I’ve uploaded the slides from the presentation for your convenience – next year, I’ll try to record any presentation(s) and put them online as well!

Slides: Practical Computer Science with Python (eTech 2010)

Links:

• Think Python: How to Think Like a Computer Scientist (free textbook, great for intro CS courses)
• Python

On Macs, there is a wonderful program called The Unarchiver that “just works” for extracting ZIP files. It works very similar to the built-in unzipping functionality, but supports many more file formats. I love it – when I double-click on a compressed file, it expands it in the current folder, and it’s very smart about how to do that. If there’s only one file inside the compressed file, it puts it right in the current folder. If there’s more than one file, it puts it into an appropriately-named subfolder.

On Windows, 7-zip is probably the most versatile unarchiving program. 7-zip will extract practically anything you throw at it, and it also is great for creating archives as well. However, it’s double-click functionality feels very “Windows”-like – it opens up a file-manager style view, and then you have to decide what to do with it. If you set up the context menu, you can right-click on an archive and choose 7-zip -> Extract Here, and it will sort of do what The Unarchiver does… but not really. It’s not smart about folder placement/creation, and did I mention you have to right-click then go through a submenu?

I decided to write a script in VBS that would work more like The Unarchiver, and bind this to the default action for double-clicking on any archive files. I’ve created an installer file that will copy the VBS script to your 7-zip folder, and optionally set up the file associations for you. I haven’t tested it thoroughly, but it appears to work fine on two of my computers.

Note that this requires that you have 7-zip installed. I recommend installing it to the default directory of C:\Program Files\7-zip, but if you install it somewhere else, just make sure to specify the real 7-zip directory in the 7zunarchiver installer.

7zUnarchiver Installer v1.0

Not too long ago, I won an Amazon Kindle while attending the CS4HS workshop at CMU (Yay Google and other sponsors!). Perhaps not surprisingly, I fell in love with reading on it. I quickly looked for all the free e-books I could, discovering sites like MobileRead and FeedBooks, with their ultra-sweet Kindle-based book browsers (Click and download right on the Kindle? Awesome!). However, their selections of recent works are certainly… lacking. Only a few authors have really embraced free e-book distribution, and while I do love Cory Doctorow, my brain really craved more.

I investigated the local library, which has been a great source of mind-food ever since we moved to Cleveland a couple years ago. Alas, they had nothing expressly for the Kindle – though they did have an e-library with MobiPocket and Adobe EPub books. I had remembered reading somewhere that the Kindle could do MobiPocket, so I decided to investigate further.

I found out that Igor Skochinsky had done some investigations into the Kindle’s handling of MobiPocket DRM, and had put together a set of Python scripts that were supposedly able to convert normal MobiPocket DRM’d files into a format readable by the Kindle. I tried them out on my own, and lo and behold, I was reading Barack Obama’s book on my Kindle, straight from the library! Well, not quite straight from the library – it made some detours on my Mac, involving some Terminal usage and some copying-and-pasting. But it was still pretty sweet!

I decided that dealing with the command-line was not something my wife would be interested in doing, and so I decided to write a little AppleScript application that handled all the tedious stuff, like running the commands. This, in fact, only ended up taking me the lesser part of an afternoon, despite having never worked on a similar project before. However, the end result was a fancy little droplet, that I could simply drag and drop a PRC file onto and get an Amazon-compatible AZW file back.

I contacted Igor, and he gave me permission to redistribute his code. So I created a logo and a Google Code project, and am happy to share it with you, readers of the world! I give you:

Mobi2Kindle!

cd \xampp\php

go-pear

pear install MDB2

pear install pear/MDB2#mysql

Being a web developer (if only part-time at the moment), I found myself having to remember how I went ahead and set up my testing environment so I could do rapid virtual host set-up, and I decided to write it down for posterity’s sake. This is just my own desired set-up – your own preferences may vary, but perhaps you’ll find this useful!

