Recall These Thoughts
Inanity from Scott Lougheed, PhD

Comedian and late-night host John Oliver devotes a segment to food waste. The whole thing is just damned funny and damned…. damning. I particularly enjoyed (in my own cynical way) the footage of food in landfills:



However, agreeing on the existence of a problem never automatically led to solutions, which has become painfully evident in this case. Despite the issue’s uncontroversial character and seemingly lack of political sensitiveness, food waste has long been a political non-issue. Everyone seems concerned but no one seems to take any action to deal with the problem.

It turns out that learning about and targeting food waste leads to a deeper understanding of what is really wrong with our food system, an insight which severely weakens assumptions about increased agricultural yields as the solution to future food security challenges. I will come back to this point in a later post about the “Human right to cheap food”.

– Helena Robling at Food Policy for Thought

This is a very interesting insight from Helena Robling at the Food Policy for Thought blog. I think she is definitely correct that, in many respects, waste in the food system is a political non-issue. There are a large number of possible explanations, but one potential factor is that, to an extent it is the very politics of food production (such as subsidies for certain crops) that contribute to wasting (For some more on this precise issue see for example: Gille, Z. 2012. From risk to waste: global food waste regimes. The Sociological Review, 60, 27–46).

In my own research on food recalls, there are also issues of brand reputation and protecting profitability that can sometimes already be marginal under the best of circumstances. Destruction is often the path of least cost and resistance. Moreover, in most cases companies are reluctant to disclose just how much product was recalled and subsequently destroyed. 


I suspect a quick Google search will show a few thousand results comparing Evernote and DEVONthink intended to help you decide which one to use. However, I think it is important to explore the differences between these applications not for the purposes of choosing one over the other, but for understanding the different needs that each application offers. It is not an either/or situation. I chose these two applications because they are very frequently referred together or suggested as alternatives to one another. However, they are not alternatives, they are not replacements, they are different products with different purposes and uses. And while I’ve chosen to compare these two, there are plenty of other options, Microsoft’s OneNote and Google’s Keep are two prime examples of options that are also along the lines of Evernote (and less like DEVONthink), as you shall quickly see.

I originally set out to write an elaborate post that described how I use both applications, with the hopes that such an explanation would reveal the differences and similarities between the two applications. That quickly proved cumbersome. Instead, I have produced a pair of short DEVONthink/Evernote posts instead. The two posts highlight some salient differences between DEVONthink and Evernote. The posts are formatted as pairs of contrasting sentences with elaborations below, focusing on a short list of key areas chosen off the top of my head (so not at all systematic or exhaustive). These posts are not instructive, I’m assuming you either have existing knowledge or are willing to look up how to do these things in the documentation.

Part I focuses on some critical functional differences with some important implications. Part II focuses on features that are largely shared between the two applications but which are implemented in different ways.

Evernote allows you to create one type of document

basic formatting in evernote

Evernote allows you to create “Notes”. Notes are documents that contain formatted text, tables, images, and links. On the surface they look similar to a rich text file (under the hood, they are written in ENML which is largely HTML). You can also attach files of any type (except for a few gotchas, like .pages files) that can be opened in a suitable external application. Evernote notes and the (not entirely) basic formatting they offer are great, low-fuss ways to get ideas down quickly. You can actually generate some fairly nice looking documents very quickly in Evernote using their native Note format.

I rarely actually create Evernote notes with the intention of adding text. In most cases I am attaching files for storage and retrieval. Most notes I do make are quick jottings, such as the weight of each of my cats. Heavy-duty writing is done in Microsoft Word (yes, yes, I know), and depending on the project, is indexed by DEVONthink (more on that below).

DEVONthink allows you to create multiple document types

Standard document types in DEVONthink

DEVONthink can natively create a variety of documents. These include plain and rich text, markdown, basic spreadsheets, and even HTML documents that it can display as the source code or render as a web page. These can all be viewed and edited right in DEVONthink. There are also a number of advanced “templates” that come with DEVONthink, and you can make your own. For example, some default templates include a Cornell note document, a contact card with a link to that card in your Address book, or a blank task list you can add to. The “Project” template shows off the power of templates in DEVONthink the best. As you can see from the screenshot, it creates a master “About this project” file and automatically generates sub-directories for various project elements that you can populate with your own project files as you being your planning.

