»   About   VinylDiscography   Archive   Feed  «


While perusing the DEVONthink Pro manual for information on variables (to be used in a Markdown based annotation template) I stumbled on this little gem:

defaults write com.devon-technologies.thinkpro2 RenderMarkdown True

This will make DEVONthink Pro always render Markdown per default and so inverts the Best Alternative (CMD+ALT+P) function. This means that you’ll be required to toggle off the rendering so you can edit a file.

I definitely prefer to toggle off rendering to edit a file over having to toggle it on everytime I want to read a file, since I search and read notes more than I write them in DEVONthink.


I read stuff all the time. In the morning, in the evening and in between, when time allows. I feel most comfortable reading on my iPad and secondly my iPhone (but not for longer periods of time). I have 3 primary sources when I read:

  • .pdf files which I read using iAnnotate on my iPad or PDF Expert on my iphone.
  • the Kindle app for both iPad and iPhone.
  • the Instapaper app which gives me access to my Instapaper account.

With some of the stuff I read (specifically the non-fiction stuff) I do highlights. Now, as I’ve described in earlier blog posts I use DEVONthink Pro to manage all kinds of information, including .pdf’s, web archives, Office documents and what not.

I also use nvALT for jotting down smaller notes, often formatted in markdown, and I’ve configured it to use a Dropbox folder for storage. This enables me to let iOS apps (like Notesy, Nebulous Notes or SimpleNote) access the notes when I’m on the go. As part of this setup I let DEVONthink Pro index my nvALT folder, which will include it’s contents in the database along with all the other files.

I also like to import my highlights to DEVONthink Pro. For the Kindle website and .pdf’s I have Apple scripts which gets the job done. So my Instapaper account was the odd one out.

But just the other day, IFTTT released a new Instapaper channel which enables a couple of triggers.

  • New item saved: Fires when you save an item to your Instapaper account.
  • New liked item: Fires when you like a new item in Instapaper.
  • New highlighted item: Fires every time you highlight a piece of text in Instapaper.
  • New archived item: Fires every time you archive an item in Instapaper.
  • Item moved to folder: Fires every time you move an item to af specified folder.

And one action:

  • Save item: Saves a new item to Instapaper

What’s most interesting to me is of course the New higlighted item trigger. With this I’ve created an IFTTT recipe which appends the highlight to a text file in my nvALT folder. The filename is created if it does not exist, and is named by the article title with ‘-InstaHigh’ appended. The format of each highlight goes like this:

##Daring Fireball: Only Apple
>"Google’s mindset a decade ago was centered around web apps running in browsers. Google didn’t need its own platform because every PC had a browser and people would use those browsers to do everything Google provided in browser tabs. That meta-platform approach has limits, though, particularly when it comes to post-PC devices. Their stated reason for buying Android wasn’t because they wanted to design and control the post-PC device experience, but because they wanted an open mobile platform on which their web services could not be locked out." 

*on June 17, 2014 at 10:47AM*, [via Instapaper]( 

which will render it like this in html (of course depending on the stylesheet):

Daring Fireball: Only Apple

“Google’s mindset a decade ago was centered around web apps running in browsers. Google didn’t need its own platform because every PC had a browser and people would use those browsers to do everything Google provided in browser tabs. That meta-platform approach has limits, though, particularly when it comes to post-PC devices. Their stated reason for buying Android wasn’t because they wanted to design and control the post-PC device experience, but because they wanted an open mobile platform on which their web services could not be locked out.”

on June 17, 2014 at 10:47AM, via Instapaper

This way, the highlights of my 3 reading sources all go into DEVONthink Pro, either through Apple Scripts (.pdf’s and Kindle) or my nvALT folder on Dropbox.


  1. Use a Dropbox folder as storage for nvALT
  2. Activate below recipe and modify to suit your needs
  3. Ensure that DEVONthink is indexing the nvALT Dropbox folder

You can get the recipe here:

IFTTT Recipe: Append Inst. Pap. Highlights to nvALT folder in Dropbox connects instapaper to dropbox

Continue reading →


As mentioned in a previous blogpost, I use DEVONthink Pro (DTP) as my Information Management system. Basically, this means that I throw all kinds of information, in all kinds of formats, into DTP for indexing and archiving. Be it blogposts, .pdf articles, Word documents, Excel spreadsheets, Powerpoint presentations or plain text files. DTP does it all.

This post won’t cover DTP that in-depth, but rather the proces it is part of when actually producing some kind of writing based on, or including, the information I keep in DTP.

