Window Snapper

Many programs have a nice dividers in their GUI - and you can usually drag the divider to resize the panels on either side. I've often been a little frustrated when resizing windows to tile them to fill the screen. Perhaps a Word document on the left, a PDF in the upper right, and email in the lower right. It was quite tedious, and because of that I didn't do it often, and I feel this slightly diminished the power of windows. So I wrote an AutoHotkey script and now I have what I wanted.

To use it, while the script is running, you can right-button drag window borders, and the script will move all neighbouring window edges at the same time. The script will grab window borders within 10 pixels (configurable) of the border/s you are right-button dragging. It also assumes a seven pixel wide window resize border, but that's also configurable at the top of the script. You can resize from edges or corners - giving you lots of power when resizing windows.

The most common use-case I have is with just two windows positioned with Win+Left and Win+Right (like explained here). With this script you can resize both windows effortlessly.

Here's an example of it in action:

 I can resize all five windows with one right-button drag, in the area where the four window corners meet in the middle of the screen. If I drag towards the upper right, I can get something like this:

 And things like this are possible too, but unsnapping window edges like I've done below with the left and right can be a little fiddly. I could potentially add something like Ctrl+Right-Click to only resize windows actually near the mouse...

Please have a go with the window snapper script. Please let me know what you think after using it!

The script is licensed under a Creative Commons Attribution 4.0 International License.

Engagement in the Family

My brother-in-law Rowan Barton is engaged to Elpida Savvidou. This is awesome news (even if not recent :-). Their engagement party is movie themed, and we've got some western style props, so I've created custom wanted posters! Here they are:

Once printed we are going to brown the edges with the heat from a fire.

The SVG files were created with Inkscape (Trace Bitmap is fantastic) and are available for use if you'd like. I put them in the public domain (but I'd be thrilled to hear from you if you use these). They make use of the Playbill font.

I'm interested to see what other people have created. Enjoy :-)

Minecraft vs Real Life

I have been digging a hole in the ground. It is just a bit more than one and a half metres deep, and around a metre in diameter. Here is the hole:

Here is the hole with the dirt I have emptied out of the hole in view:

Here is an equivalent sized hole in Minecraft:

Apart from the Minecraft hole being digital, it was also a lot neater. And it took less time to dig. The real hole above has taken several weekends of digging - probably close to 20 hours of digging. The Minecraft hole took around 20 seconds, that's including deciding where to dig. Perhaps that's one reason why games like Minecraft are enjoyable to play - they let you do mundane things really quickly.

Bits of Minecraft

Minecraft has sunk a lot of my time over the few years it's been around. Here are some of the things I've done in it (in order of construction):

Rainbow! ... and cake

Triple helix near an overworld plate

Home, satellite dish and "Craft"

The rainbow, triple helix and satellite dish were all planned using Lisp. Here's some Lisp code you can use to plan rainbows:

(loop with freq = (list)
       and reach = (1+ 20)
       and width = 8
      for j from reach downto 0
    do (loop for i from (- reach) to reach
             for r = (round (sqrt (+ (square i) (square j))))
             for c = (cond ((= i j 0) #\+)
                           ((< r (- reach width)) #\-)
                           ((<= r (1- reach)) (- r (- reach width)))
                           (t #\Space))
        when (not (eql #\Space c))
        do (aif (cdr (assoc c freq))
            (incf (car it))
            (push (list c 1) freq))
        when (or (minusp i) (not (eql #\Space c))) do (princ c))
    do (terpri)
    finally (return freq))

It produces output like this (this one's a small version to fit in the width of the blog):

                 777777777
              777666666666777
            7766655555555566677
          77665554444444445556677
         7665544433333333344455667
        766544433322222223334445667
       76554433222111111122233445567
      7655433221110000000111223345567
     7655433211000-------0001123345567
    766543321100-----------001123345667
    7654332100---------------0012334567
   7654432100-----------------0012344567
   765432110-------------------011234567
  765443210---------------------012344567
  765432100---------------------001234567
  76543210-----------------------01234567
 765433210-----------------------012334567
 76543210-------------------------01234567
 76543210-------------------------01234567
 76543210-------------------------01234567
 76543210------------+------------01234567
((#\+ 1) (#\- 256) (0 45) (1 45) (2 43) (3 57) (4 57) (5 57) (6 59) (7 57))

The last bit is a count of each block type so you know how much to gather of each type. Three dimensional objects like the dish and triple helix use more code which I can post if people are interested.

P.S. Don't spend too much time playing computer games.

Diode PN Junction Explained

I never felt like I properly understood how the PN junction in a diode works. So I looked it up and came across some explanations that really helped me understand it better. I've put it all into a diagram. But let's introduce some of the concepts first.

A diode is made of a semiconductor (like Silicon) with two regions. The semiconductor makes up a crystal. One region in the crystal has been doped with atoms that have a spare electron when in the crystal lattice. It is known as the N region. The other region in the crystal has been doped with atoms that have spare room for an electron when in the crystal lattice (the spare spots are known as holes). It is known as the P region. That's where the name PN junction comes from.

The N region with its spare electrons conducts via movement of these spare electrons (the electrons are the charge carriers in an N region). The electrons are the charge carriers. Conversely, the P region conducts via the movement of the 'holes' (the holes are the charge carriers in a P region). As an aside, undoped semiconductors don't conduct as well because they don't have mobile charge carriers.

An important note to remember is that while the regions may be called N and P, this does not mean that they are overall positively or negatively charged. The regions have an overall neutral charge. Following on from that, if you were to take away the electrons in an N region, the crystal structure would be positively charged. The opposite would occur if you filled the holes in a P region with electrons.

So, with that explanation, here's the diagram:

So the depletion region gives a potential barrier (from the fixed charges). The width of the depletion region (and thus also the size of the potential barrier) depends on the applied potential (in effect, it depends on the direction of the attempted flow of current).

Thanks to:

• http://library.thinkquest.org/12666/junction.html
• The Wikipedia page about diodes
• The Capacity Factor for the SVG graphics in Blogger instructions
• And Inkscape!

The PN Junction Explained diagram by Jonathan F Johansen is licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License.