First, what is livestats? It’s the end point of this Peepcode tutorial video.

Second, why the completely convoluted setup? If you watch the tutorial (it’s only $9 and I highly recommend it), you’ll see that they set up their Faye/Node server on port 8000. I want my setup to work on port 80 so that most firewalls won’t block access to it (including mine at the office). I also have a bunch of other websites running on port 80 (including the geoip site). I need to be able to split up access to each of these based on host names in the request headers (ergo, livestats.gmtaz.com takes you to my node.js app and geopip.gmtaz.com takes you to my geoip app – both of which are running on port 80).

Ah, so what to do, what to do…

First step was to get Node.js running. It works on linux so I had to get a linux distro, hopefully Ubuntu or something similar (ended up going with Mint), and some kind of virtualization software to run it (VirtualBox). Both of these products are free.

Actually, now that I think about it, all the software I used is free. Awesome.

Step 1.A: Get Node.js installed.
Step 1.B: Get nginx installed. “sudo apt install nginx”.

Ok, easy enough. Next step is to complete the tutorial and get the node.js app running. Go do it.

…a few hours later…

So now, you should have your node app running on localhost:8000 without issue. Great! Wait, not great. I(you) want to access the node app on port 80. Ok, well, let’s configure nginx to proxy over port 80 to port 8000.

default.conf

server {
listen 80;
server_name example.com *.example.com localhost;

}


proxy.conf

proxy_redirect off;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
client_max_body_size 10m;
client_body_buffer_size 128k;
proxy_connect_timeout 6000;
proxy_send_timeout 6000;
proxy_read_timeout 6000;
proxy_buffer_size 4k;
proxy_buffers 4 32k;
proxy_busy_buffers_size 64k;
proxy_temp_file_write_size 64k;
send_timeout 6000;
proxy_buffering off;
proxy_next_upstream error;


Check out the nginx wiki for some good tutorials and sample code.

Now, when I go to http://localhost:80, nginx loads my node app which is running on port 8000.
You can configure multiple server objects in the config. Took me a few minutes to figure that out. This means you can have multiple name based proxies.

Ok, this is great. Now I point my home firewall to forward port 80 to my virtual server’s IP (which should be bridged to your network, not NAT *muy importante*). This allows me to go to my public ip and get routed to my virtual server. Just make sure that in the nginx config you are specifying the server name as the public IP or the DNS friendly name instead of example.com.

This is where the trickiness comes into play. I need a geoip server. I can get one running on apache and php. I used the pure PHP module, but feel free to use the module for Apache. This brings me to my my Apache config. *Awesome segway, I know*

I have apache listening to port 80, which means it can’t be running on the linux VM (you can only bind a port to an app once). Installing Apache 2.2.16 on the host machine (running Windows 7) was my next step. What about IIS? Well, I could use my IIS setup on Windows 7 too but I initially used Apache for my proxy server. Yeah, yeah, now I’m doing a double proxy. Totally unnecessary, but you’ll see why I did it in a minute. Anywho, I have my Apache config set up to run multiple virtual hosts. One vhost proxies out my http request to my IIS server, which then serves up the site in IIS. The other vhost serves up the geoip app.

Another thing to be aware of is that I needed 2 different versions of PHP installed on my computer because Apache 2.2.16 requires a different version (VC9) than my IIS setup (VC6). Also, IIS is configured to hit sql and mysql servers for different websites that I have running, and I was getting errors in php5ts.dll when enabling those modules in PHP 5.3.3 on Windows.

Quick summary of the situation so far:

Windows 7 (HOST):
Apache 2.2.16 with PHP 5.3.3 VC9
– Apache is set up to proxy requests based on host names to IIS or to geoIP.
– In case I decide to turn off the virtual machine, I still want to be able to host all my other websites.
IIS 7.5 with PHP 5.3.3 VC6 (PORT 8000)
– IIS is serving up a few other website that are developed in ASP.net C# and PHP (This is why I need IIS )

Linux Mint (VM)
Nginx
– Proxy server for node.js running on port 8000 (in case I want to run multiple node apps – each one must be on a different port, which is why I need the proxy).
Node.js
– Hosting the livestats app on port 8000.

Got that? Ok. The next part stumped me for most of the day. My node.js via proxy worked fine in Firefox, but not in Chrome. WTF?? Let me rephrase. Chrome worked when going to the direct IP and port of the node server which means that the proxy was causing problems. I’ll save everyone a lot of time by saying this: websockets via proxy don’t seem to work too well.
Basically, Faye was detecting that the Chrome browser can use websockets and was attempting to create the connection using that (websockets). Firefox was using the long-polling method. How did I figure this out? Long story but it had something to do with looking at the Faye object in Chrome inspector and seeing that it was rocking the websocket, then looking at the same Faye message object in Firebug and seeing that it was using long-polling. Ok, I guess it wasn’t that long of a story…

So now I know that Faye, Node, websockets and proxies don’t mix all too well. Turns out there was a simple solution. Force Faye to use long-polling! What do you mean “it’s not an option”?