And that’s going to introduce yet another Mac app favorite: Scrivener produced by Literature and Latte.


Now, Scrivener ought to, I imagine, be a writers idea of heaven. I don’t write novels or long form literature. But I have spent the time required in Word to handle 40 to 50 pages more than once. And honestly, it was a pain. My top 2 of personal annoyances include:

  • being forced into a sequential writing process; sure, you can copy/paste and see an outline. But moving stuff around? Better not hope it’s too complex.
  • the inherent focus on form, not content. I find this even more stressful than looking at blank sequential pages. This might be personal, but I think Microsoft Word sits squarely between being a wordprocesser and a typesetting tool. I dare you to use it with a focus on content alone! So for me, I prefer using Word as a typesetting tool in the very end of the process.

Not so with Scrivener. In my perception, Scrivener is all about the content. Sure, you can muck about with fonts and margins. But it’s not something that jumps into your face or is required. Choose a font. Make your headings bold. Write.

Scrivener main window

I’m just going to explain the primary parts of the interface along with the ‘compile’ feature, to give you an impression of Scrivener. It won’t be in-depth per se, but should give you a general idea and inspiration as to trying the app. Yup, free trial at the Literature & Latte site.

The Binder

On the left hand side is the Binder. Basically, it immitates how a regular paperbased binder would function. You can create new collections (also based on search criteria) with different colors to supplement the ones you see on my screenshot (Updated Documents, Search Results and Binder).

Below the collection titles you can see the contents of the selected collection. And this is one of the cases where you have to think differently from Word. Because what’s presented here under the Draft top level, are Scrivenings or, pieces of text, with a title. As you can see, these can be placed in a hierarchy. But the really neat thing is, that they can be dragged around as well. This makes toying around with the contents damn easy.

Oh, initially, the Binder will be split into two distinct parts: Draft and Research. The Draft part is where you will put your text, you product, so to speak. The Research part is where you can put images, .pdf’s, Word documents and other goodies to keep them close while writing.

The Text Window

In the middle, we have the text window. And this is where it gets special. Because the text window has 3 different view modes:

  • Scrivenings which is what you see on the screenshot above. This window can show one scrivening or several scrivenings in one long view. See where the cursor is at in the screenshot? This is where the first scrivening ends, and the next begins.

  • Corkboard which is exactly what you might think. It’s a corkboard where the individual pieces of text are represented with their title and outline. I use this a lot for brainstorming and shoving around different scrivenings, it make it really easy, and a lot more visual, to discover the right contents and sequence of your sections.

Corkboard in action
  • Outline Is exactly what it says. It’s an outline which also shows the Synopsis, Label and Status for each part of the outline. Each part translates to the Scrivening you have selected in the Binder. You should know Outlines so I will not go into detail about that part.

Now, a neat thing about the text window is that it can be split. I use this a lot when I am writing parts that are based on other parts of the work I’m doing (for example, I find it useful with a whitepaper to have the Introduction in the top and the Summary/Conclusion in the bottom). Might also be research material in one of the windows.

And guess what, it’s your choice (Your choice, your choice, peace, or annihilation - sorry, the A Perfect Circle cover of this excellent protest song was on the turntable while writing this) whether the windows you have split into are in Scrivenings, Corkboard or Outline mode. Fancy an outline or corkboard in the top window, while writing in the bottom one? No problem, Scrivener’s got you covered.

Corkboard and Scrivenings split

The Inspector

On the right hand side, is the inspector which, in classic Mac style, can show many different types of information (the contents of the Inspector changes with the selected Scrivening). Standard has the Synopsis on top, followed by a General part which keeps track of Label, Status, Modified Date, wether it should be included in the compile, if there’s a line break before (when compiling) and if it should be compiled ‘as-is’.

Below the General part, different views can be selected: Notes, References, Keywords, Custom Meta-data, Snapshots and Comments & Footnotes. Going in-depth with these will explode the scope of this post, plus, they’re pretty self-explanatory.

Of course, the Binder and Inspector can both be hidden, leaving the text window the only window.


Remember when I said that Scrivener focusses on content more than form? Now, with all the different view modes and the ability to shift your scrivenings around, Scrivener uses its own file format, to keep all the relevant data.

Also, I do recognize that typesetting is important for the final product. It’s just that I would not find Scrivener to be the best tool for this. I’d rather use Final Draft, Word or Open Office for this part.

So how to actully export your material from Scrivener? The answer in the world of Scrivener is: Compiling. As seen in the screenshot below, there are indeed a lot of options available, when performing the compilation of your writing.

Compile time!

