Alain M. Lafon

I love Dropbox and use it daily. From sharing project files and source code repositories in professional life to saving and synching my important personal belongings, I wouldn’t want to miss it.

For the services I like and use, I usually don’t bother paying for them. But in the case of Dropbox, just no plan seems right. I’m not even talking money here, but storage – the entry plan being 50GB. On my main machine(MBA 11″), I’m running a 128GB SSD – meaning that after having installed all necessary tools, I don’t even have 50GB storage free for things that I want backed up or synched. Since I can’t even physically use the plan properly, I don’t see myself paying for it. I would have been glad to pay for a 10GB or 20GB plan, though.

Not even having 50GB storage might seem very little space to the reader – or for my former self, that is. But for the stuff that I previously needed a NAS or even dedicated servers, I now have the cloud. Mail is on my private server (and backed up), music is on Spotify and Soundcloud, movies are in iTunes. The only thing I’m still hosting and managing myself are pictures.

Still, I wanted more than the 2GB storage that Dropbox offers for free initially. Luckily Dropbox offers up to 16GB extra space for bringing in referrals. Over the last year, I’ve gotten a few people to use Dropbox and got my account up to over 4GB. But then I thought that I could improve this process and set up an AdWords campaign.

Just one day later and having only spent 25USD, my Dropbox account is now 16.8GB.

Here are the details of the campaign. You can see that it only took a day to accumulate enough referrals.

So now I can start putting more files into the bigger Dropbox.

Update: As mentioned in the comments, Dropbox normally grants 250MB per referral. If you own an edu email address, however, they will give you 500MB – even retroactively. If you happen to live outside the USA and your university uses a different TLD, don’t worry – there’s a form to put your address, they will verify it and let you join the edu program.

If you liked this article, please feel free to re-tweet it and let others know.

You should follow me on twitter here

Finally, the day has come for Spotify to offer its services to us Swiss folk. Previously, I have spent days trying to ‘hack’ an account so that I can start paying them money instead of Grooveshark.

Why did I switch? Let’s just compare two numbers – after that, nothing else needs to be said.

Cold startup time

And that’s without Adobes’ ever annoying “There’s a new Adobe AIR update available” message that pops up every other day and takes another 30s to click through – not counting the inevitable restart of the application afterwards.

Cold Start + searching and playing a song (Paul Kalkbrenner – Aaron)

Mind you, the numbers are including typing the artist and song. Spotify would serve the song right in under 3s if I were to type fast enough. Anyway, there’s a difference of 3 magnitudes in speed!

The benchmark was done on a contemporary Macbook Air with 4GB Ram and SSD, so not the sloppiest machine in the world. 124s to start listening to a song – in other words that’s two friggin’ minutes! I honestly can’t imagine a worse user experience.

Like I proclaimed, nothing else needs to be said.

Switch now and enjoy your life instead of waiting for the music to play!

P.S.: Migrating your existing playlists from Grooveshark to Spotify is made kinda easy with playlistify - so that’s not a reason to hold back.

Spoiler: This post is primarily gonna be an excerpt of my bookmarks collection. That’s because more intelligent men than me have already written great articles on the topic of how to become a great Python programmer.

I will focus on four primary topics: Functional programming, performance, testing and code guidelines. When those four aspects merge in one programmer, he or she will gain greatness no matter what.

Functional programming

Writing code in an imperative style has become the de facto standard. Imperative programs consist of statements that describe change of state. While this might sometimes be a performant way of coding, it sometimes isn’t (for example for sake of complexity) – also, it probably is not the most intuitive way when compared with declarative programming.

If you don’t know what I’m talking about, that’s great. Here are some starter articles to get your mind running. But beware, it’s a little like the red pill
- once you tasted functional programming, you don’t want to go back.

• http://www.amk.ca/python/writing/functional
• http://www.secnetix.de/olli/Python/lambda_functions.hawk
• http://docs.python.org/howto/functional.html

Performance

There’s so much talk going on about how inefficient these ‘scripting languages’ (Python, Ruby, …) are, that it’s easy to forget that very often it’s the algorithm chosen by the programmer that leads to horrible runtime behaviour.

Those articles are a great place to get a feel for the ins and outs of Python’s runtime behaviour, so you can get your high performing application writting in a language that is concise and fun to write. And if your manager asks about Python’s performance, don’t forget to mention that the second largest search engine in the world is run by Python – namely Youtube(see Python quotes).

• http://jaynes.colorado.edu/PythonIdioms.html
• http://wiki.python.org/moin/PythonSpeed/PerformanceTips

Testing

Testing is probably one the most misjudged topics in computer science these days. Some programmers really got it and emphasize TDD(test driven development) and it’s successor BDD(behaviour driven development) whereever possible. Others simply don’t feel it yet and think it’s a waste of time. Well, I’m gonna be that guy and tell you: If you haven’t started out on TDD/BDD yet, you have missed out greatly!

It’s not about introducing a technology to replace that release management automaton in your company that mindlessly clicks through the application once in a while, it is about giving you a tool to deeply understand your own problem domain – to really conquer, manipulate and twist it the way you want and need it to be. If you haven’t yet, give it a shot. These articles will give you some impulses:

• http://www.oreillynet.com/lpt/a/5463
• http://www.oreillynet.com/lpt/a/5584
• http://wiki.cacr.caltech.edu/danse/index.php/Unit_testing_and_Integration_testing
• http://docs.python.org/library/unittest.html

Code guidelines

Not all code is created equal. Some can be read and changed by any great programmer out there. But some can only be read and only sometimes changed by the original author – and that maybe only a couple of hours after he or she wrote it. Why is that? Because of missing test coverage (see above) and the lack of proper usage of coding guidelines.

These articles establish an absolute minimum to adhere to. When you follow these, you will write more consise and beautiful code. As a side effect it will be more readable and adaptable by you or anyone else.

• http://www.python.org/dev/peps/pep-0008/
• http://www.fantascienza.net/leonardo/ar/python_best_practices.html

Now go ahead and spread the word. Start with the guy sitting right next to you. Maybe you can go to the next hackathlon or code dojo and start becoming great proficient programmers together!

All the best on your journey.

If you liked this article, please feel free to re-tweet it and let others know.

You should follow me on twitter here