Category Archives: Writing

Introducing Plinky

I constantly begin blog posts with the pronoun “I” and I know a blog is definitionally “a personal journal [… which usually showcases] the work of a single individual” but just shut up about yourself, Kevin, for five minutes find it hard to come up with ideas for what to write about these days. I don’t really have a problem coming up for ideas for designs—earlier today in fact, I was really proud of my first drafts for the wholesale newsletter we at Mister B send out to our clients every week. But sometimes, all we need is a prompt and we’re on our way.

Let me introduce Plinky: a site currently managed by the team behind blogging platform WordPress, Automattic. While you can write and publish content directly from the site, I would be using it to generate a rough idea which I would then start hacking away at—yeah, that sounds an awful lot like stealing to me. It’s a shame there’s no random prompt feature built-in. Luckily, the permalink format is random number generation-friendly: insert a number ($) into the URL$/answers/new and Bob’s your uncle. I hope to be using it to overcome my acute writer’s block over the coming weeks. Expect posts on whether or not ghosts exist (they don’t) and how I got to where I am today.

Actually, that second one might be a good one to start with.


Written after furthering my Main Document

This article was written for the ‘Without the Internet’ series. I’ve been without the Internet for almost a month now and, as a result, I’ve been spending most of my free time pruning my Mac mini inside and out – coming across interesting snippets from the last few years.

I haven’t mentioned my ‘Main Document’ in a while. That might be because I haven’t actually been keeping up with it; it’s been one of those projects that, while having a explicit end, I’ve just never had a chance to complete.

I’ll first explain what the ‘Main Document’ is and the reason behind it being at all. Towards the end of my 2009–10 course at Cambridge Regional College into art and graphic design, I wrote a detailed account of what was called the “final major project”. While I didn’t enjoy the project itself, I did very much enjoy documenting it in detail and I thought I’d carry the practice on into my next course in radio. I was also, on the art-design course, having problems with certain tutors grading my coursework factionally; I thought that I might appreciate having a precise and comprehensive text full of the coursework that wouldn’t get submitted (the ‘working out’) later on, so I could hand it into the qualifications body for reassessment.

So between September 2010 and April 2011—which was the first of three ‘courses’, each one lasting about thirty college months—this ‘Main Document’ document became the centrepiece of my coursework and a place to collate notes, scripts, pieces of copy, despatches and so forth. The document was always written in arrears but, owing to the fact that a great deal of assignments had to be written in late March and early April, I fell a little behind on writing up old notes (the most recent day’s work was always written up soon after) and transcribing old programmes (which takes ages).

In the three-week holiday between the first and second ‘courses’ of the course, I was interning at Radio Netherlands Worldwide in Hilversum, the Netherlands. My plan was that each evening after interning, I’d return to my bed-and-breakfast and write up a page or two more of the document. To cut a long story short, this wasn’t the case; perhaps the distractions of living in a new country were simply far more interesting than writing up old college work. I managed to get myself another month-long internship at RNW before finding full-time employment at Mister B. Again, the distractions of my new home were just too damn interesting (they still are).

It’s nine months to the day since I arrived in the Netherlands and the ‘Main Document’ has come forward little since that Spring day of ferry boredom and suicidal cycling across provincial Holland and Utrecht: at times, I was so damn tired, I almost fell into one of a number of rivers and canals. But being without the Internet this past month means there have been fewer distractions and interruptions, and so back to writing I’ve gone. My parents brought over two cardboard boxes worth of notes and documents a few months ago and, one by one, I’m slowly sifting through the mountain of paper; typing each one up and putting it in an appropriate place in the now nearly 700 kB LaTeX file of 52,660 words. This blog post, by comparison, is hovering at a meagre 4 kB.

Memory and where I dump mine

While doing twenty things at once this afternoon, I realised just how many of my hundreds of ideas a day don’t make it to paper or a keyboard before they disappear. My memory is like a sieve and I require not one but (and I’ve counted them) eight services to hold everything that spews out of my brain somewhere until I can deal with it. I haven’t included Google Calendar or my HTC’s Voice Memos application in that magic figure: Google Calendar is, organisationally, something slightly different; and I’ve begun to use Voice Memos less and less, since the transcription was beginning to piss me off.

Of course, no two ideas are the same; but my ideas tend to be more suited to a specific one of these eight applications or services over all the others. I’m going to show you two of the eight that I use most regularly, Instapaper and Remember the Milk.

Instapaper is something that I abuse constantly. Instapaper is supposed to be about finding an article or post on the Internet but because you don’t have the time there and then to read it, you ‘save it for later’. It’s an amazing service but it comes with one downside: it’s too convenient. While I do use it in its indented fashion 75% of the time, I also use it as a place to store YouTube videos; as a wishlist; or as a list of Wikipedia articles which I’ve found typos on (I really don’t have the less-than-fifteen seconds it takes to click ‘Edit’ and correct the typo; I’m that busy). And so, to Marco Arment, I apologise: this behaviour will stop.

