Fortunately it’s pretty easy to configure the default template so you don’t have to do this.

1. Open Inkscape
2. Set up your blank sheet however you like
3. Save the file in /home/username/.config/inkscape/templates/default.svg

That’s it!

You know when you hear something and it sends ideas pinging off in different directions like a particle collision? Reading about this gave me one of those moments.

Think about any scenario where you’d want real-time location specific information pushed to you. There’s the obvious stuff, like retailers notifying customers of up-to-the minute promotions and offers. Imagine getting a message from the Deli round the corner, telling you that there’s only two peri-peri chicken baguettes left?

This could really take off for events organisers. Wouldn’t it be great getting info about bands in the signing tent right now? But more importantly not getting the info if you’re on the other side of the site?

But what about calling for help when you need it? In disaster situations, spreading information is essential. You can be told, not where all the heath centres are, but the nearest one to you, right now. This could make managing resources much easier. Messages directing people to the better equipped centres can be sent out over a larger area than the less equipped ones.

These are a few thoughts from a few minutes thinking. My brain’s going to be chewing on this for a while…

When using the following code, inside a widget, to return posts from a custom query, you’ll find that the post title and perma-link are returned, but the post ID is not.

<?php $loop = new WP_Query($query);
while ( $loop->have_posts() ) : $loop->the_post(); ?>
	<a href="<?php the_permalink() ?>"><?php echo $post->ID.' : @; the_title(); ?></a>
<?php endwhile; ?>

Adding a global call to the $post variable at the top of the loop fixes this issue.

<?php $loop = new WP_Query($query);
while ( $loop->have_posts() ) : $loop->the_post();
	global $post;?>
	<a href="<?php the_permalink() ?>"><?php echo $post->ID.' : @; the_title(); ?></a>
<?php endwhile; ?>

Lets start with our two objects, a blobby object and a rectangle below it.

Crate a clone of the rectangle (Alt+C), the cloned shape will appear on top of the others. A cloned object is linked to the original, so any changes you make to the original will also happen to the clone.

Select the cloned rectangle and the blob object (Shift + Click to multi-select). Then clip the blobs with the clone (from the menu select Object > Clip > Set). This will only show the portion of the blob object that is covered by the cloned rectangle, the cloned rectangle itself will become invisible.

Now you have the pattern displayed on the shape. But more than that, you can resize the rectangle and the pattern will expand to fit it!

TaDa!!!

If, like me, you’ve never known which colour space you were working in, chances are it was RGB. CMYK is an alternative, commonly used in the printing industry. I’ll spare you further details, but if you want to know more, you know where to go.

Support for CMYK in Inkscape is ongoing, as is the discussion surrounding it. Click Click Click But we cans till configure Inkscape for CMYK to make things easier further down the road.

Before I go on, a disclaimer: I am in no way an expert in Inkscape nor CMYK colour profiles. These settings seemed to work for me, they may also work for you. I am more than happy to accept comments & criticisms in the comments below.

Download the Adobe Color Profiles and copy them all in to /usr/share/color/icc
http://www.adobe.com/digitalimag/adobergb.html

Follow the instructions on this page to generate an ICC profile for your monitor
http://www.ubuntufieldmanual.com/?q=node/38

Copy the icc file in to one of the locations that Inkscape searches
/home/username/.local/share/color/icc
/usr/share/gnome/color/icc
/usr/local/share/color/icc
/usr/share/color/icc (best for all users)

Open/restart Inkscape
Go to Inkscape preferences (Shift+Ctrl+P)
In ‘Color management’
Set ‘Display Profile’ to your monitor profile (lcms RGB Virtual Profile)
Tick ‘Simulate output on screen’
Tick ‘Mark out of gamut colors’
Select a highly visible color for the warning color (I prefer a bright green)
Set ‘Device profile’ to an apropriate output profile for your region (US Euroscale etc) and print medium (Coated, Uncoated etc)

Open the Document Properties (Shift+Ctrl+D)
Click the Color Management tab
Link the ‘lcms RGB Virtual Profile’

Now, if you enable Color Managed View (in the view menu), colours that are out of the printers range will show up on the page. You’ll also be looking at colours more-or-less as they come out the printer.

Working with Colour and Gradients

If you’ve never touched Inkscape before, this tutorial is a total must.
http://www.kendraschaefer.com/2009/06/inkscape-tutorial-color-and-gradients-for-absolute-beginners/

Spiro Swirls

A new feature in the latest version, this video tutorial (no sound) shows you haw to create killer spirals. Push the algorithm to the edge to see what it comes up with.
http://verysimpledesigns.com/vectors/inkscape-tutorial-spiro-swirls.html

Text Styling

This tutorial runs through the great range of control that Inkscape gives you with text, and shows you how to make a great page title.
http://troy-sobotka.blogspot.com/2008/04/inkscape-tutorial-2-text-and-simple.html

Creating Icons

This tutorial takes you through the process of creating a Tango style icon. Even if you’re not interested in creating icons, there’s a lot of good tips about working with paths in here.
http://vector.tutsplus.com/tutorials/illustration/how-to-create-a-vector-light-bulb-icon-with-inkscape/

I found all these tutorials through the Inkscape Tutorials Blog, jump on in!