A more advanced template in DEVONthink. This is the pre-installed Project template. You can see the pre-generated file structure to the right, ready to be filled in!

Above is a more advanced “template” in DEVONthink. This is the pre-installed “Project” template. You can see the pre-generated file structure to the right, ready to be filled in. DEVONthink is not so much about easily creating pretty looking notes natively, though between HTML, Markdown, and Rich Text, you can do a fair bit natively. What lacks in its native ability to create pretty looking documents, it makes up for in its ability to handle almost any type of file you could possible need access to.

Evernote handles files as “attachments” to notes

If you want to store an arbitrary file in Evernote, such as an image, PDF, Word document, and so on, you have to first create a Note, then attach the desired file. Attachments can only be accessed via Evernote and cannot be stored or easily accessed from the Finder or opened via the suitable app. If you attach a word document to a note, for example, you cannot open that document from the Finder or Word, you must open it from Evernote. This can cause some grief if you have an attachment open, close Evernote, and make further edits to the attachment. Because Evernote is no longer running it will not sync the changes to the server, and it is possible though rare, that if you make a mis-step here to lose data. A very crude analogy would be to think of Evernote notes as email messages with attachments.

DEVONthink can handle any files type on its own

You can add any file you like to DEVONthink. It will display a wide variety of files natively, allow viewing of an even wider variety of files beyond what can be edited in DEVONthink itself. Any file that cannot be edited or viewed in DEVONthink will open in the appropriate external application. A contrasting crude analogy would be to think of DEVONthink as a file system, where files stand on their own, rather than email messages with attachments.

Evernote defaults to synchronizing your content

Evernote is, in some respects, a cloud service with a robust client (i.e. the application you have running on your Mac), rather than a product. Much of what makes Evernote great is done server side (making images of text searchable, for example). As a result of this, Evernote works best if you synchronize your stuff, which is the default state for everything in Evernote.

You can create “local notebooks” which, as the name implies are only stored locally. This can be good for some sensitive information but I don’t trust this system. It seems all too easy to uninstall your Evernote client (say, when troubleshooting or setting up a new machine) and lose local data, because you failed to export it prior to uninstalling. It is also possible to accidentally place sensitive information into a synchronized notebook or to have your local notebook actually accidentally uploaded to servers (and there have been reports to corroborate my suspicion of both of these happening). In either case you’ve either lost important data, or uploaded sensitive data to places you shouldn’t upload to. Neither of these are risks I’m interested in taking.

The synchronization in Evernote is great. It is so straightforward and works so well (most of the time). One major perk is it makes sharing simple. My partner and I share a few notebooks. For example, we have a “household” notebook containing all the documents related to our household affairs, such as receipts for joint purchases, financial planning, vacation planning, and things related to our cats (we recently kept a photo journal of a minor ailment afflicting one of our cats. Being able to both contribute to that journal daily without fuss was truly amazing). We also share a notebook with recipes for cooking. It is easy to scan printed recipes or clip them using the web clipper and have them immediately accessible on our iOS devices in the kitchen or do quick searches for “rhubarb” or “Cookie” (we use tags to assist with this too) if we need inspiration or ideas. All of this is so easy and is fueled by Evernote’s synching.

Synchronization is less great when in comes to very sensitive information. Because of this, there is a lot of information I do not, or cannot store, in Evernote (and I do not like the opacity and unreliability of Local Notebooks).

DEVONthink does not synchronize content to any cloud by default (or at all)

DEVONthink’s synchronization capabilities are more diverse and flexible than Evernote’s but they are a bit less straightforward. DEVONthink doesn’t, as yet, offer any seamless synchronization, and does not offer its own cloud service a la Evernote. The major disadvantage is that it makes sharing or collaborating as I described doing with Evernote impossible, or at least nearly so nearly so. DEVONthink does provide a system for synchronizing databases directly between two macs, and this can be facilitated with a service like Dropbox, directly over your LAN, or via your own webDAV server. It is great if you have a home and work machine, or a desktop and laptop. It does not aid in collaboration. Collaborating with DEVONthink’s sync options can be done, and the folks at DEVONtech outline one way of collaborating with DEVONthink and Dropbox on the Devonian Times blog.