Remember the Milk is a truly awesome task management site. Tasks are, again, a different kind of idea: they don’t require too much work in themselves but everyone needs a reminder now and then. I can input a task in RTM and forget about it; then, every morning, I receive an email (I used to also receive direct messages via Twitter) giving me an overview of what needs to be done. The one problem I have with these emails is that they don’t include overdue tasks.

But even adding tasks can be a task in itself. Because I work a lot in the shell, I don’t just want the ability to type something like:

kevin@computer:~$ rtm "Figure out which of these Smart Add characters need to be escaped ^today !1 #rtm #investigate"

to add a task to Remember the Milk (which can be done by adding this function to your .bashrc); but I want a proper manager for my tasks. If I could get RTM and something like Taskwarrior to communicate, for instance, I’d be a happy happy man.

Idea: clause depth changer

My style of writing uses clauses, as you’ve probably read before. Clauses are extremely useful for Erkläreideen und Erzählung—as well as just generally preventing the flow of a sentence from jarring to an abrupt halt. Since I’ve begun using the editor vim, the speed at which I write text (I hope) has improved and a few months on, I’m starting to get the hang of some of vim's—I guess you would call them—‘navigational’ shortcuts; these shortcuts are supposed to speed up your typing, though I still prefer using Ctrl + arrow keys over for word-by-word navigation—though not for changing or deleting text, which is where vim becomes very powerful indeed.

Clause depth for me is three-fold: in ascending order of ‘weight’, a comma affects a sentence the least, an mdash affects a sentence a little more and a semi-colon I use to combine related sentences or groups of related clauses; and, sometimes, I’m required to flip the weight of the semi-colon and mdash around—just so I don’t use one in close proximity to another.

Sometimes, it happens that I change my mind on how deep or with what glyph I want to clause (v.) a clause (n.); to go back and change, let’s say, a comma followed by a space to a mdash with spaces on either side would require, even using Ctrl + arrow keys, between seven and nine keystrokes. Couldn’t there be a simpler way to find the last comma-space and replace ‘, ’ to ‘ — ’ while retaining my cursor position in vim? I suppose it could be done with a macro—though I’m far from fluent in vim ‘beyond the basics’ .If you know of such a way, please: I’d be very grateful to know and the offer of a beer is… well, on offer for the most helpful response, though we may have to have a vote to eliminate (most) bias.


Graffiti 2 vs Dell

In this post, I’ll be wasting your time by pointlessly comparing the Palm technology Graffiti to other methods of text input.

Graffiti may be retro but it’s one of the slowest ‘innovations’ I’ve recently gone back to. Don’t believe me?; remember Graffiti as a quick method of Palm input?; remember Palms? I wrote the following passage first in gvim on my netbook, then using the software keyboard on my ancient PalmOne Zire 31—yes; they do have software keyboards—and finally on my Zire 31 using Graffiti 2 as the input.

This expanded edition of John Stuart Mill’s Utilitarianism includes the text of his 1868 speech to the British House of Commons defending the use of capital punishment in cases of aggravated murder. The speech is significant both because its topic remains timely and because its arguments illustrate the applicability of the principle of utility to questions of large-scale social policy.

The above passage, taken from the blurb on the back of my copy of Utilitarianism, took ninety seconds to write in gvim and that writing-out is the quote above. I rewrote-out the passage, using the book’s back cover as the source each time and waiting five minutes between typing each version to reduce memory bias. Taking a stylus to the Palm’s software keyboard took four minutes and 35 seconds (275 seconds). To write the paragraph using Graffiti 2, it took seven minutes and thirteen seconds (433 seconds): 1.6ish times longer than the Zire’s software keyboard and almost five times longer than using gvim on my Dell Inspiron mini. I’m not sure what scientific relevance or use this information has but remember: you’ve read it; you can’t unread it.

Palm Zire 31 and its software keyboard—accessible by tapping ABC in the bottom-right of the main Graffiti area

Palm Zire 31 and its software keyboard—accessible by tapping ABC in the bottom-right of the main Graffiti area

I’ve also been meaning to compare typing on one’s netbook and typing on one’s HTC Desire and its software keyboard (termed ‘Touch Input’ by the Android OS). The same-as-before passage of text took 108 seconds to type on the Desire; this figure versus the ninety seconds it took in gvim on my netbook. Not a huge difference and a lot faster than I was predicting.

So, as predicted, my netbook is the fastest mobile/portable device I have when it comes to typing; possibly slightly slower than my MacBook Pro was (rest in peace) owing to its slightly-more-springy keyboard. My Mac mini uses a Apple wireless keyboard as its keyboard input and, while the keys are a little smaller than those on my netbook and the spring is a little subtler on my Mac’s keyboard, I think I prefer my netbook’s keyboard—though maybe that’s because I’ve been using my netbook as my sole computer for the last fortnight and over the last few months, I’ve seen my Mac mini use dramatically fall as I begin to get more and more comfortable with using my netbook at college and for ‘around the house’ typing and editing.