Click Here To Download XOR.zip

File neuralNetwork.h

This is the file that handles the datatypes. The model I’ve used here represents the neuron output values as a float array, the connections between neurons as a 2D boolean array and the connection weights as a 2D float array. To work with the data, the class contains the following functions.

• neuralNetwork() – Default constructor, creates network with 10 neurons.
• neuralNetwork(int n) – Constructor creates network with n neurons.
• ~neuralNetwork() – Destructor, cleans up data objects.
• getSize() – Returns an integer representing the number of neurons in the network.
• setSize(int n) – Creates the data structures for a network of size n. Calls to this function are not required if using the neuralNetwork() constructor.
• zeroData() - Sets all the values of data in the network to zero, effectively clearing the network. Calls to this function are not required unless resetting the network at run-time.
• setVal(int n, float val) – Set’s that value of neuron n to val.
• getVal(int n) - Gets that vlue of neuron n.
• addConn(int i, int j) – Creates a connection in the network from neuron i to neuron j.
• addConn(int i, int j, float weight) – Creates a connection in the network from neuron i to neuron j with weight weight.
• delConn(int i, int j) – Removes the connection between neuron i and neuron j.
• is Conn(int i, int j) – Returns a boolean, true if there is a connection from neuron i to neuron j.
• setWeight(int i,  int j,  float weight) – If there is a connection from neuron i to neuron j, the weight is set to weight.
• getWeight(int i, int j) – Returns a float representing the weight of the conneciton from neuron i to neuron j.
• printVals() – Prints to screen the data in the output value array
• printConns() - Prints to screen the data in the connections array
• printWeights() – Prints to screen the data in the weights array
• printAll() – Prints to screen all of the data, with headings.

File nnXOR.cpp

This file uses the neuralNetwork class as an example to create a neural network to compute the XOR relationship.

At the top of the file, the class nnXOR is created extending the neuralNetwork class, it contains the following functions:

• nnXOR(int n) – Constructor, creates the network of size n, note the use of setSize().
• updateNeuron(int n) – The update process for neuron n. This process multiplies each input neuron value by the conneciton weight, sums these values then performs a simple threashold function to determine the output value of neuron n.
• update(int epoch) – Sequentially updates each neuron in the network epoch times.

The bottom half of the file contains the main method, calculating the XOR relationship between variables a and b.

1. Set the variables a and b from the command line, if no data was supplied from the command line, prompt the user.
2. Create the neural network object with 5 neurons.
3. Create connections and set weights between neurons in the network.
4. Set the input neuron values to a and b. Values can be either 1 or 0.
5. Print the initial state of the network.
6. Update every neuron in the network once.
7. Print the final state of the network.
8. Print the XOR result.

Feel free to use and abuse as you wish.

MyLineup is a simple app for Android phones which allows you to keep track of the artists you want to see at all the festivals your going to this summer.

Please note, this is still in (very early) beta, so if you encounter any bugs, let me know in the comments below.

To install, visit this page on your android phone and click the link below.

Click Here To Download myLineup

I’ve been trying to writea script to test authorization against a list of servers, and fire off an email if any of the servers doesn’t respond. The lis tof servers is in a text file, and I’ve got a while loop to itterate through, testing line by line. If any of the tests fails then, I’d change a flag variable from its default value of “FALSE” to “TRUE”. After the loop I can test the value of the flag variable, and send an email accordingly.

The problem I was having is that the variable scope didn’t seem to extend beyond the while loop, so the final check always showed that the flag was set to FALSE, even if it was changed within the loop.

My original script looked something like this…

SENDMAIL= grep "^server:" servers.file | while read line; do
    if [ testauth() ]; then
        SENDMAIL="TRUE"
    fi
done

if [ SENDMAIL ]; then
    sendMail()
fi

The problem is that, I’ve piped my file through grep and in to the while loop. Because of the way the shell handles piped commands, the variable scope is limited to within the while loop. To get around this problem, you have to remove the pipe in to the while loop. In this case I chose to output to a temp file then read it in, but you can also pipe to a variable, then loop over the variable.

SENDMAIL= grep "^server:" servers.file > servers.tmp

while read line; do
    if [ testauth() ]; then
        SENDMAIL="TRUE"
    fi
done < servers.tmp

if [ SENDMAIL ]; then
    sendMail()
fi

But not any more! I’ve found Xming, which is brilliant. It’s hardly new, but it’s new to me, so I’m excited. It works wonderfully with PuTTY. It starts when I log in and runs in the background with nothing but a system tray icon. It does all sorts of other X’y stuff, but all I want from my X server, is for it to stay out the way until needed. Hummingbird didn’t, Xming does. Lovely stuff.