The major advantage to having files stored locally is that I always know where they are all the time. There are no sync conflicts. There are no errors. Most importantly for me, I have full control over my data. My research data has personal information about third-parties. My university prohibits me from storing that in cloud services (with a few small exceptions, and Evernote is not one of them). The province of Ontario also has legal limitations on this. It is therefore important for me to have large portions of my research data explicitly NOT in the cloud. Thankfully all of the amazing features DEVONthink offers take place locally, so this is not an issue for me.

Evernote maintains its own, cryptic database to store your content

Evernote stores your data in you computer’s ~/Library/Containers directory, and it is in a database format that is pretty human-hostile. You can get into it, you can view files, but it is organized in a way that makes zero sense to a human. It was not built for you to wade into. Besides, if you mess with the database in the Finder, you risk corrupting it, potentially losing data, and so on. This means you have to always do everything via Evernote’s interface. This is not itself an issue. Evernote has its little data silo and you modify the contents of that silo through the Evernote interface. Everything only ever lives in that silo.

A major consequence of this is you cannot “reference” files. That is, files only ever live in Evernote’s database. Evernote cannot see things outside of this database, and you can only view the contents of the database through Evernote. You cannot, for example, have an Word document that is stored in your ~/Dropbox folder attached to an Evernote note.

Another consequence is that editing attachments after Evernote has closed can be problematic. As mentioned above, modifying an attachment after Evernote has been closed can potentially result in data loss since Evernote loses track of what is current or what has changed.

DEVONthink can store content in a database or reference files in your file system where they lay, or both

DEVONthink has its own databases, not unlike Evernote, however there are some important differences. The first is that files you organize using DEVONthink don’t actually have to be stored in these proprietary databases. You can “import” files, or “index” files, or both. Files you “import” into DEVONthink live in these silos along with all of the extra metadata that DEVONthink generates about these files to fuel its “artificial intelligence” engine that helps your organize and find files. If you “index” files or directories, the files remain where they lay in your file system (so, if you index a folder in Dropbox, the files live in that dropbox folder), while all the additional metadata that DEVONthink needs live in its own database. Changes made to files in that directory inside or outside of DEVONthink are automatically reflected both inside and outside of DEVONthink.

These PDFs live in a folder in Dropbox, but are indexed by DEVONthink, and thus, available through the sophisticated search features and for processing by DEVONthinks AI

For example, in the image above you see a list of PDF files. These PDFs live in a folder in Dropbox, but are indexed by DEVONthink, and thus, available through the sophisticated search features and for processing by DEVONthink’s “AI”. These files have little arrows next to them indicating that they are “indexed”. These files don’t live in DEVONthink, they actually live in a folder in Dropbox. I can use DEVONthink to read and annotate these files and they remain in Dropbox, and all the changes are immediately reflected in Finder and Dropbox. This is exactly the same as if I had just opened the PDF in Preview.app and made changes there. Likewise, if I read and annotate one of these PDFs on my iPad, the changes are synced via Dropbox and immediately reflected in DEVONthink. This allows me to take advantage of all of the searching and organizational features in DEVONthink, without rendering my PDFs inaccessible from other devices. Imported files live in DEVONthink’s databases and need to be accessed via DEVONthink, though they can be opened, viewed, and edited, in any external application you like, even if DEVONthink is closed while you have a file open in an external application. Unlike Evernote, there’s no harm in modifying files that are imported into or indexed by DEVONthink after DEVONthink has closed since there is no threat of a sync conflict.

Evernote Maintains a single database, stored in a hidden location

The major downside here is that this is also where your local notebooks live, which makes them very difficult to move around between machines if you are setting up a new Mac for example. You have to go through a slightly onerous export-import routine… and that assumes you remember to export and backup your local notebooks before you nuke your previous installation! You can also only have one Evernote database. This means that you cannot have, say, a “Work” and a “Home” database. While you can use Notebooks and Stacks to differentiate these things, it is not always ideal and can clutter searches. If you want full separation, you need separate Evernote accounts. This can mean that, once you accumulate a lot of stuff, it’s an all-or-nothing deal and extremely large Evernote databases can sometimes be slow. And while Evernote’s searching functionality is robust, it can be hard to add all of the exclusions needed to ensure you don’t have a lot of work cruft when searching for a recipe for dinner.