Scrivener supports an impressive array of formats, from Plain Text, RTF and Word and Open Office variants, to Final Draft, ePub, Kindle and iBooks. Add in Multimarkdown support which can be compiled to LaTeX, HTML and XML and you’d be hard pressed to require further formats.

A writing and typesetting process

So, that was an introduction to Scrivener. Not overly deep but it should be enough to whet your appetite and send you off to download the trial version.

I am sorry if what got you reading was my mention of DEVONthink Pro, because there hasn’t been much of DTP so far. But let’s outline my actual process (please note that it’s iterative, it’s not a one roundtrip sequential thing):

  • From DEVONthink to Scrivener is the first part. And this part can really take a while, since this is where I identify the relevant material to be used. Sometimes it’s PowerPoint presentations, other times it’s annotated pdf’s or the annotations themselves that I go through (and that’s where my previous workflow really shines).
  • Writing in Scrivener using the outline and research material gathered in the previous step. This often leads me to do more research or to remove materials from the Groups and Smart Groups in DTP as I get more into the specifics. I also use Zotero in this part, since it’s my bibliography manager, so I use it for citations.
  • Citation scanning with Zotero comes next. Basically, I compile the as RTF from Scrivener and then use the RTF Scan feature of Zotero to create the proper citations and the bibliographic reference.
  • Typesetting with Word is the final part, as I haven’t concentrated on the looks of the document so far. Setting fonts, margins and sizing images correctly is all a part of this.

Sounds great, right? Let’s dig in for some more details.

From DEVONthink Pro to Scrivener

As I’ve mentioned in my other post and in the introduction of this one, I use DTP for my collecting and managing my research material. So how do I get from a big collection of all sorts of stuff, to actually using it in my writing? The answer to that is using Groups and Smart Groups in DTP. Basically, when the subject of my writing project has been decided (like the whitepaper on Agile Software Development you’ve seen mentioned in the screenshots; hint, it’s in danish), I start searching and organizing my data in DTP.

This gives me a rough sketch of the outline of the writing project. And then I create an outline with Groups and Smart Groups. Groups in DTP are basically folders you can use to structure your content. And when I say Smart Groups, think Smart Groups in Mail. You punch in the search criterias and DTP will create a Smart Group folder with the contents that matches the search criteria.

In this screenshot you can see a very simple example:

DTP Smart Group

This will produce Groups in DTP with contents that matches my searches, and then I start pruning the results that come up. So basically, I use a combination of manual searches1 and smart groups to locate the juicy stuff. This is also where the strength of DTP’s See Also function comes in. The information quests in DTP can be pretty damn awesome as the See Also functionality guides you through all the material you’d long forgotten. Ah, serendipity incarnate!

Now, next up, is the decision to drag’n’drop or import the research that I find relevant to Scrivener. And when I say relevant I mean the stuff where having it in a split window, will provide a benefit while writing. Otherwise, I find that I might as well just CMD+Tab to DTP. It’s also worth remembering, that the material you choose to drag over to Scrivener, will increase the filesize and save/export times (not that they’re crazy or anything).

Writing in Scrivener

As I’ve mentioned, when I do my writing in Scrivener, it’s sometimes with split windows, other times I use the fullscreen distraction free writing mode. It all depends. When I encounter a part where I need to put in citations, I CMD+Tab to Zotero if I need to remember the author, or typically, the year. Then back again to Scrivener and then I write the citation like in this form which will be recognized by Zotero when performing the scan: { Boehm, 2004}.

There are pretty good guides out there showing how to do it in different ways, like this one or this one, but since I’m not a historian doing very complex research and loads of citations, I find the above way to be the easiest.

UPDATE: The lovely thing about Twitter (and the internet): It’s the perfect liquid network! Not long after publishing this post, @s_margheim tweeted that I should try his ‘Zotquery’ plugin for Alfred, which makes querying and getting cites from Zotero a blast. You should definetely check it out if you’re an Alfred user, it really rocks. Start right here at his Hackademic site which has loads of good stuff. Also, if you need more control and details in how your cites are put together, this Zotero plugin helps you get exactly that2. Instead of scanning RTF it works with ODF, and one of the good things about that, is that it’s available to all through OpenOffice or LibreOffice.

Citation scanning with Zotero

Once I’ve exported the file in RTF format from Scrivener, I use a feature of Zotero called RTF Scan. And, not surprisingly, it does exactly as you’d think. It scans the document you select for ‘codes’ like the one above, substituting them with proper citations (whatever citation style you prefer, Zotero should support it and inserting a Reference list in the end of the document. A neat thing is that Zotero will save the results to a new file of your choosing.

