Searching for a recipe

Doomed to failure

We are looking after a friend’s place in the country and perhaps the hardest work is keeping on top of the vegetable garden. Ignore it for a day, and there are five large cucumbers, mocking you. My cucumber repertoire is somewhat limited; raita, quick pickle, sandwiches and, when I remember, Yan-Kit So’s spicy cucumber. Alas, I don’t have her book with me, so I go online, and am reminded again of the absolute nightmare of recipes on the internet.

There are many different complaints. Perhaps the most obvious is straightforward thievery, people claiming long-established recipes as their own. But there are lots of other peeves too. The ones that make you scroll through endless gushing prose and fabulously staged photographs — oh, and a few adverts — before finally, reluctantly, giving up the ingredients and method. The ones that have gamed their way to the top of the search but that hide themselves with paywalls and popups and other barriers. The ones that promise more than they could possibly deliver. The ones that have hidden your chosen keywords in an otherwise robotic list of 17 things to do with a cucumber. The ones that have not yet woken up to the fact that being online means being aware that your readers may not share your understanding of idiosyncratic measurements and brand name products.

You get the point.

There have been attempts to regularise matters. One I had high hopes for in the beginning was a thing called h-recipe. This is “a simple, open format for publishing recipes on the web” according to its definition in the microformats wiki. I need to backtrack.

Microformats are ways of marking up HTML, the stuff that tells your computer how to display web pages, to give additional meaning to the information. The goal is to give computers a helping hand in deciding what kind of information the stuff represents, which should help them to make better use of the information. So, for example, I can tell any computer reading this article about its author — me — by including that information in a microformat called an h-card. In theory, if I published in lots of different places, and used the same h-card everywhere, a computer program would be able to show you all the things I have written without needing to use one of the big search engines with all that that entails.

h-recipe allows you to specify things like the individual ingredients, the yield, instructions and nutrition. That should make it possible for a search engine to show me recipes with the kinds of detail I am looking for. In the first flush of my IndieWeb enthusiasm, I actually made the effort to add the h-recipe details to my recipe for cornbread. Once was enough. It was dull work, brought me no benefit whatsoever and had no impact that I could discern on how many people found their way to that post.1

Alas, there is a singular lack of what the IndieWeb movement (which champions microformats) chooses to call “tools for consuming h-recipe”. In fact, the only working example is Pinterest, possibly the most thoroughly ensiled site on the entire Internet.

What, then, is the point? I’m honestly not sure. “Because it is possible” might be one answer but given the general emphasis on tools for consumption before experimental microformats, the few cases in which h-recipe is generally useful do not make a very strong case.2 We are left with the sad truth that it is actually extremely difficult to find good, useful recipes online. The recipes are there, no doubt, and with experience one learns which domains are more likely to yield success, but search is seldom your friend.

And that brings us right back to Yan-Kit So’s spicy cucumbers. There is, of course, a recipe online, and it is on a silo site that is a superb resources for cooks. I’m talking about ckbk which is a fully searchable database of “ingredients, dishes, authors and more” that is based on the recipes in published cookbooks. ckbk is a very clever idea, kick-started by asking a bunch of people to recommend their 10 favourite cookbooks; here are mine. One of the neat features is that you can discover who likes which cookbooks. Most of mine are pretty mainstream, although one is unique to me. Wouldn’t it be cool, though, to be able to see clusters, along the lines of if-you-liked-this-you-might-like-that?

Anyway, my problem with ckbk is the membership. To access any recipe, you need to pay €5.49 a month or €43.99 a year (currently available at a 25% discount). That is just too much when in all honestly I am unlikely to want more than one recipe a month. Would I pay on a per recipe basis? You bet, but probably not more than 50 cents. Heck, I’d probably pay €5 up front for a bank of 10 recipes that I could download at some point in the future and top up when I needed more. I realise that of course the site has overheads and is also, I hope, paying publishers and authors more than a pittance for access to their recipes, but it seems to me there ought to be some way of making a scheme like I have suggested work.

As for more useful discovery and presentation of recipes on the internet, I can see no solutions, simple or otherwise. The work involved in an independent search engine would be immense, and it would still be plagued by inconsistent mark-up. My own approach is to curate my own collection of favourites, as grubby printouts, online links and boiled down text notes, any and all of which I would be happy to be share, but not if I have to mark them up as h-recipes.

This post is a submission to the Carnival of IndieWeb, newly launched by Sara Jakša. A resurgence of blog carnivals would be a great thing.

  1. And I can think of no other reasons to do the work.  

  2. I cannot say anything about its value in Pinterest as I have no interest in exploring recipes there.  

Two ways to respond: webmentions and comments


Webmentions allow conversations across the web, based on a web standard. They are a powerful building block for the decentralized social web.

“Ordinary” comments

These are not webmentions, but ordinary old-fashioned comments left by using the form below.

Reactions from around the web