This blog post is a companion document to two Chit Chat Across the Pond segments I will be recording with Allison Sheridan on the NosillaCast over the next two weeks. The first of the two shows is now out, and and can be found here. One the second show is out I’ll add that link in too.

In episode 474 when Allison was chatting with Donal Burr about Apple’s new Swift programming language said she didn’t understand what a compiler was, so I thought it might be fun to try address that! But rather than focus in on just that one very specific question, I thought it would be more useful to take a high-level look at computer programming in general, so that some of the conversations around various developer platforms will make more sense to the majority of NosillaCast listeners, who are non-programers.

Read more

I’m betting most people won’t be interested in this, but if anyone’s curious how this programmer goes about building up a perl module from scratch, you can watch along on as I build XKPasswd 2.0 over on GITHub:

I’m re-implementing XKPasswd from scratch. The resulting functionality will be mostly the same, but with some additions. The reason for starting over is two-fold. Firstly, the first implementation of XKPasswd was very much a prototype, and as with any prototype, I learned some valuable lessons, and there are lots of things I’d do differently if I was doing it again. Secondly, the first version of XKPasswd is almost three years old now, and since then, my Perl skills have increased a lot. Probably the single biggest difference between the me of 2014 and the me of 2011 is that I read Perl Best Practices, and started to run all my Perl code through Perl Critic. Another big difference is that, thanks to JQuery, I’ve fallen in love with Code References in all the languages I use that support them, including Perl.

Since this is a re-write, I’m really focusing on building a solid foundation, so I’m starting with the utility functions that will let me build up functionality gradually, and I’m writing the user documentation in parallel with the code. Before every commit to GITHub, everything that’s done so far is getting documented with POD, and, the code has to pass Perl Critic with no warnings.

L in Semaphore Quite a few months ago I was joking with Guy Serle of the My Mac Podcast, and the topic somehow turned to flag semaphore. I think I’d sent the guy an iTunes review in mores code for a joke, and then Guy challenged me to do it in flag semaphore. I figured it would be easy to find a converter on the net, but, for the first time in a long time, the internet let me down! When I couldn’t find a converter I registered the domain with every intention of getting a converter up and running in a few days. Then, real life got in the way, and the domain sat there for months, until yesterday, when I unexpectedly had a free afternoon, and I finally got my converter written!

You can now go to the site and convert text to flag semaphore, play a crude animation of the signal, and even share a link to the signal. E.g. this link takes you straight to the conversion of “I love semaphores” to semaphore.

I sometimes take some stick for having a very defensive coding style. When ever I find myself making an assumption I throw in code to test it. “This function will only ever be called with a positive number”, OK, then add a test to throw an exception should a negative number be passed. You don’t want bad data ricocheting through your code because goodness knows what damage it will do! Similarly, my style is to always use explicit syntax, and, to avoid syntax shortcuts – sure, the ternary operator takes up less space on the screen, but there’s a price to pay for that terseness – it makes your code harder to read and hence to debug and maintain.

However, one of my very biggest bug-bears is the failure to brace control statements like conditionals and loops when they’re operating on a single line of code. This is the trap Apple fell into so spectacularly this week.

Read more

Inspired by a recent episode of The Mac Cast I decided to see if I could come up with a simple way of getting a word count of a PDF on OS X using only tools that come standard with the OS.

Because of OS X’s Unix underpinnings, all Macs have access to the Unix wc command which calculates word counts on given input. OS X also has a handy built in Terminal command to access the contents of the clipboard (pbpaste). This leads to an obvious simple manual solution:

  1. Open the PDF in Preview
  2. Select All Text
  3. Copy to clipboard
  4. Run the Terminal command: pbpaste | wc -w

This is a bit cumbersome though, so I went on to create a simple OS X Service to calculate the word count of any selectable text in any app (the fact that this is even possible, let alone easy, is why I love OS X).

For those of you just looking for a copy of the Service, you can download it here:


To install the service simply extract the automator file from the ZIP archive and copy it into either the Library/Services folder in your home directory, or the system-wide service folder /Library/Services.

Once the Service is installed you can use it in almost any OS X app (specifically in any app written using the standard Cocoa libraries) by selecting some text, right-clicking on it, and selecting the Word Count service:

Right Click Selected Text to Invoke the Service

When done the results will look something like this:

Sample Output

Those of you who want to see how easy this Service was to write, read on and I’ll walk you through it.