DEVONthink allows you to create any number of databases, stored where you see fit, and which are portable

With DEVONthink, I can create any number of databases. This can help mitigate slow-downs due to large databases (you can split databases at logical points), and isolates contexts making sure searches are focused and that the AI that helps categorize or surface related content is bringing up meaningfully related items.

I don’t need my receipt for DEVONthink to come up as a search result or as a “See Also” suggestion when I’m looking for academic literature. Having my personal files in a separate database means that I don’t have to string together a slew of exclusions in my search terms, nor do I have to sift through false positives.

EVONthink Database files in a location chosen by me, and ready to be moved around at any moment!

These databases also live in any location. I store them in ~/Documents/DTPO. I can move them to a new location if I want with zero consequence. It’s also trivial to move items between databases should the need arise, such as if you want to split a database that is getting too large or if something suddenly shifts contexts. Another advantage: I don’t need my multi-gigabyte academic literature or government document databases open when I just need to quickly consult my comparatively small personal database. This means the application loads quickly and requires fewer computer resources when all I have is a small job.

One advantage of this is that, while you cannot restore parts of databases easily (as was also the case with Evernote), restoring entire databases, or multiple versions of entire databases from backups is very easy. This means, and unlike with Evernote, you can restore one or more versions of a database from a backup. You can also have these restored versions open simultaneously. This means if you are trying to find a past version of a single file, you can have your current database open, along with a past version, and transfer only that file from the backup to the live database. Put another way, backups are much easier with DEVONthink’s system, and so is restoring.

That’s it for Part I. Part II will be posted in a few days. Stay tuned, and try not to let the suspense get to you in the meantime.



Lots of insight in this article.

Yet, looking at the above eight outbreaks together, I find it a bit hard to parse out why some have been targeted – OK, perhaps the Parnell prosecution is a bit easier because it was so clearly intentional – and some have not, or at least not yet.

Honestly, what are the differences in prosecuting the Jensens, DeCosters and ConAgra and leaving the others – so far – unmolested by section 402(a)(4) of the FDCA? Is it the number of sick, the number of dead? Is it the economic consequences? What really are the criteria, or, should it simply be left to the discretion of the prosecutor as to who or what feels the sting of the criminal justice system?

Bill Marler, Marler Blog

Striking is the seemingly arbitrary decisions about when to prosecute and when not. Short of the Parnell brothers and the Peanut Corporation of America scandal, it seems like throwing darts at a board. 



My friend and food studies colleague Sarah hasn’t had Tutta la Vita online for too terribly long, yet it has to be one of the most well done food blogs on the internet. Beautiful photos, exquisite web design, and creative recipes make for a pretty top tier combination. Sarah has a Master’s degree in Nutritional Science and has a very down-to-earth approach to food. No hard-to-swallow pseudo science, no stagnant and boring “crash diet” crap either. Just good wholesome food made with real stuff. 

I strongly recommend checking out her Pomegranate & Hazelnut Chocolate Bites. We made them for guests recently and they really were an impressive show-stopper. 

So go check it out! This is a blog—and a blogger—to keep an eye on.


In this entry I talk about how I use OmniOutliner and DEVONthink Pro Office to summarize academic literature to lay the foundation for a literature review (or any type of written work, for that matter).

Setting Up DEVONthink

To begin I need to explain a little bit about how I organize my academic literature. All of my academic literature is roughly organized into folders for broad subjects and which are stored in Dropbox. This is primarily in order to give reasonably convenient mobile access to these files in the event I want to read them on my iPad (PDF Expert 5 is my PDF reader of choice on my iOS devices). I can set a given folder to be available offline in PDF Expert, avoiding having to keep my multi-gigabyte PDF library on my iPad.