Some background: I work on sites for many different clients. When it comes to organizing things, I like keeping a folder for each client’s site (with documents, PSD files, etc.), along with a separate folder for the “web root” of the site. I usually keep the web root in Subversion as well (refer to my earlier article for info on setting that up). With the instructions below, I’ll show you how to easily set things up so you can just create a new folder and have XAMPP automatically get you there, without having to edit any configuration files. With the special DNS sauce, I can even type in a URL like “http://jamesmallen.net.imac” and view my local test version of jamesmallen.net.

First off, you need XAMPP. I use the Mac version, available from Apache Friends. I probably could use OS X’s built-in Apache server, but I prefer to keep this separate from my “system” files, so I can easily drop in different versions of things. XAMPP is nice and self-contained. If you decide to go with XAMPP, make sure you disable Apache by unchecking the “Web Sharing” option of the “Sharing” system preference pane.

With XAMPP’s default install, XAMPP will serve pages from /Applications/xampp/xamppfiles/htdocs. This isn’t a very convenient location, and besides that requires root access/su authentication to edit files. I prefer to keep different websites in my ~/Sites folder (which is part of my Time Machine backup), as I like to think of them as works-in-progress. Once you’ve decided where you want to put your sites, you need to create a new configuration file to have Apache include.

Save something like the following as a new file called /Applications/xampp/xamppfiles/etc/extra/httpd-dynamic-vhosts.conf:
UseCanonicalName Off
VirtualDocumentRoot /Users/james/Sites/%-2+

You’ll want to replace “/Users/james/Sites” with whatever folder you’re using to keep all of your different domains. The %-2+ indicates that you want to use all but the very last part of the domain name in determining what directory to use. I leave off the last part, because I want to type “http://mydomain.imac”, but I don’t want to have to include the “.imac” in every directory I create.

Now you need to include above .conf file, so add the following line to the bottom of /Applications/xampp/xamppfiles/etc/httpd.conf:
Include /Applications/xampp/etc/extra/httpd-dynamic-vhosts.conf

If you want to be able to access the default xampp htdocs, you’ll need to create a link to them. The following command in a terminal window will create a symbolic link, letting us keep things separated:
ln -s /Applications/xampp/xamppfiles/htdocs /Users/james/Sites/xampp

Again, replace “/Users/james/Sites” with whatever root you’re using.

Finally, the DNS sauce. This will depend on the capabilities of your router. I have an older Linksys WRT54G, which lets me run the most excellent . If you don’t have a highly-configurable router, I highly suggest you get a cheap one that is compatible with Tomato – it’s wonderful to have all the power of a Linux box in such a low-power, always-on device. Tomato uses dnsmasq for DHCP services as well as DNS caching.

First, set up your development machine to have a static IP (using Tomato, you can do this by clicking on the particular machine in the “Device List”). Next, get into the advanced dnsmasq configuration by clicking on “Advanced,” then “DHCP / DNS” in the Tomato menus. Enable the “Internal Caching DNS Forwarder” if it isn’t already enabled, then add the following line to the Custom Configuration box:
address=/.imac/192.168.1.100

Replace “.imac” with whatever fake TLD you want to create (I used to use “.dev” before I started using multiple dev machines). Make sure it isn’t an ACTUAL TLD, or else you’ll cut yourself off from all sites that end with that! Also, replace the “192.168.1.100″ with whatever static IP your dev machine has on the network.

Save the changes, wait for the services to restart, and then fire up XAMPP. I recommend using the XAMPP Control Panel.app to just start the services you need. Test things out by navigating to http://xampp.imac. If all is well, you should see the XAMPP splash page. Congratulations!

My most recent project involves editing a slew footage of which I was not involved in shooting. One out of five cameras that were recording during the event (a musical) was shot using 12-bit audio. This is my first experience trying to use footage with 12-bit audio in FCE, and frankly, I’m not impressed.

It is my understanding that “12-bit audio” doesn’t tell the full story – besides having a “vertical” resolution of only twelve bits, it also has a lower sample rate than 16-bit DV audio. 12-bit has a sample rate of 32,000 Hz, whereas 16-bit has a sample rate of 48,000 Hz. So far, so good.