Read more

I spent quite some time this afternoon finding a reliable way of converting Unix Time Stamps (UTS) to the ISO 8601 format used in SQL databases like MySQL and PostgreSQL that does not get confused by timezones. In the end the final result is, as is so often the case with Perl, very short and simple, but since the Googles failed to find the answer for me today, I thought it would be worth sharing in case it’s of use to others in the future.

  1. use DateTime;
  3. # Function to convert Unix Time Stamp to ISO 8601 Date
  4. # Arguments:
  5. # 1) The Unix Time Stamp to convert
  6. sub uts_to_iso8601{
  7.     my $uts = shift;
  8.     my $date = DateTime->from_epoch(epoch => $uts, time_zone => 'UTC');
  9.     return $date->ymd().q{ }.$date->hms().'z';
  10. }

The algorithm is very simple, use DateTime‘s from_epoch function to a DateTime object in the UTC timezone (AKA Zulu). Then assemble the output as YYYY-MM-DD HH:MM:SS, and append a z for Zulu. MySQL and PostgreSQL can now use the string to populate Date or Timestamp columns.

As an example, the Unix Time Stamp 1369410796 converts to 2013-05-24 15:53:16z.

This is a minor bug-fix update for XKpasswd (my Perl random password generation module). It squashes two minor bugs which came to light while updating to use version 2 of the module.

  1. When the custom_separator option was left blank, no separator was used, rather than the expected random separator.
  2. When the custom_separator option was left blank or set to RANDOM, and the pad_char option to SEPARATOR, the results were un-expected, different random character was used for each, rather than the same random character.

For documentation and detailed release notes on version 2 of the module, see the release notes for version 2.0.


Automator + XKpasswdA few weeks ago on the Chit Chat Across the Pond segment of the Nosillacast, I mentioned that I had an OS X service set up to generate a random password using my XKpasswd Perl module and copy it to the clipboard. Listeners enquired as to how they would go about doing that, so as promised, here’s a quick tutorial.

Obviously this tutorial is for Mac OS X users only, because OS-wide Services and Automator are OS X features. The screenshots are taken on 10.8 Mountain Lion, but this same technique definitely also works on OSX 10.7 Lion, and probably even on 10.6 Snow Leopard. This tutorial also assumes that you have downloaded the XKpasswd module, and saved it somewhere on your computer, along with either the sample dictionary file included with the module or one of your own making, and that you know where on your computer those files have been saved. In other words, you need to have and a text file with one word per line somewhere on your hard drive. In my sample code I’m going to assume you’ve installed the Perl module to the suggested location, /usr/local/xkpasswd/, and that you have customised the sample dictionary a little (more secure that way), and saved it to /usr/local/xkpasswd/dict.txt.

Read more

I spent the weekend majorly re-factoring, my Perl random password generation library. V0.1 was the last thing I wrote before reading Perl Best Practices, and looking back on that code really illustrated the value of that book when used in combination with the perlcritic code analyser.

The new version of the module provides all the functionality the old one did, and more. The refactoring has made the module simpler to use from within scripts, as well as easier to modify and extend. Some new features have also been added, including the ability to use the web service as the source of randomness for the library. A full list of bug fixes and new features is included below.

I had hoped to distribute this version as both a ZIP file and a .PKG file, but XCode 4.4 is not being cooperative on the new Mountain Lion, so that will probably have to wait until version 0.3.

Update – 6 August 2012: The link below has been updated to point to version 0.2.1 of the code. Details of the bugs fixed in the release notes.


Read more

Yesterday I posted my thoughts on watermarking images for uploading to the web. This post formed the basis of myself and Allison Sheridan’s discussions in the Chit Chat Across the Pond segment on tonight’s Nosillacast Mac Podcast. During the discussion we talked a little about how I watermark my images, and I realised that I hadn’t released a version of my image processing scripts since 2008! I promised I’d remedy that, so today, after a little tidying up, I’m releasing the current snapshot of my scripts under the FreeBSD license.

I’m not releasing these scripts as a polished software package that’s ready to use, but rather, as a starting point for anyone who wants to create their own watermarking scripts. If you’re not prepared to get stuck into the command line and a little Perl (VERY little is needed mind), these scripts are of no use to you!

The scripts rely on the free and open source ImageMagick command line image editing tools, so you’ll need to install ImageMagick before you get stuck in. If you’re running OS X, I’d recommend installing ImageMagick via the free and open source MacPorts package manager.

Read more

keep looking »