Zotero RTF Scan options

Typesetting with Word (or OpenOffice or Pages)

Now comes the final step. Typesetting. Yup. Typesetting. Sometimes I enjoy it, but most times I find that it actually facilitates [procrastination][14]. I’ll simply start diddling with the form, instead of producing content. But the beautiful thing about this process is that all you need is an application that can read RTF. So you can use Pages, Nissus Writer, OpenOffice, Libre Office or Word, for that matter.

Of course, if you want to, Scrivener can help this proces along with some of its features when you ‘build’ the compile. But that’s a story for another day!

Now, seriously

I honestly do recommend Scrivener, especially for more technical pieces, where you need to throw things around or go about your writing in a more complex way. For me it enables a different way of working with the text and its structure.

One last neat tip, is that Scrivener can automatically export your individual Scrivenings when you quit the application. I use that to put the text into a folder in my Dropbox account for use when I’m on the go with either my iPad or iPhone. And what happens when you fire Scrivener up next time? It actually syncs the changes you’ve made. Enough said, go try it out !


  1. From the Search Results window you can drag items directly to the group and create Replicants using ⌥+⌘ while dragging.

  2. {See | Smith, (2012) |p. 45 | for an example |zu:2433:WQVBH98K} is an example that’ll turn into (See Smith, 2012, p. 45 for an example) once scanned.

Continue reading →


Read this post at the New York Times ‘Fashion & Style’ section and I think it brings a nice and critical perspective to the way some people use Twitter. Or, in fact, many other social media.

It can sometimes be really difficult to get the grasp of a social media and it’s inherent etiquette, but like I mentioned yesterday, keeping it real with integrity is paramount. You will not get succes by spamming people with tweets, favorites, likes or whatever currency the given soclal media has.

Some quotes stood ou to me, and should be part of any basic introduction to social media:

But a random appeal to Twitter royalty is only one of several gambits by which users of the site blur the lines between good and bad manners. Some Twitter users, on seeing that something they’ve tweeted minutes ago is gaining no traction, will wantonly fire off a spray of retweets or Favorites (similar to the Like option on Facebook) in a desperate attempt to prompt reciprocity.

But if there’s no algorithm for determining when self-promotion has crossed the line, are there any general principles to be considered? Two come to mind. First, self-promotion becomes unseemly when it is viewed as repetitive. Your followers probably are willing to tolerate two or three newspaper reviews of your new monograph about combat ethics of the Boer Wars, but not 17. Second, readers are turned off when they’re made more aware of a tweet’s strategy than its content. We were so wowed by the fact that you’re sending the tweet to @madonna and @GwynethPaltrow and @RinglingBros that we lost the fact that you just secured your first booking as a yoga clown.

Ms. Li suggested a third metric: usefulness. “Your content has to be useful to people,” she said. “If it doesn’t have value to your followers, then it’s seen as spam or self-promotion.” Under this guideline, Rod Stewart’s tweet about having a child from each decade would seem to justify its swagger, because it’s information that his fans can share with others, or perhaps bait their grandparents with.

My takeaways on publishing your content on social media:

  • Trust in the quality of your content
  • Don’t push yourself in front of peoples faces
  • If people did not react the first time, spamming will not help


Interesting piece by the venerable Matt Gemmel:

I don’t think there’s any reasonable way, or any need, to separate vanity and ego from a personal blog. Writing is inherently about its author, and is a product of their personality and opinions – that’s not something to be shy about, and we shouldn’t try to change it either. So, write for yourself – and hold yourself to an appropriate standard, because you’d better believe that others are judging the person as well as the piece – but as soon as you publish your views, you’re inviting readers to take a look. I think that the needs of the reader and the author are more aligned than many blogging systems seem to believe.

I haven’t really thrown myself 100% into the art of blogging, but I’m pretty sure that if I do, it has to be real. Not that it isn’t now, but I don’t spend much time on it, resulting in a few bursts of posts every now and then. Which means I also have very low expectations on readership. But quality still matters the most.

And I know, from the blogs I follow, that I need a nerve that shines through, giving me a sense of who the person behind the blog is, for it to be really interesting. Otherwise, I will probably know sooner or later from someone in my RSS feed.

There’s plenty of sites out there, just bouncing the echo of the few and talented, as it is. So be real (even if that’s boring), for the sake of integrity and the readers you might have accumulated. And if that mens no readers, so be it. You don’t want Trent Reznor calling you ‘a copy of a copy’ (because he’s badass).