Check out Faye-node.js in the tutorial code. In there you’ll see an array for CONNECTION_TYPES. Just replace “websocket” with “long-polling”. Amazingly enough, this works.

So now you should have a pretty complicated network of proxies and web servers running but everything should be working fine! Magic? Nay. Just a lot of tinkering and coffee.

Things that made development and testing easier:

1. Running the node.js dev server so that it automatically restarts when a file is changed
2. Installing Dropbox on all servers and running the node site from the application folder within Dropbox
3. Firebug and Chrome’s Inspector
4. Espresso on the mac for editing the app within the Dropbox folder
5. VirtualBox and Linux Mint

These things made it incredibly easy to make changes and verify their functionality with a simple refresh.

EDITORS NOTE: Since writing this, I realized that for some reason the proxy isn’t working all that well in nginx for me ATM. I’m using apache to proxy to nginx for now and not really using nginx at all. The nginx proxy from vm to host works fine, but the proxy from nginx to node doesn’t work…

Something to do tomorrow. PEACE!

EDITORS NOTE 2: I figured out the proxy thing. Now I’m just using nginx for proxy services. Turns out I had statically set livestats.gmtaz.com in my hosts file on windows to 127.0.0.1. Forgot about that. Yay for testing!

I’m going to assume that you already have a valid, legal copy of Windows 7 Enterprise and an installed copy of Windows 7 RC Ultimate and they are both the same kernel (either 32bit or 64bit) and that you already backed up all your data.

Within Windows Ultimate:

1. Open up the Registry Editor (Start > Run > regedit)
2. Go to HKEY_LOCAL_MACHINE\Software\Microsoft\Windows NT\Current Version as seen in the screenshot below
   
3. Replace the word “Ultimate” with “Enterprise” in ProductName and EditionID.  You’re done modifying Windows 7 RC.

Now comes the (kind of) tricky part.
See Here (howtogeek.com) for another article describing this process and here (blogs.msdn.com)

1. Take the Windows 7 Enterprise DVD and create an ISO image.  You will need to open the ISO and modify a file, more specifically, cversion.ini.
2. Once you have the ISO image, using an ISO image editor (I used PowerISO), and browse the DVD image.  Locate \sources\cversion.ini and extract it to your desktop, then open it with notepad.
3. Change the winclient value to 7100, save and close.(winclient=7100)
4. Replace the file in the ISO with the file you just edited and burn the ISO to a DVD.
5. Start the installer from within Windows.
6. Be cooler than the next guy.

That’s it.  Enjoy what is, arguably, the best version of Windows yet.  I’m assuming that this registry hack will work for versions other than Enterprise although I have not tried any other versions myself.

reg add "HKLM\SYSTEM\CurrentControlSet\Control\Terminal Server" /v
 fDenyTSConnections /t REG_DWORD /d 0 /f

This will make Terminal Services on the remote computer allow for incoming connections. Then, for Windows 7, type in:

net start "Remote Desktop Services"

This will start the listener service on the computer. You should now be able to access the computer using the administrators account (granted, you must have admin access to the ssh or telnet terminal to do any of this).

EDIT: One person commented about not being able to set up a different wallpaper for each space on OSX. Check out Spacesuit. This looks like a pretty good solution. Let me know if this helps out!

Click on a link below to download:

Brief: Dropbox is an application that runs in the background on your computer (mac, linux, windows) and automatically syncs files when connected to the internet.  If you’re doing web dev stuff, chances are you are connected online.  Dropbox also creates a local folder on each computer to store the files that you are working on, etc.

Enter brain power:

1. Set up Apache on all computers since it’s a very powerful web server that runs on the previously mentioned operating systems. Here’s a guide on installing Apache on Windows.
2. Set up the root directory in Apache to be the ~/user/Sites folder and create a sym-link (Windows Vista) to your dropbox directory. This way you can run multiple websites (some that sync and some that don’t).
   To take this a step further, you could put all of this config info in a separate .config file (I called mine, “dropbox.config”) and drop it in the Dropbox directory for automatic synchronization throughout your systems.  This works best if the operating systems are the same or the root structure is the same.
3. So now, when you browse http://localhost, you should see this:
   
4. Verify that you see the same directory structure on all computers.  If this is the case, then you’re done.  Just start by creating PHP site within the Dropbox folder (I created a subfolder called WWW and in Apache I create a call to the virtual directory /dropbox which takes me to the WWW folder) <= This is important if you want to use Dropbox for other things!  This way, when you click on the Dropbox link listed under (~/Sites) it will redirect you to the dropbox/WWW folder.

