TangoRangers.com's Blog

Almost two years ago I wrote a how to on tethering with my G1 and Slackware, located here. This was back before I rooted my phone to breathe more life back into it. I still have the phone and plan on having it for a long time. Sure it is as slow as can be, but I still love the full qwerty keyboard… oops, sorry, I’m not here to talk about that. Anyways, like I was saying. I rooted my phone and I’m currently running CyanogenMod 6.0 on my G1.

Wireless Tethering allows you to connect your computer through the phone and use the 3G (or 4G, EDGE, etc) as your internet connection. Some new phones can do wireless tether without rooting, but often times you must pay extra for the service. If you rooted your phone there are two ways of doing wireless tether. First check and see if you have wireless tethering built into your MOD. Go to Settings –> Wireless & networks –> Tethering. If you only see USB tethering then there are extra steps you must take to enable wireless tethering. See the 3 images below. Note that my phone only lists USB Tethering. If your phone is the same, then continue.
  

Some people I know are also running CyanogenMod and they have wireless tether built in. Why is that? For starters, since the phone is newer than my G1, they have a more up to date wireless card. These cards can be put into AP mode. This is the same style mode that your wireless router at home uses. Some phones that have wireless tether can also put the card into Ad-Hoc mode. This mode is also known as Computer to Computer without an Access Point. However, it seems that you still need the extra software to do wireless tethering. It is kind of hard for me to explain as I don’t have access to many phone to see how they work. Let alone ones that are rooted. With my G1, I can say that the wireless card was not even designed to operate in Ad-Hoc mode. Thanks to the software I’m running I can force the card into Ad-Hoc mode and connect to it with my laptop.

Keep in mind that wireless tethering will kill your battery quickly. Be sure to have a USB charging cable with you as your battery will not last long. Also note that many service providers do not want you to tether without paying extra. Be sure to read the terms of service of your provider before continuing.

If you made it this far you are going to need extra software. Once again, if your phone is not rooted this will NOT work! Head on over to Google Code to download. At the time of this writing I am running version 2.0.2. A later version has been released, but I have not tested it as mine works very nicely. Download the software to your phone and install. If you don’t know how, I can’t help you.

After installing there are several features to the software. I highly recommend encryption and if you want, Access Control. By default the software uses wpa_supplicant to set the WEP key. This is done via a passphrase. If you like the shell (like I do) you will need to use iwconfig to set the key. Below are some pictures of the settings menu in Wireless Tether.
  
  

If you use Access Control, before a newly connected device can access the net you will have to allow it. I don’t have pictures, but if I remember it will be in the notification bar. Just drag it down, select the computer/device and hit allow. Later you can look in the Access Control Panel by hitting Menu then Access Control. Here is what mine looks like (and no, the computer is not actually connected. I just have it set to always allow the machines based on it’s MAC address.

My laptop has a Broadcom wireless card. I’m using the b43 drivers found here. As it turns out there are several problems found with my current kernel version (2.6.28) and the rc.wireless script that Slackware uses. There are two fixes. One is to remove the b43 module and reload it (this is so we can put the card in Ad-Hoc mode) and I have had problems setting the WEP key via iwconfig. With iwconfig you can set the key (default for this software being ‘abcdefghijklm’) by typing:

iwconfig <card-id> key s:<passphrase>

I always got an error. I solved the problem by using another computer to set the key then you can type iwconfig again and see the key. In this case, the key is 6162636465666768696A6B6C6D. If needs be, write down this key so you can use iwconfig to set it again.

Ok. I hope by now you have the wireless software installed on your Android phone and it is running. Note: I have found that sometimes even when encryption is set it doesn’t turn on. Be sure to scan for wireless networks to see your phone and if encryption is turned on or not. Also note that the software has created an Ad-Hoc style connection. If everything seems to be working here is what you do. In the next steps I will have some bash commands, then a note on what each command does. You may need a make changes to it to suit your needs. I turned it into a script I can run on my laptop so I don’t have to type every line in every time I tether.

rmmod b43   -unloads the b43 module
modprobe b43   -reloads the b43 module
ifconfig wlan0 down   -take the card offline
iwconfig wlan0 mode ad-hoc   -set to Ad-Hoc mode
iwconfig wlan0 essid android   -set the essid (be sure to check your configs)
iwconfig wlan0 key 6162636465666768696A6B6C6D   -set the WEP key (may differ!)
ifconfig wlan0 up   -bring the card online
dhclient wlan0   -get an IP address to your phone

Now type ifconfig and make sure you have an IP. If so, try surfing! If you are a hardcore Slackware user you may note that I use dhclient instead of dhcpcd. For some reason dhcpcd never gets me an IP. If this doesn’t work try it again. Sometimes I have to run the script twice, but I always get it working eventually.

