Markdown vs Microsoft Word

I’ve begun using a markup language called Markdown recently for almost everything text. The pastime of choice at my workplace is to unnecessarily put text into Word documents, which (in my opinion) slows down productivity and makes text “documents” impossible to search with utilities like grep. It also makes the files much, much larger than necessary.

The text Something super secret is almost 15 times larger as a Word document.

Now, I say “documents” in the above paragraph. There is a time and place for Microsoft Word, as indeed there’s time and a place for Windows, Exchange and other Microsoft technologies. But plain text or text with little formatting does not belong in Word – press releases; notes to yourself; even essays and reports in college, though you should be using LaTeX for academic typesetting anyway.

I use Markdown in almost all my projects, both at home and at work. This blog post, for example, was written in Markdown in my editor of choice vim. My new website at kevinsteinhardt.net is Markdown-based as explained in the FAQ – just Ctrl-F/Cmd-F and search for “Markdown”.

Consider giving Markdown a go. It’s fast and simple to learn and best of all, you’ll be able to concentrate on content rather than trying to figure out the ribbon interface in Microsoft Office.

The iPod shuffle playlist

Some people can listen to the same few songs on repeat all day long and I can have days where all I want to listen to is Fångad av en stormvind or whatever – but that’s a story for another time.

Quite a few of my belongings are still at my parents’ house in Cambridgeshire. Whenever my parents visit me or vice versa, something of mine is almost always given back to me. Sometimes a car full (September); a small something I forgot last time I stayed (November), stuff to occupy the space the Christmas presents took up on the flight there (December); and me and my mum prepared a small list this time (March) complete with a sewing repair kit she made and an extra lock for my bike which would have taken me over my baggage allowance on the flight back at Christmas time. An important item I’ve missed since I broke my HTC Desire in that completely unavoidable(!) jacket-on-floor-meets-foot accident is my age-old first generation iPod shuffle.

I keep a Smart Playlist in my iTunes library based on the last time I listened to a particular track:

Media Kind is ‘Music’
Genre is not ‘Live’
Last Played is not in the last 21 days
Size is less than 20 MB

These parameters produce a playlist of fresh songs that I literally haven’t heard in weeks. The track size threshold is owing to the capacity of my Shuffle, 512 MB: pathetic by today’s standards and expectations but for what I want it for, perfect. What’s going to happen is that I plan on syncing my iPod every evening and re-Autofilling it from the Smart Playlist above. When you listen to music on the Shuffle (or any other iPod), the play counts for the tracks in question get updated in iTunes on every sync – therefore, music I listened to (rather than skipped; which I occasionally do) won’t make its way back onto my iPod for at least three weeks and even then, the probability of it being Autofilled is roughly 1 to 125.

It’s not about forcing myself to listen to music I don’t like; music that terrible either gets deleted from my iTunes library or never makes it in in the first place. It’s always nice to listen back to a song you haven’t heard in a while and today, I came across many an old favourite. If I wanted to be really evil, the playlist would take into consideration the skip counts of tracks too.

Gmail and saved searches? – not yet

I love being organised. Anyone who’s worked with me knows this, and probably because I’ve tried to impose some sort of régime upon them because I know (and I do know) it’ll save us time and effort in the long term. This weekend, like most, I’ve been sifting through my inboxes and also through Google Reader in a task I like to think of as pruning.

I was in Gmail and searching for all unread emails from before I moved to the Netherlands, using the query is:unread before:2011/04/09. I knew I wouldn’t finish properly dealing with all eighty emails retured by that query; so I wanted to save the search for later. But after sniffing around Gmail’s interface for a bit, I discovered that there isn’t a native saved searches feature.

The excellent Greasemonkey tool is built into Google Chrome (something which most people aren’t aware of) so I searched for a Greasemonkey script to solve my problem; I didn’t need to worry about version compatibility or performance.

Well actually, it turns out I did need to. The only decent script I found can be found on this blog post from March 2005. Subsequent updates to Gmail (rather than subsequent versions of Chrome; what I meant by “version compatibility”) have broken this script so I’m still without a solution to my script lust.

But if you I think about it, it’s been almost seven years since that script’s feature on persistent.info and Gmail still doesn’t have a saved searches feature. I’m not sure how many people would use it but it sounds as if there isn’t that much demand out there except from me and possibly a handful of others; so it’s unlikely it’ll ever get included.

“Overweg vrij laten”

A few mornings ago, on the slow bus into Hilversum, I saw an sign which intrigued me at the level crossing between the platforms at Hilversum Sportpark railway station. The flashing traffic sign—two lights on top, two lights below; flashing like those motorway-side FOG signs that always seem to be on when there’s clearly no fog at all—read:

“Overweg vrij laten”

I saw this sign the other day but alas: it was unlit and I took less notice of it. It turns out this sign is lit during rush hour periods, and reads something along the lines of “keep level crossing clear”. Basically, it’s to remind thick drivers not to blindingly follow the car in front onto the level crossing—only to get stuck with nowhere to go, then hit by a train calling at Hilversum Sportpark.

I had to ask a colleague about this sign’s meaning since Google wasn’t much help. It was even less help when I stupidly tried to Google Translate (v.) one result of the search query ‘overweg vrij laten’—the Dutch Wikipedia article on level crossings—into English, since I knew the English article on level crossings wouldn’t have anything at all on vrijs or overwegen. Computer translation, or machine translation, is notorious for failing to take the context of words around a particular word under translation into account. When one Google Translates a website—rather than just a few words or a paragraph—upon hover, small bubbles appear over the translated text, showing the user the original text that was translated but, more interestingly, a “contribute a better translation” link—which when clicked, expands to a simple input field with the English (or rather Dunglish) text ready to be corrected.

First of all, files (pron. fee-lers) are traffic jams; the singular form of the word files is file, perhaps unsurprisingly. Dutch ‘file’ ≠ English ‘file’—as in a document—though file can also mean a file as in a document in Dutch. I’m only at the first word and the machine translation has already proven itself to be context-insensitive. The lead sentence of the ‘crossing equipment’ section—which should really read something like “…”—reads “depending on the level of security, Dutch consider the following items fitted”. The word order isn’t as bad as in some machine translation cases I’ve seen but it’s wrong, for a start; beveiliging, which the translation engine subsititue with “security”, would be more appropriately translated as “protection”, especially since we’re talking about level crossings. The Dutch noun overwegen—‘level crossings’—has been confused with the Dutch verb overwegen—to consider. The word kunnen (lit. ‘can’, as in “Dutch level crossings can be fitted with the following protections…”) has been ignored completely. Again, the word order towards the end of the sentence is getting raped pretty badly. And this is just one sentence in one of millions of Wikipedia articles. Imagine if the other Wikipedias were simply Google Translate’d ‘copies’ of the English-language site; fuck, that would cause a few arguments on the talk pages …in broken English of course.