23, 20:00 | Report

Originally Posted by EasternPA View Post

I think the biggest issue here is figuring out how the internal card is detected by the installation host when the USB is attached. Apparently Ubuntu sees the card as /dev/sdc but for me on CentOS, the card was detected as /dev/sda. And does it register as a different scsi device if the usb cable is connected to a different usb port? With all of the different distros out there and kernel module options, I don't think there's an easy way to tell which scsi device is the tablet. Perhaps do an fdisk -l and look for the only 2GB scsi drive? Is there something in proc that shows the vendor of each attached drive?

I'm waiting for approval to formally host the project at a community development site. Once that gets up and running, perhaps we can get some more visibility and start up an installer sub-project. I know someone who knows dialog. Its just a matter of getting his time. As I've said before, my ultimate goal is to publish a bootable USB key for each release.

The /dev/sdc was just an example. The script works like this:

Code:

mount|grep -e /dev/sd

It then asks you for the last result, which would be the just-plugged-in tablet. It then asks for it again without the number on the end, then with a 3 on the end. These entries are used as variables that are referred to during the partitioning, mounting, etc.

The problem with looking by size is that different people have different sizes (that's what she said) of SD cards and having used a bunch of them, many have weird sizes when coming from strange places in Shenzhen, China. (931MB instead of 1024MB)

EasternPA | # 36 | 2009-01-23, 21:35 | Report

Originally Posted by jakemaheu View Post

The /dev/sdc was just an example. The script works like this:

Code:

mount|grep -e /dev/sd

It then asks you for the last result, which would be the just-plugged-in tablet. It then asks for it again without the number on the end, then with a 3 on the end. These entries are used as variables that are referred to during the partitioning, mounting, etc.

The problem with looking by size is that different people have different sizes (that's what she said) of SD cards and having used a bunch of them, many have weird sizes when coming from strange places in Shenzhen, China. (931MB instead of 1024MB)

Sorry if we set you off a bit

But seriously, fanoush posted code on the previous page that generates a list of scsi drives and the vendor. If we were going to go for a polished installer, I would use his code to find the drive with "NOKIA" as the vendor.

And to answer your earlier question, the only "easy install guide" I'm aware of is in the "getting nitdroid running" thread. I ran your script, when 0.3.2 first came out but all it did was download the latest versions of everything and untar them on my linux host. It didn't actually try to mount or untar them to the nit. Plus at the time, solca missed out on updating the symliks so I ended up having to download everything by hand before untarring again and reflashing by hand.

If there's a newer guide somewhere else, or if you've gone back there and updated it, then awesome, please post a link! Finally, the Ubuntu CD when booted into live doesn't recognize any of my three Thinkpads' built-in wired or wireless Ethernet adapters, and I can't blow away my Rocks Cluster frontend, so I never got to test out the instructions in an Ubuntu setting.

chyang | # 37 | 2009-01-24, 07:55 | Report

the power button doesn't work and the original worked maps application cannot work any longer.

chyang | # 38 | 2009-01-24, 18:06 | Report

found some of the clue to make maps and some other applications run available:
--- platform.xml.save 2009-01-25 00:47:53.000000000 +0800
+++ platform.xml 2009-01-25 00:48:00.000000000 +0800
@@ -117,6 +117,10 @@


+ <library name="com.google.android.maps"
+ file="/system/framework/com.google.android.maps.jar" />
+ <library name="com.google.android.gtalkservice"
+ file="/system/framework/com.google.android.gtalkservice.jar" />
<library name="android.awt"
file="/system/framework/android.awt.jar" />
<library name="android.test.runner"

But now, i am running meeting with the issue like:
I/MapActivity( 1372): Handling network change notification:CONNECTED
E/MapActivity( 1372): Couldn't get connection factory client
E/MapView ( 1372): IllegalStateException on drawMap. Wiping cache.
E/MapView ( 1372): java.lang.IllegalStateException: Null Bitmap!/loading_tile.png=17301864
E/MapView ( 1372): at com.google.common.graphics.android.AndroidImage.<i nit>(Unknown Source)
E/MapView ( 1372): at com.google.common.graphics.android.AndroidImageFac tory.createImage(Unknown Source)
E/MapView ( 1372): at com.google.googlenav.map.MapTile.c(Unknown Source)
E/MapView ( 1372): at com.google.googlenav.map.MapTile.getImage(Unknown Source)
E/MapView ( 1372): at com.google.googlenav.map.Map.a(Unknown Source)
E/MapView ( 1372): at com.google.googlenav.map.Map.drawMap(Unknown Source)
E/MapView ( 1372): at com.google.android.maps.MapView.drawMap(MapView.ja va:997)
E/MapView ( 1372): at com.google.android.maps.MapView.onDraw(MapView.jav a:443)
E/MapView ( 1372): at android.view.View.draw(View.java:5660)
E/MapView ( 1372): at android.view.ViewGroup.drawChild(ViewGroup.java:14 47)
E/MapView ( 1372): at android.view.ViewGroup.dispatchDraw(ViewGroup.java :1211)
E/MapView ( 1372): at android.view.ViewGroup.drawChild(ViewGroup.java:14 45)
E/MapView ( 1372): at android.view.ViewGroup.dispatchDraw(ViewGroup.java :1211)
E/MapView ( 1372): at android.view.ViewGroup.drawChild(ViewGroup.java:14 45)
E/MapView ( 1372): at android.view.ViewGroup.dispatchDraw(ViewGroup.java :1211)
E/MapView ( 1372): at android.view.View.draw(View.java:5663)
E/MapView ( 1372): at android.widget.FrameLayout.draw(FrameLayout.java:3 24)
E/MapView ( 1372): at com.android.internal.policy.impl.MidWindow$DecorVi ew.draw(MidWindow.java:1725)
E/MapView ( 1372): at android.view.ViewRoot.draw(ViewRoot.java:1167)
E/MapView ( 1372): at android.view.ViewRoot.performTraversals(ViewRoot.j ava:989)
E/MapView ( 1372): at android.view.ViewRoot.handleMessage(ViewRoot.java: 1418)
E/MapView ( 1372): at android.os.Handler.dispatchMessage(Handler.java:99 )
E/MapView ( 1372): at android.os.Looper.loop(Looper.java:123)
E/MapView ( 1372): at android.app.ActivityThread.main(ActivityThread.jav a:3779)
E/MapView ( 1372): at java.lang.reflect.Method.invokeNative(Native Method)
E/MapView ( 1372): at java.lang.reflect.Method.invoke(Method.java:521)
E/MapView ( 1372): at com.android.internal.os.ZygoteInit$MethodAndArgsCa ller.run(ZygoteInit.java:745)
E/MapView ( 1372): at com.android.internal.os.ZygoteInit.main(ZygoteInit .java:503)
E/MapView ( 1372): at dalvik.system.NativeStart.main(Native Method)

solca | # 39 | 2009-01-26, 06:29 | Report

0.4.1 userspace released.

2.6.28-5 kernel released too.

Last edited by solca; 2009-01-26 at 06:31. Reason: Fix links.

The Following 4 Users Say Thank You to solca For This Useful Post:
Bundyo, maxkilla, meizirkki, spjndizzy

EasternPA | # 40 | 2009-01-26, 15:51 | Report

Went to 0.4.1 and latest kernel. OSK is back but crashes after one use. ADB hasn't worked for me under Linux in 0.4.x at all. Has the device id changed again? Is it time for a new version of adb?

I believe the system clock interface may have been improved, as I didn't have to set the date and time after reformatting for this install. Software directory comes up reliably now but I still can't install software without an SD card.