Just a little bene: Graffiti 2 tries to be more ‘logical’ in terms of imitating handwriting than Graffiti (or Graffiti 1 as I’ll call it here) was. Graffiti 2 introduced multiple strokes per character—for example, a lowercase ‘t’ character requires a stroke down (which types a lowercase ‘l’) and a stroke across from left to right (which changes the ‘l’ to a ‘t’; without the stroke down to begin with, a space would be typed)—and I’m not the only one who considers this change, from ‘one stroke, one character’, less easy-to-use and difficult to adapt to. The interesting thing is is that I never had any contact with Graffiti 1 until February 2011 (when I installed a Graffiti 1 input method on my HTC Desire; simply because of retro impulsion), and the only device I’ve owned to have any version of Graffiti on it—this dear-old Palm Zire—had always been Graffiti 2-based. What probably happened was that I’d forgotten most (if not all) of the gestures from Graffiti 2 and, in relearning Graffiti 1 in those crucial five minutes, it now means until I forget everything again, I’m stuck with feeling awkward and frustrated every time a ‘t’ won’t write the way I ‘expect’ it to. JSYK, a ‘t’ character is written/drawn in Graffiti 1 by stylus-ing a glyph that looks similar to mirrored capital gamma (Γ)—but right-and-down and not left-and-down; left-and-down would produce an ‘f’.

Stopwatch used was an iPod nano (3rd generation). Blurb used as passage-to-be-transcribed was back cover of Utilitarianism and the 1868 Speech on Capital Punishment (2nd edition; ed. George Sher) (ISBN 0-87220-605-X). Control variables: same subject (Kevin Steinhardt), same passage of text; variable variables: input devices and methods (Dell Inspiron mini 10 keyboard with Windows key, Palm Zire 31 using Graffiti 2 first then using software keyboard, HTC Desire using Android 2.2 default software keyboard).


The interruption mark

Unless you’re a scriptwriter, this post probably won’t interest you. I’m a stickler for uniform-looking documents in series: one of the main reasons why I use (La)TeX. Recently, I’ve been writing what are basically ‘design guidelines’ at the beginning of my upcoming coursework publication, Wine Glass Moustache—don’t ask—so, when proofreading it, I have a definiative solution to common problems like how to arrange radio scripts if they’ve got timecodes, or how to cite people’s quotes rather than their publications (in these cases, an omiliography is what I’m after). Today, I’ve been musing on what interruption mark to use; an interruption mark is something put at the end of a line in a script to indicate an interruption. For example, …

KS: I think all religions are, in practice, cul-
JE: Stop right there.

In the above example, I’ve used the hyphen to denote an interruption to my (KS) lines; I think this is the mark I’m going to stick with. Alternatives were things like ellipses, but then the line looks incomplete:

KS: But they are. Just ask the OED: it defines ‘cul…

It looks like I haven’t finished writing the line, at least to me it does—kind of like a programmer’s TODO. [1]

I think I’m going to stay with the hyphen but I’m interested to hear what you think, not just on boring bollocks like how to write interruptions in scripts. Do you have your rules of writing written down?; is it so you don’t waste time having an internal conflict with opposing brain halves?



I write a lot. I write copious amounts of waffly shit, mostly, but sometimes I’ll write for quite a long time—being passionate about the subject helps but so does being passionate for and about writing itself. In June last year, I got ~21,000 words into a book called Waffle (the title was changed to Untitled when I began a podcast of the same name) but the format wasn’t working for me, I stopped writing for a couple of months and when I returned, the content I’d written was not only irrelevant but it gave me an opportunity to see how much of a twat I can be when I’m busy writing / hating people. I’ve had a few other ideas for books, all non-fiction: a transcription of my best AudioBoos or at least a book based on transcriptions of said AudioBoos, a collection of my ND Radio essays (currently in singular, unredacted form at or perhaps some articles authored using a multi-collaborator editor like SubEthaEdit or even Google Docs (not sure what the subject would be or who I would author the work with). Tomorrow, if I get time, I’ll begin writing my most daft idea to date: in 100–120 words, my opinion on the article returned by Wikipedia’s random article engine. Who knows: might even get some of my own.


Garish in colour, high in functionality

Yesterday, I cooked-up a fancy spreadsheet in Google Docs (which, by the way, is pretty good; haters gonna hate) to track the progress of … it tracks the progress of the progress of me writing my Main Document, a term for my seriously-OTT notes portfolio.

I should probably explain this. Rather than actually doing coursework, it’s been observed by some that I seem to spend more time organising the doing of coursework. This spreadsheet (link might be dead) is an example of just that.

Cells containing the number 100 (coloured baby-vomit green, at the time of writing) relate to “100% written” sections of my Main Document, and cells containing the number 0 are empty sections; the other numbers are (similarly) percentages of written-up-ing-ness. “Check” is a simply flag to say “Kevin: evaluate this section’s completeness”; “N/A” is self-explanatory. The Main Document sections which need my most urgent attention are low figures between Christmas (possibly rows 16–17) and this week (half-term; possibly row 25, full of N/As).

The objective of the spreadsheet—that’s if I stick with the whole spreadsheet idea—is that all cells that are meant to be coloured baby-vomit green will be coloured baby-vomit green. To give you an idea of the scale of the task: you see all the complete lack (bar one cell, N15) in the pretentious photograph up there… all those incomplete sections and 0-marked sections total 27,302 words.