I don’t have a problem with Windows 7 on the same machine. Windows will allow you to type in the paraphrase and it works great. If you’re running Linux and using the b43 module there are several things you can try to fix any problems. I hear the 2.6.33 and newer kernel have these problems fixed. Also if you don’t run a wireless script on boot, that may help. Most newer Linux distributions have GUI programs to help you connect to networks. I’m still running an older version of Slackware, and at the time, it didn’t come with those programs.

Now for some more pictures I took of the program. No reason. I just thought I would share.
  
  

I hope this all makes sense. If not you can always ask, if I know the answer I will reply. I would also recommend asking on a help forum or asking Google. Good luck and Happy Hacking!

After trying to get my Android (aka, T-Mobile’s G1) working on my Slackware laptop and having many problems I looked for help at androidcommuinity.com. Even tho they didn’t technically help me solve the problem they were still a great help, and there quick replies kept me thinking of what the problem could be, and I thank them for that. In the end it was my own fault. Here’s the story.

My laptop runs Slackware 12.0 with a new custom 2.6.28 kernel and there was were the problem was. As it turns out I failed to turn on and modularize certain USB functions, now I’m unsure of what I enabled that made it work.

First, if you think it’s a problem with your kernel check your USB device settings. I use the make menuconfig when I build kernels, if you use xconfig it may appear slightly differently.

Check under Device Drivers –> USB Support. I turned on just about everything and modularized just about every USB device, don’t forget to turn on USB Gadget Support I turned on the top 3 options, they are for debugging and also Serial Gadget was modularized. Click Here to download my .config file. This file is in the root of where you extracted your kernel (usually /usr/src/linux-2.6.28/). Be sure to rename the file from DaijoubuKun.Android.config to .config otherwise you will need to tell your kernel to use that file. Quick note: I know I have a lot here that does not need to be turned on, I like large kernels with lots of modules.

UPDATE!
Put your 50-android.rules file in /etc/udev/rules/ and chmod it to 755!
Also in your kernel build make sure you turn on USB debugging mode!!!

If you think your kernel is OK check out graha.ms. There is a lot of the stuff your going to need to know there.

Now, at this point I’m assuming you visited the link above but your Tetherbot still isn’t working. This may be because the commands google gives may not work in Slackware. I wrote 4 lines more designed in Slackware’s udev style. Below are 8 lines. I’m sure you only need one, but it didn’t hurt me none to use all 8, the bottom 4 are mine. Here is my 50-android.rules

SUBSYSTEM=="usb", SYSFS{idVendor}=="0bb4", MODE="0666"
SUBSYSTEM=="usb_device", SYSFS{idVendor}=="0bb4", MODE="0666"
SUBSYSTEM=="usb", SYSFS{idVendor}=="0bb4", MODE="660"
SUBSYSTEM=="usb_device", SYSFS{idVendor}=="0bb4", MODE="660"
SUBSYSTEM=="usb", ENV{DEVTYPE}=="usb", SYSFS{idVendor}=="0bb4", MODE="0666"
SUBSYSTEM=="usb", ENV{DEVTYPE}=="usb_device", SYSFS{idVendor}=="0bb4", MODE="0666"
SUBSYSTEM=="usb_device", ENV{DEVTYPE}=="usb", SYSFS{idVendor}=="0bb4", MODE="0666"
SUBSYSTEM=="usb_device", ENV{DEVTYPE}=="usb_device", SYSFS{idVendor}=="0bb4", MODE="0666"

I know it’s got to be overkill to have 8 lines in there, so if you figure out which is the one that works post it in a comment and I will edit this to show that line with credit to you.

