Goddchen ’s Blog

Hey everybody,

i think many of you know the problem with android 2.x phones: they can’t connect to pptp vpn servers which have encryption enabled.

You get the error:

E/pppd ( 6070): unrecognized option ‘+mppe’

But since yesterday this problem is solved I had a similar idea for a while now but didn’t have the time to try it. As the error message states “pppd” is missing the +mppe feature. So, why not just compile it WITH the mppe feature and replace the original binary. And guess what: a guy “qvr” at Alldroid (http://alldroid.org/threads/15944-VPN-PPTP-modules-needed) just did exactly that. And now guess again: IT’S WORKING JUST FINE

Now i got vpn working at my university Thank you so much qvr!

Just go the alldroid posting and follow the link to his binaries. I just backed up the original files and replaced them with qvr’s. Everything with the Root-Explorer (of course, you need root to do that!).

Greets, Goddchen

Hey everybody,

today i read that google is introducing gesture search for american Android devices with Android 2.0 and higher. So i thought: why only for american phones?

A quick google search brought me to this xda-developers thread where “Firerat” posted the apk he got from the market.

Just download to your SD card and install as any other app. Works perfectly on my Milestone (Android 2.0.1)

Greets, Goddchen

Hey,

Motorola released a news regarding the update plans for Android 2.1 on the Motorola Droid, Milestone, Cliq and Dext.

Have a look here:

https://supportforums.motorola.com/community/manager/softwareupgrades

If this is correct we (European Milestone users) will get the update within the next six weeks. That would be really great!

Greets, Goddchen

Hey everyone,

according to  recent posts in this thread on android-hilfe.de the german update for the Milestone will be available at 13.01.2010.

This was assured by different support employees at the motorola hotline.

Anyway it’s not certain that it will be 2.0.1, it could also be 2.1 but i personally think that 2.0.1 is more likely. Also it was only spoken of the german update, but i think it will be the european wide release…

let’s hope…

Goddchen

Hey everyone…

Since i (like everyone else with a Milestone) am waiting for 2.0.1 to be released i did some analysis of the wifi traffic of the Milestone when it checks for an update.

Here is what i found out:

1: A DNS query packet: DNS    Standard query A motosb1fe.bitfone.com

2: The DNS response: DNS    Standard query response CNAME mvfe1.bitfone.com A 216.90.91.100

3: After that there is some HTTP traffic going on which seems to encapsulate SYNCML data

POST /mvpdm/dm HTTP/1.1

Accept: application/vnd.syncml.dm+wbxml

Accept-Language: en

Accept-Charset: utf-8

User-Agent: Mototorla Android Sync DM Client

Cache-Control: private

Content-Type: application/vnd.syncml.dm+wbxml

x-syncml-hmac: algorithm=MD5,username=”xxxxxxxxxxxxxxx”,mac=xxxxxxxxxxxxxxxxxxxxxxxx

Connection: Close

Content-Length: 508

Host: motosb1fe.bitfone.com:80

…j.-//SYNCML//DTD SyncML 1.2//ENmlq.1.2..r.DM/1.2..e.4C9C..[.1..nW.http://motosb1fe.bitfone.com:80/mvpdm/dm..V.Motorola…gW.IMEI:xxxxxxxxxxxxxxx..V.xxxxxxxxxxxxxxx…Z..L.20000……kFK.1..O.1201…`K.2..TgW../DevInfo/Mod…O.Milestone…TgW../DevInfo/Man…O.motorola…TgW../DevInfo/Lang…O.de_DE…TgW../DevInfo/DmV…O.1.2…TgW../DevInfo/DevId…O.IMEI:xxxxxxxxxxxxxxx….FK.3..O.xxxx..TgW../DevDetail/Ext/SystemUpdate…Z..G.xml..S.org.openmobilealliance.dm.firmwareupdate.userrequest…..O……..

And the response:

HTTP/1.1 200 OK

Date: Sat, 02 Jan 2010 13:12:48 GMT

Content-Length: 591

Content-Type: application/vnd.syncml.dm+wbxml

Connection: Close

…j.-//SYNCML//DTD SyncML 1.2//ENmlq.1.2..r.DM/1.2..e.4C9C..[.1..nW.IMEI:xxxxxxxxxxxxxxx…gW.http://motosb1fe.bitfone.com:80/mvpdm/dm..V.Motorola…a.http://motosb1fe.bitfone.com:80/mvpdm/dm?sid=xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx..NZ..G.b64..S.syncml:auth-basic…..xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx=…Z..L.xxxxx……kiK.1..\.1..L.0..J.SyncHdr..o.http://motosb1fe.bitfone.com:80/mvpdm/dm..h.IMEI:xxxxxxxxxxxxxxx..IZ..G.b64..S.syncml:auth-MAC..xxxxxxxxxxxxxxxxxxxxxxx……O.401…iK.2..\.1..L.1..J.Alert..O.401…iK.3..\.1..L.2..J.Replace..O.401…iK.4..\.1..L.3..J.Alert..O.401……

So, it seems like the host is already active (anyways, ping is blocked). But unfortunately it sais that there is no update…

Looks like we just need to wait…. longer….