I then set DEVONthink on my Mac to “index” the root directory that contains all of these subject folders and their contents. In DEVONthink parlance, to “index” is in contrast to “importing”. Importing into DEVONthink means that files are kept in special DEVONthink databases that are not readily readable outside of DEVONthink, and so works well if you don’t need regular, easy access to these files on outside of DEVONthink (DEVONthink Server, included as part of the Pro Office version, however, could work for some use cases though). Indexing, on the other hand, allows you to reference files as they live in your computer’s file system. This means I can have my PDFs in folders in Dropbox, and still view them in DEVONthink and take advantage of various organizational and search features in DEVONthink. So in DEVONthink, I end up with a listing of all those subject matter directories and their PDF contents:

DEVONthink Pro Office showing the root directory with all of roughly organized folders
DEVONthink Pro Office showing the root directory with all of roughly organized folders
Tags in DEVONthink Pro Office
Tags in DEVONthink Pro Office

From here I can use tags to impose a more elaborate and in my opinion, useful, organizational system on these that doesn’t mess too much with their organization in Finder (and thus, does not impact their accessibility on my iOS devices!). I’ll write more about this in another entry, but just to demonstrate, here is a segment of my tag list in DEVONthink. This is my primary way of organizing in DEVONthink and offers much more flexibility than the rough folder categories I described above (again, that is almost solely inservice of offering easy access on my iOS devices). Example of my sloppy tagging to the right.

Setting up OmniOutliner

This is a much less sophisticated task compared to DEVONthink. If you aren’t already familiar with OmniOutliner, familiarize yourself with its features and its functions. Now, you don’t have to use OmniOutliner in particular for this task, but I do find it to be one of the most intuitive and powerful outlining applications. Using a text editor or word processor for outlining has, in my experience, made what is an essential part of writing into an onerous task. OmniOutliner makes it a breeze. If you have a favoured outlining application of your choice you’ll likely be able to translate some of these ideas into that system (and likely have an effective system of your own). Once you’re familiar with OmniOutliner pick a template or create a template that suits your writing needs. I’ve built my custom template for outlining paper and literature reviews which I’ll link to, but its no major chore to make your own.

The workflow

Now that we have our files in DEVONthink I’ll outline the workflow. Essentially the purpose of this workflow is to create a single document with paraphrased passages and quotes, with direct links back to the original document for future reference (and proper bibliographic reference when I finally write it up). The idea is that by the time I’m done in OmniOutliner, I should be able to take the paraphrased passages, incorporate any of the extracted quotes that are essential, and transfer them to my word processor to generate a more-or-less complete literature review to stand on its own for my reference, or to plunk into a manuscript. Once it is in the word processor it should only require some polishing/fleshing out, and proper bibliographic references (I use Papers 3 as my reference manager currently, though its been a long road with lots of reference manager casualties along the way. This subject deserves its own post).

Step by step

  1. Open your PDF in DEVONthink’s PDF viewer. It is somewhat important that you read using this viewer rather than in Preview for reasons that will become apparent shortly.
  2. As you encounter important passages select them and copy their contents to your computer’s clipboard.
  3. Move to OmniOutliner. At this point I do one of two things depending on what I’m working on. I’ll either write a passage paraphrasing the quote I’ve copied which will eventually serve as the actual contents of the literature review, or I’ll write a keyword or concept that the quote pertains to. I usually do the former when I’m at a more advanced stage of writing, while I’ll do the latter when I’m making a literature review simply for my own reference or in the earlier stages of writing.
  4. Create a new outline item and press tab to make it a child of the paraphrased text you just wrote. Then “paste and match style” (shift-opt-cmd-v) to make sure the content matches the formatting of your OmniOutliner template.
  5. Wrap the text in quotation marks and fix any nasty line-breaks and hyphenated words if you’re nit-picky like me.
  6. Return to your PDF in DEVONthink, copy the page link for the page you are currently viewing (shift-ctrl-opt-cmd-c).
  7. Return to your outline, add the link as a “note” for the quote you just pasted (cmd-‘) with the page number in parentheses for your reference.
Early stages, not much paraphrased content in this example, but it should convey the idea
Early stages, not much paraphrased content in this example, but it should convey the idea

At this point you should end up with something that looks more or less like the image on the right.

The major advantage of using OmniOutliner for this task is the ability to easily move things around. During the course of my reviewing, I’ll change the order of things, create new sections, combine previously separate sections, and so on. Using a regular text editor or word processor for this task can be a bit painful once you start having copy/paste or drag errors, lose your hierarchy and so on as you clumsily shuffle things about. These are total non-issues in OmniOutliner.