However, when I import the 12-bit tape into FCE, I get some serious synchronization issues. After about a minute of tape, they’re noticeable – by the end of the tape, the audio and video are off by more than a second. FCE has an “Easy Setup” option for importing 32KHz DV footage – I tried this setup, as well as the normal DV setup.

After fiddling with different settings, I came across an Apple KB article about iMovie and 16-bit audio. It describes similar symptoms when exporting a Quicktime movie from iMovie. I decided to try importing the problematic tapes into iMovie, and see if I could follow the suggested instructions for “converting” the tapes to 16-bit audio. I should mention that this was using iMovie (HD) ‘06.

Well, when I tried outputting the iMovie clips back to a new tape, they were being sent with only 12-bit audio, so no dice there. However, I think I found a workaround that fixes the audio sync issues.

1. Import your DV clips using iMovie.
2. Save your iMovie project, and then find the .iMovieProject file in the Finder. Right-click it and choose “Show Package Contents”.
3. In the “Media” directory inside your iMovie project’s package, you should see a bunch of .dv files. FCE can work with these, but it doesn’t like to – you’ll have to render any audio before you can hear it, and playback will be SUPER-slow. To work with these files nicely in FCE, you need to convert them to Quicktime movies.
4. Open up MPEG Streamclip (It’s free, in case you don’t already have it). If you’ve only got one DV clip that you’re working with, just drag the .dv file to the main window. If you’re working with multiple files, open up a Batch List window (Cmd+B), and drag all the .dv files into the Batch list.
5. For your task, you want to “Export to Quicktime”, using the following settings:
   

Note that here we’re converting the audio to 48 kHz. It may also work if you leave the sound setting at “Auto” (which will result in a file with 32 kHz sound) – this may be “truer” to the original footage, and you can leave the upsampling to FCE. YMMV.

6. You’ll also want to check out the DV Options by clicking the “Options” button at the top right, and make sure that you’re set up for 4:3 Interlaced (or 16:9, if you’re using anamorphic 16:9 footage):
   
7. Now, start the conversion process, and when it’s finished, drag the new .mov files into Final Cut, just like importing any other hard disk footage. Voila! No more sound sync problems!

I wish I knew why FCE had these issues in the first place – it seems like a rather significant oversight that projects can’t cope with these, especially since many consumer camcorders nowadays shoot with 12-bit audio by default. Oh well – in the future, I’ll try and make sure all the potential videographers for an event are consistent with 16-bit, now that I know what a pain it is to import.

For the Mac, there’s SCPlugin, which seems to offer similar functionality and Finder integration. Of course, there are a couple additional (and, in my opinion, poorly-documented) steps required to use SCPlugin with an svn+ssh connection. (Note: these steps worked for me on OS X 10.5, Leopard. YMMV)

1. Get SCPlugin from http://scplugin.tigris.org/ and install it. There is no need for a separate installation for SVN – SVN 1.4.4 appears to come pre-installed on Leopard.
2. Restart the Finder. I just opened up Force Quit (Command-Option-Escape) and chose to Relaunch Finder.app, but you can also just log out and log back in if Force Quit scares you.
3. Here’s the tricky part – you need to set up ssh to use key pair authentication as opposed to password authentication when you connect to your SSH server. Detailed instructions can be found here.
4. Now check something out using SCPlugin. In Leopard, this is done by right-clicking in the target directory in a Finder window, choosing “More,” “Subversion,” “Checkout,” and entering the repository URL and other information as necessary.

I didn’t see any mention of this on any recent blogs, so I’m guessing they’re rolling it out slowly? In any case, you can clearly see the “Older Version” link in the screenshot above, and if I click on it, I then have a “Newer Version” option.

So far, I’ve only noticed two differences – the window title is much longer (now “Gmail – Inbox – [EMAIL_ADDRESS]“), and the tooltips for Quick Contacts are a bit fancier. I personally prefer the shorter window title – with the older version, when I’m in another tab, I can tell at a glance if I’ve got new messages simply by whether or not the title for the Gmail tab has a “…” at the end.