Motorola what are you doing? What’s taking so long? We don’t want a major upgrade with many cool new features. We just want the currently not working stuff to be fixed, like that auto-focus bug which is really really annoying. Fortunately it will only annoy us to the 6th of january since the 24,5 days are over then. And hopefully you will get 2.0.1 ready before the next 24,5 days are over!

Gr33ts Goddchen

Hey,

hab gerade von einer coolen Aktion von druckerei.de erfahren.

Ihr könnt dort einen richtig geilen Twitter-Wandkalender gewinnen. Alles was ihr dazu tun müsst ist in dem Blog-Eintrag von oben beschrieben.

Druckerei.de möchte wissen, wieso ihr twitter-süchtig seid. Über ein paar kreative Gründe in den Kommentaren würd ich mich sehr freuen

Na dann wünsch ich euch mal viel Glück

Hey,

i finally got the bootloader ethersex working on our ATMega2561 board.

The problem was that it worked fine on our ATMega128 board, but the same software didn’t work on the ATMega2561 board. The problem was that the bootloader downloaded the firmware via tftp from a server but then didn’t start the loaded firmware, but seemed to reboot and download it again, and again, and again…

The problem was that it wasn’t able to jump to the application at address 0. Why is that? A look at the generated assembler code showed that the ATMega2561 uses EICALL, instead of ICALL (which is for chips with PC of 22 Bits). EICALL uses the Z-Register plus (!!!) the EIND byte to store the jump address (which is 0 in our case). So i just added “EIND=0;” before the jump to our application (which is at address 0) and now everything works just fine .

Hopefully this helps many other developers who try to get ethersex working on ATMega2561 /ATMega2560 / ATMega1280 / ATMega1281.

Gr33tz Goddchen

Auf Chip gibt es ein VMWare Image einer frühen Entwicklerversion von Googles neuem Betriebssystem ChromeOS zum Download.

Hab es gleich mal gezogen und in VirtualBox geschmisse. Virtuelle Maschine gestartet und dann *WTF* ?!?! Nach 3 Sekunden (!!!) begrüßt mich bereits der Login-Screen. Hier muss man sich mit seinem Google-Account einloggen und fünf weitere Sekunden später sitzt man bereits vor seinem bekannten Google Mail Bildschirm welcher mit Googles eigenem Browser Chrome geöffnet wird.

Die Startdauer dieser frühen Entwicklerversion finde ich einfach phänomenal. Allerdings muss man sagen, dass es bei der Benutzer dann doch sehr sehr langsam ist. Was jedoch keineswegs an ChromeOS liegen muss. Schließlich ist es eine Entwicklerversion und läuft dazu noch in einer virtuellen Maschine.

Mein Fazit: Äußerst vielversprechend als Betriebssystem für Netbooks und andere Geräte mit denen hauptsächlich gesurft wird.

Hey,
today i turned my computer on and got surprised by the raid bios with a red blinking “degraded” on two (?!?!) of my raid array. At first i asked my self “why are there two arrays”? My configuration is the following: i have two 500 Gigabyte Samsung S-ATA drives in a raid-1 mirror. So again, why are there two arrays?
A quick search at google directed me to this article at techbytes. It describes exactly the problem i’m having. Looking at the drive management of windows it showed a second hard drive with the same partition as my raid array but without drive letters.

It seems like the RAID controller devided my raid-1 array into two raid-1 array with one drive each. And of course therefore both were shown as degraded. So what to do now? As described in the techbytes article i had a look in the nvidia system control (the “Storage” tab). There were shown the two arrays with one drive in each array. In one array, it showed me drive letters, the other didn’t. So that was the “bad” drive. I noted it’s channel (0.2).

So what to do now? Following the techbytes article i went to the raid bios, deleted the array with the channel 0.2 drive in it. And rebuilt the correct array with the 0.2 drive.

Problem fixed…

But still i have no clue how this can happen. Seems like many other Nvidia raid user had the same issue before… Strange, strange…

Gr33tz Goddchen

Hi everybody,

on my current work with openembedded i found a problem with the resulting image.

I built a console-image for a “qemuarm” machine to test it in qemu. The image itself built just fine. Starting the resulting kernel and image in qemu was fine too. But the boot process hangs after INIT is called. It just hangs, no login-shell, no error message, nothing…

A little conversation with a guy from a mailing list i found via google showed that it was an issue with the starting order of the init scripts.

The solution is that udev has to be started after checkroot. Currently the files are /etc/rcS.d/S03udev and /etc/rcS.d/S03checkroot.

My current fix is to do:

sudo mount -o loop <yourimage> <yourmountpoint>
sudo mv <yourmountpoint>/etc/rcS.d/S03udev <yourmountpoint>/etc/rcS.d/S15udev
sudo umount <yourmountpoint>

This way, udev is started after checkroot and everything run’s just fine up to the login prompt

Another way is to change the corresponding bb file. Here you have two options:

1. Change the bb file directly
2. Use your local overlay

The file we’re talking about is “recipes/udev/udev_<version>.bb”. For me it’s udev_141.bb.

You need to change INITSCRIPT_PARAMS from “start 03 S .” to “start 15 S .”.

Now recompile, that should do the trick

Hopefully i could help…

Gr33tz Goddchen