Repeat this process until you’re satisfied and feel ready to create the formal literature review in the word processor of your choice. Take some time to use OmniOutliners organizing capabilities to your advantage here. Make sure you have all of your things in the order you want them to appear in your final document. It’s much easier to move these things around in OmniOutliner than it is in your word processor.

If you’ve been diligent with your paraphrasing, if you’ve made sure to use page links for every quote and paraphrase, and you’ve used the easy drag-and-drop organizing to make sure things are in a logical and sensible order, this part should be easy. You should, with little effort, be able to dump your paraphrased stuff (through exporting or copy and pasting) into your word processor, adding the references with your reference manager as you go along, incorporating quotes where needed. You should now have a reasonably well organized, fully-referenced not-totally-rough draft of your literature review. Obviously at this point there is still work to do to make it publication-ready, but you’ve done a lot of the hard work already.

Additional notes

I generally keep my OmniOutliner-created files around forever since they are great reference material. It is also easy to copy-and-paste nodes or entire trees into other OmniOutliner documents if I’m in need of it in another OmniOutliner document. You might be asking why I put the Page Links into a “note” for each node. Having them as a “note” has two important benefits: First, it ensure that the link and the quoted passage are inseparable, so you never end up with orphan quotes. The second benefit is that notes can easily be hidden and shown, individually or en masse. This helps clear up the clutter and improve the readability of the outline.

The Contents pane in OmniOutliner allows you to focus on one particular section.
The “Contents” pane in OmniOutliner allows you to focus on one particular section.

You could also use OmniOultiner as a means of outlining an entire manuscript, making the literature review section just one portion of the broader outline. You can then use the “Content” pane in OmniOutliner to show only the literature review section as I’ve done in the image to the left.

You might also be thinking that this is a lot of clicking and switching between applications. Indeed it is, but once you really build the keyboard shortcuts into your muscle memory and get the hang of using cmd-tab to switch between applications, you’ll find that you can get a quote and page link into OmniOutliner quickly and with minimal effort.

I hope this post inspires you to think about how you review your research literature and that maybe it will help you with your own workflow. I had been using DEVONthink for a while now, but it was OmniOutliner that really revolutionized how I digest my academic literature. The two work together so well when using “Item Links” and “Page Links” that I hardly know how I go by without them!



For some reason this particular recall is trending on Facebook… this is the first time I’ve ever seen that happen. Interestingly, some news reports are highlighting the fact that it was prompted by a “recall in another (undisclosed) country”. It takes about 5 minutes to find out that the country is the USA, and that this is actually just a second wave of an earlier recall from July 2nd that didn’t extend to Canada. 

What is interesting about recalls like this is that you get to see just how truly limited our consumer choice is. Look at the list of products being recalled, all of which are manufactured in the same facility.

This facility is producing food for low-cost private label brands (e.g., No Name) as well as for food service companies. These products are likely at vastly different price points for the end consumer. While obviously standards set by the customers and other variations in QA can impact quality and price, this certainly does make plain that much of our food comes from a very small number of places, even if the perception of a plethora of choice suggests the opposite. 



CWN logo

I’m pleased to announce the launch of Canada’s Waste Network. This is an initiative that intends to bring together researchers in Canada studying waste. It is a new and growing community and new resources and events will be added regularly. If you are a scholar at a Canadian institution doing research on waste (in Canada or abroad), I’d encourage you to become a member and connect with the community!



Inevitable. 


Food Manufacturing writing about a recent report on plastic production and use:

Packaging is an especially big piece of that pie as consumers continue to replace glass and metal with plastic. The study shows that the average person living in Western Europe or North America consumes 100 kilograms (220.5 pounds) of plastic every year, mostly from packaging.

Food Manufacturing

This is a staggering amount of plastic per person per annum. What is often also neglected is the extent to which consumers (and the municipalities that manage these wastes) are rather beholden to the whim of manufacturers and the market for recyclables in their region. Thus while I laud efforts to expand a lagging recycling infrastructure worldwide, the reality is that the most effective response is simply to produce and consume less plastic, recycled or otherwise.