On usability, a case study

02f December 6, 2014 -- (tech)

Given nowadays' popularity of the web, usability is, I suppose, still something of a hot topic in the field of human-computer interaction. More than half a decade since the first electronic computer, the concepts behind making our computing devices usable are still largely left to exploration. But what is "usable"? I will defer to my friend, Merriam-Webster:

1: capable of being used
2: convenient and practicable for use

So, is a hammer usable? Well, it is, in the sense that I can use it to drive nails into pieces of wood. Is then a guitar usable? Well, it generally is, in the sense that it allows me to make music; in particular, it might or might not be, depending on the guitar's neck and the length and width of my fingers. So, is the computer, a general-purpose device typically comprising a keyboard, a mouse and a monitor as input/output tools, usable? It's hard to say, this depends on what I want to do with it, on the operating system and the user interface it implements and so on; if I intend to play a game, maybe I'm better off using a gamepad or some other type of specialized controller1.

Getting back to the World Wide Web, in the last few decades companies such as Google have been pushing to move at least a part of their applications on the web. JavaScript has become the most widely used programming language on the web for exactly this purpose: the web had an untapped potential for improved interaction from the very beginning, as semantic content can be easily implemented on top of it using pages, hyperlinks and "rich text" elements. JavaScript merely allows this content to be modified dynamically, allowing, for example, the browser to change a section of a page when the user presses a button, as opposed to reloading the entire page. Along with the great potential for developing web applications, this has also opened a few more perverse avenues for developers; for the sake of sticking to the subject, I won't go into any details regarding this aspect.

I wish to present a comparative case study for the purpose of illustrating the usability of web applications. I will dive into the bowels of one of the web's most widely used applications, Gmail, comparing it with Mutt. I've been using both of them extensively for at least one year now, so I am quite able to distinguish between the pros and cons of both.

Firstly, I should mention that Gmail comes with many features in comparison to Mutt. For example Mutt doesn't have filters and it doesn't offer any interface for editing mail by itself; it doesn't have any support for built-in chat, nor does it allow configuration for multiple accounts, since it doesn't really have a well-defined concept of "accounts". Fortunately, that functionality can be integrated using many third-party applications, which is why I will focus on the basics, i.e. reading mail and making sense of the great e-mail organization mess of which we are all aware.

One major advantage that Gmail has over all the other clients is that it replaces folders with labels. This is compatible with the IMAP folder view, but with the addition that you can have a single mail residing in multiple folders (or under multiple labels, to use Gmail terminology) at the same time. This is indeed very useful, mostly because you can keep an e-mail in the inbox and in another folder simultaneously. However, Gmail's biggest and greatest advantage is the search function, allowing anyone to find e-mails almost instantaneously2.

Mutt on the other hand was created back in the 1990s, when folders weren't very popular, so the focus of the main window is on the current folder and only it3. Mutt's main disadvantage is the steep learning curve: you have to sit a few hours to configure it before obtaining a usable interface. After you waste that time, however, the interface will be blazingly fast, albeit keyboard-driven instead of mouse driven. This feature is so useful that it was borrowed by Gmail's keyboard shortcut interface, whose documentation, in case you're not aware, can be accessed using the question mark (?) key.

Regarding composing and answering to mails, I mentioned earlier that Mutt doesn't have a built-in editor. Well, no, but it simply opens your system's default text editor whenever you want to edit an e-mail. This hard separation is, I believe, Mutt's greatest strength and Gmail's greatest weakness. Did you ever open a draft in Gmail's big, shiny "composer", started writing, focusing some other window, then re-focusing the "composer" window, pressing enter and finding out that your draft was just sent? Well, that, dear reader, is the very opposite of usability: an e-mail client should never, ever, ever send your e-mail when you press enter, because that's one of the largest keys on your keyboard and one of the most commonly used. Not to mention that when I write e-mails, I want to write e-mails in that precise context, without any useless clutter.

Mutt's greatest disadvantage is that it sucks. However, as its author mentions, "All mail clients suck. This one just sucks less".

So, is Gmail usable? Mostly. Is Mutt usable? Not much more than the previously mentioned client, but it does the same essential stuff at a much lower price. At the end of the day we use whatever we feel comfortable with, regardless of their usability. Or "usability".

  1. Although I personally never enjoyed using controllers to play games. The mouse and the keyboard are the perfect interface for, say, a first person shooter.

  2. Note that this is also partly a feature of e-mail being text-driven, rather than HTML-driven or whatever nonsense "modern" "enterprise" e-mail tries to push nowadays. HTML isn't usable in e-mail, because not everyone can or wants to look at HTML, and not everyone wants cross-site scripting embedded as a "feature" in e-mails. Also note that this is not a matter of preference, despite how much your mileage may vary.

  3. Although certain forks come with support for sidebars.