It’s called Actiontastic, and it’s geared toward using the Getting Things Done approach to productivity/time management. I myself have not read David Allen’s book on the subject, but simply the description of using the software (supplemented by a Wikipedia article) has given me the basic gist of the approach.

When I used my previous task management software (ToDoList by AbstractSpoon), I didn’t really think too hard about the way in which I organized tasks. I separated according to projects, but beyond that, my tasks ranged widely in scope and clarity. One task might be “E-mail quote to client,” another task might be “Make website.” The former is a good example of something that can be clearly handled and marked as done, but the latter, aside from having a somewhat ambiguous completion status, does not lend itself well to figuring out how best to spend a chunk of productive time.

There are a few major points of GTD that I’ve gotten just from reading Actiontastic’s release notes which stick with me as being good:

1. Granularity. Focusing on every open project having a defined “next action,” that leaves no uncertainty about what must be done.
2. Contexts – With the above actions, determining what context they must be done in – and thus being able to group actions according to context. (E.g., I am now in my car doing errands, so I should do other actions that require me to be in my car doing errands).
3. Immediacy – Before “actions” are created, an “inbox” of items needing attention is supposed to be processed. During this process, if an action can be taken care of in approximately 2 minutes, it should just be done and gotten out of the way.

I personally welcome the changes to my organization encouraged by this new approach. I also find it interesting that, at least in my case, a search for software has led me to adopt a very different approach to the problem which I was originally searching for software to help me solve. As a software developer, I wonder about the potential for writing other types of applications that introduce people to new, otherwise unthought-of approaches to problems they face.

Now, we’ve been a “cross-platform family” ever since we were married, with a 12″ PowerBook G4 as our designated Mac, a couple Windows machines, and we’ve even used (though didn’t stick with) Ubuntu on one of our computers. This new purchase is notable, however, as it means that I will be moving my primary development work to the Mac. While I used a PowerBook for a few programming projects in college, for the most part I’ve done my work in a Windows environment.

I use a large number of Windows-only tools as part of my work, and this has me worried. Many of them are free/open source, but currently lack Mac-native versions. Some (such as KeePass) have Mac versions, but they are sorely lacking in terms of the UI – they tend to feel very “un-Mac-like” (which I suppose is to be expected of a cross-platform application). I will certainly be using either Boot Camp and some sort of virtualization software, if only for the purpose of testing rendering of websites in Windows, but I’d prefer to use native apps for everything possible.

For reference, here is the list of software that I’m going to miss.

• WinSCP
  The most perfect SFTP/FTP program for Windows.
  So far, I’ve tried CyberDuck and Fugu, both of which fall short in terms of execution and lacking key features (specifically, live “Synchronization”).
• KeePass
  A password manager program.
  This is not of the “fill out web forms” variety, but rather an encrypted database designed to hold information about user accounts of various forms.
  I’ve looked at the Apple-included “Keychain Access” application, but I like the additional features/organizational options of KeePass. I’ve also tried KeePassX – the cross-platform version of KeePass – and its interface is horribly non-native.
• Notepad++
  A very versatile text editor.
  I think I will be switching to TextMate, even though it’s not free, because it looks to be also very versatile and good at what it does. I still have to spend some more time with it before I can be sure, though.
• ToDoList by AbstractSpoon
  A heirarchical to-do list application.
  Again, there’s an Apple application that sort of does this in the form of iCal, but the implementation doesn’t have as many features (subtasks being the most noticeable).

I’m still looking for replacements for the above tools, but I’m not ruling out the possibility of “rolling-my-own” applications (or, in the case of the open source apps, using the same core but completely re-doing the interface in Cocoa). I’ve been wanting to try out native Mac development for awhile – this migration might be my chance!

If anyone has any suggestions for replacements not mentioned above, I’d appreciate them. What did you use before and after you made the switch?