Now, when you are writing code, editing images, or anything else, the code will automatically synchronize to each workstastion, allowing you to continue right where you left off, anywhere.

An example of how this helps:  I was recently working on writing some Ubiquity commands so I set up a Ubiquity folder within Dropbox, subscribed to the [ubiquitycommand].js file in that folder on every computer I work on, and that was it!  Now, whenever I make a change to those commands, the changes get synched to each computer and because Ubiquity is subscribed to the command on localhost, that change affects all computers.

This method also works extremely well when you have different operating system specific browsers that you would like to test with.  It may seem a little daunting at first, but let me tell you, once it is set up, you’re done.  No need to reconfigure anything.  Best of all, Dropbox is FREE!

1. Get Quicksilver if you dont already have it.  So many good things come from this app that it’s worth it if not for this, then for it’s pure awesomeness.
2. Enable fast user switching (System Preferences>Accounts>Login Options-last check box)
3. Go into Quicksilver Preferences (default is option+space>command+,) and make sure you have the Extra Scripts and User Accounts Module enabled under plugins.
4. Open the Quicksilver interface (option+space) and select “Fast Logout”, press ‘Tab’ then ‘right arrow’ to open up the command selections and select ‘Run: run a shell script’ and press enter.  This should bring you to the user login window.
5. If that worked, go back into Quicksilver Preferences (option+space>command+,) and select ‘Triggers’.
6. Add a new trigger by pressing the ‘+’ symbol at the bottom. Select ‘Add new hotkey’.  You should now see the default as the last command you picked (Fast Logout and ‘Run’).
7. Click Save
8. Now click on the ‘Trigger’ (it should say ‘none’) and select the hot key you wish to use.  I chose to use ‘command+option+L’ as command+L has other functions.

That’s it.  You can now lock your Mac to your heart’s content with a simple keystroke.

So far I’ve written some custom commands to search my brother-in-law’s photogallery site, some wordpress blogs, and I’m currently working on a quicknotes plugin.  When these are done, you’ll probably be able to find them under Projects.  Speaking of, I really need to update that page… I haven’t worked on either this site or the Rome in a Day site in the longest time.

Anywho, enjoy Ubiquity.  Hopefully it will really take off.

Links:

Free Music Zilla

Songerize

First, a warning: After hacking my phone, I get periodic application crashes. In example, the iPod app crashes every so often. Not while playing music, but just after initial startup when it displays all my playlists. It takes about a second or two and if I click on a play list, the crash doesn’t happen. This also isn’t a “hard crash” per say. The app just closes. Safari was doing the same thing when browsing certain community websites. As annoying as this was, I was more concerned about the permanence of the situation. It seems that a soft reset solved the problem. If it happens again, I’ll post something and let you know if the reset continues to be a work around.

[Edit: The cause of the crashes is summerboard.  Uninstalling summerboard solved this problem]

Ok, now for a couple of tips/instructions.

1. Back up the phone before doing this. I did, thankfully, because I had to perform a restore before the hack went through.
2. After backing up the phone, perform a restore bringing it back to it’s initial 1.1.1
3. Connect to wifi network. Don’t do this over EDGE. It’s too slow. I tried it and had to try a restore before doing so.
4. This one is probably the most important and most overlooked: turn off the auto lock feature on the phone.
5. Hack the phone by going to http://jailbreakme.com/ and following the instructions.
6. Some sites say that you may have to power cycle the phone after the jailbreak is done. I highly recommend this. Computers have to reboot after an upgrade
7. Now perform a sync and everything should be back to normal. I got all of my settings back including my GMAIL settings and texts. Oh, and I’m syncing to a PC.

After all this is done, I highly recommend installing SummerBoard with the Icarus theme. Then disable the transparent status bar and wallpaper. They’re hideous. I hope this helps some people. Also, if you’re worried about hacking the phone and then another firmware update breaking or ‘bricking’ your phone… I wouldn’t worry too much, unless you hacked the SIM. As much as I hate at&t and love the iPhone, I don’t recommend unlocking the SIM. This will most likely end up bricking your phone. If you break your shit, don’t blame me. This is what I did to get hack my phone and it works. Try at your own risk*

*Standard disclaimer. I don’t want to get sued.
**I’m using the ugly kid photo again cause I think it’s funny. And apparently there are 3 degrees of separation from me to this guy… weird.