NOTE: When you plug in your phone you will be asked if you want to mount the sdcard to your computer. SELECT NO! According to rynosaur of androidcommunity.com.

Note, to make things easier copy the files is your tools/ directory (from the android sdk) to /bin. This way you do not need to go to the download directory to run the adb program. You can also copy the libs to /bin/lib

Now run adb devices. You should get some out put like

List of devices attached
HTxxxGZxxxxx    device

If you do then we are ready to go!
If not did you remember to run on USB debugging?
Start the Tetherbot program and hit Start Socks
Now type ‘adb forward tcp:1080 tcp:1080‘. This will forward and thing you send to ‘localhost:1080′ to the phone.

Using nmap I found that Tetherbot uses socks 5 with no authentication. This makes things a bit easier. Once you have gotten this far you are ready to setup your web browser to use the socks proxy. I have done this in SeaMonkey and Firefox (haven’t tried anything else yet) so here are the instructions for FireFox 3.0.x (should stay about the same for future versions).

Go to Edit –> Preferences –> Advanced (top tab) –> Network (lower tab) –> Connection Settings (button).
Select ‘Manual Proxy configuration‘ and put localhost in SOCKS Host and port 1080 in port. ONLY put that information in the SOCKS Host! Do NOT put it in HTTP Proxy! Be sure to select Socks v5 and (this may not be necessary) remove anything in the field No Proxy for:
Here is a picture:

Once you do that everything will go through your phone. I did read that there is a 1GB per month limit on the phone. I do not know if this is true or not. I’m only going to use Tetherbot when I’m doing a service call, I need a file and there is no other way to get online.

Lastly, just as quick bit of info. I live in an area where T-Mobile doesn’t have 3G! It sucks, but I’m told it’s coming later this year. It damn well better with how much I pay every month for this thing. I ran a bandwidth test through the EDGE network (it’s like 2.5G) and below is my speed.

Granted, I only ran the test once, and early in the morning. Plus I don’t think the phone was designed to have these sort of things run through it, but it will have to do for now. P.S. Thanks for the awesome speed test program Speakeasy.

Special note: I could be wrong, but it seems that after you put the required information into 50-android.rules you may need to either restart or run /etc/rc.d/rc.udev restart to restart the udev service. You may also need to replug in your phone, make sure you don’t mount your sd card as it may cause tetherbot to not function properly.

If this helped you in anyway please let me know. If it didn’t help post your problem and I will try to help.

If your like me you have scoured the internet trying to find a simple php script that will let a user upload an image and then rename the image to something unique. Well I have just the solution.

First, the only real requirement is PHP and a web server (I use Apache).

Create a new file, lets call it ‘image_upload.php’
At the bottom of the page we are going to put in a basic form to give a browse and upload button.

<form action="./image_upload.php" enctype="multipart/form-data" method="post">
Select Image to upload
<input name="thefile" type="file" />
<input name="submit" type="submit" value="upload" />
</form>

The form name isn’t important. The action is. That is the name of the file that needs to be accessed when the upload button is clicked.
Input type is very important, and so is the name.

Now that we have a starting point go to the top of the file. Just to make it easy I’m going to paste the entire code here then explain it part by part.

if(isset($_POST['submit']))
{
 
	$tmp_name = $_FILES['thefile']['tmp_name'];
	$newarray = explode( ".", $_FILES['thefile']['name']);
	$thecount = count($newarray);
	$fileprefix = time() . "." . $newarray[$thecount - 1];
	move_uploaded_file($tmp_name, "../img/$fileprefix");
	exit();
}

To start the if statement at the top is to ensure that the ‘submit’ button was hit before running the script.
The ‘$tmp_name‘ variable is used to get rid of that damn array.
$newarray takes the file’s real name (see note below) and explodes the array into several smaller arrays by the ‘.’. For example, if the file uploaded it called ‘best.friends.pic.jpg’ it breaks that name up to ‘best’, ‘friends’, ‘pic’, and ‘jpg’. Trust me, this is important (If you don’t understand how arrays work I recommend doing some research and learning).
The next line I use the count function. This will help in the event there is more than one ‘.’ in the name of the file being uploaded.
Now to create the new name of the file. $fileprefix equals the time of the upload. This works well, it will grab the time from the server and use it for the file name. Then it adds a ‘.’, then adds the suffix of the file (example: jpg)
Time to move the uploaded file. In my example I added the directory before the newly created file name. This will go back 1 directory and then to ‘img/‘.
Lastly, we use the ‘exit();‘ function. This is so that after the file has been uploaded the script stops. If you remove the function it will display the upload form again.

NOTE: There is a reason (altho I don’t know it) on why there are 2 arrays. It seems to me that the first ($_FILES['thefile']['tmp_name']) holds 2 parts, a pointer for ‘thefile‘ and the temp name php gives files before we do something with it, and the second ($_FILES['thefile']['name']) contains the original name of the file.

btw, make sure apache has full access to the folder you are uploading images to.
For example, if this is on your own server check /etc/apache.conf (or httpd.conf) for the user and group apache runs under. This is usually either ‘nobody‘ or ‘apache‘.
If you have SSH into your server (or are sitting infront of it) goto the directory your files are being sent to and type

chown apache.apache img/ -R

This will give both user and group ‘apache‘ access to write files.

p.s. I will be creating another script here that deals with multiple picture uploads.

Here are my instructions on how to get VNC in KDE 3.5+ working through an SSH tunnel. It’s easier than you might think.

To start all you need is 2 or 3 linux machines with OpenSSH installed. Most distros come with it (altho I know Ubuntu does not).
NOTE: All my machines run Slackware 12.0 or higher.

Step 1 – Setup the host.
This is fairly simple, open up you Control Center, and find Desktop Sharing. Just look at my picture below and see the settings I would recommend for this.

Just make sure you set a STRONG password!

Now comes the fun part. Creating the SSH tunnel. By default the VNC connection is on port 5900.
For this example you have 2 computers. Your at a coffee shop with free wifi but your smarter than everyone else, so your going to use encryption to your home desktop and surf the internet from there.
Your home computer (lets say) has a domain name. For my examples it will be daijoubu.net, and your internal computer is 192.168.1.2.
Make sure you set your router to forward port 22 (the SSH default) to 192.168.1.2
Open up a terminal (some times it’s called Konsole) and type:

ssh dkun@daijoubu.net -L 5931:localhost:5900

The user name I’m using is dkun, just put in your user name
You will be prompted for your password, after entered you have formed the SSH connection. What this command does is it takes all traffic from your desktop port 5900, and forwards it to your laptop (at the coffee shop) to localhost port 5915.
Seems complicated, but trust me, it works!
Now open up Krcd and type

vnc:/localhost:5915

Just as shown below.

If you have 3 computers. For example, you don’t forward to your desktop (for security reasons) but you do forward to a file server. Lets say your file server is 192.168.1.3 and your desktop if 192.168.1.2 type:

ssh dkun@daijoubu.net -L 5915:192.168.1.2:5900

This will form the SSH tunnel to your server (192.168.1.3) then forward port 5915 from 192.168.1.2 through the SSH tunnel back to you.
Reminder: Doing it this was results in plain text from 192.168.1.3 to 192.168.1.2. This is only a problem if you don’t trust your internal network!

From here is gets simple, after you click Connect you will be prompted for the following window.

These are the settings I recommend for over the Internet, VNC can take a lot of bandwidth.
Next you will get a password prompt, type in your password and hit OK

Your remote desktop will appear! if you look quickly you will see this at the bottom right of the screen

That’s it! Now you can use your remote desktop over a secure connection!

WARNINGS! If you attempt a connection without the SSH tunnel your passwords will be sent in plain text! That is BAD!
DO NOT FORWARD PORT 5900 ON YOUR ROUTER!

Extra Notes: If you do not have a domain name to work off of, you can put in an IP address after the username@, make sure it’s an internet IP address, 192.168.1.2 will NOT work
If you don’t have a static IP address you can use dyndns to get you one. They are really good, but if your IP changes you will have to update your opendns account. I would recommend checking before you head out.

Special Thanks to Spyder_3lite of UCoD.com. If it weren’t for you showing me something way cool with SSH, I never would have been able to do this.

Note: This was originally written on my other site Daijoubu.net. I have moved it here for better indexing from Google. ^_^