Wikimedia Cookbook

Take your example of heat on a pan. I’ve always wanted a stove where I can control the BTUs with a dial. You could test your own stove for it and calibrate yourself for example.

There is a slightly more scientific blog here, but it won’t satisfy what you’re looking for I don’t think. I’ve never found that depth, but would appreciate it too! https://www.seriouseats.com/

I like this book too: https://en.wikipedia.org/wiki/The_Food_Lab

The most famous example is how most baking recipes call for 2-5 cups of flour, since the humidity in the air, the season the flour was harvested, etc. all affect how much you might need in the end. So you just learn when you've added "enough". That might make the engineer in you upset, but that's the reality of it.

There are no precise steps to "brown an onion", but you will know when you've done it correctly. And if you did it wrong, that just means it takes more practice and more learning about your skills.

Generally, things you cook on the stovetop don't need strict quantitative controls. Exercise your own judgement and experience- you can understand how the onion will taste based on how it looks. Mix it, heat it, season it, eat it.

As for regular recipes, I really don't think this type of precise writing is popular because everybody's got different tastes and palates so it would be tough to say precisely how one should 'brown an onion', although I get where you're coming from. Personally, I've gotten in the habit of finding recipes, trying them out, and saving them with personalized tweaks. So when a recipe calls for a 'dash of salt', I replace it with 'exactly 1 tbsp of salt' or 'no salt necessary'.

I wonder if a similar analog is sheet music. Ten different musicians could follow the sheet and produce 10 different sounding results.

Hate to break it to you, but it's even worse. Further variables include the moisture content which varies per-onion, the type of onion, the size of onion, the thickness of your pan, and the material it's made out of.

The reality is that cooking is an art, not a science. Even if your equipment stays the same, every chicken breast and every tomato you buy is different.

Now it totally drives me nuts when people say "if you follow the recipe you can't go wrong!" There are usually 100 different ways you can go wrong.

But it's just part of cooking. You learn to cook the same way you learned to walk or throw a ball or speak your language: through trial and error and careful observation and practice.

The great resource now is YouTube videos, which have the huge advantage of letting you see exactly what it's supposed to look like when it's done (how brown is brown? how thick is thickened?), and just by seeing the types of pans they use and the sizes of flames, and the amount of bubbles or sizzling, you get actually get a feel for it pretty quick.

Finally, the point of recipe books is to tell you which ingredients, rough quantities, and the steps. Just because recipes require interpretation (similar to a music score) doesn't mean they're useless.

With such a huge margin of error for just one ingredient it immediately makes me wonder if anyone actually cooked this before they wrote an article about it.

That would actually be useful, unlike this random assortment of non-curated recipes.

Have you tried Julia Child’s first books “Mastering the Art of French Cooking”? Her books are very in depth and start with clearly written descriptions of techniques and what to look for. This does mean that recipes can be many times longer than most other books.

Julia Child’s reputation is well earned. She may be one of the first celebrity chefs, but her status came directly from the quality of her research and writing.

Absolutely! If you are looking for this kind of specificity, Cooks Illustrated is your source. The will discuss multiple variables of basic techniques. Like in your example, dice size vs. browning time vs. just what the hell does "browned" mean anyway?

:

:

The other solution to this: practice. If you cook thoughtfully, you will eventually come to understand how these variables impact your dish. But you need to pay attention while you cook, like a good engineer!

I'll give you a 3 examples, specifically about onions, that took me years of practice to understand.

Indian gravy base (essentially all dark curries and masalas) A <1cm dice, cooked very slowly until dark brown with no spot of white and not burned. You literally want a uniform brown paste because that gives the gravy a solid base and has a very earthy taste to it. Hard to do without burning.

Mire Poix (a base used in many, many dishes in Europe: 50%carrot, 50%celery, 100%onion by mass): A <1cm dice, cooked slolwy until translucent. You don't want a toasty flavor at ALL, you want the savory taste of the onion with no zing. (easy, can be done in ~8-10 minutes)

Spanish Caponata: A <1cm dice, cooked until browned edges (higher heat), but still white center, meaning sweet with some zing but still a toasty part. (tricky to not burn and still get a brown edge)

And all of this skill changes if you buy a new pan or move to a new house with a new range! Or if your yellow onions are dried and cured properly, or overly moist or greenish.

But cooks Illustrated will cover a lot of this, you just need to buy a lot of books though because they spread this out across several volumes. :(

Practice though. Read cookbooks. Literally read them, the intros are usually very thoughtful. Marcella Hazan, Madhur Jaffry, Alma Lach, OTTOLENGHI... just to name a few of the greats you should check out.

There's a second edition with some more techniques added.

One book that helped me was Vegetarian Cooking For Everyone. You may not want to cook vegetarian, but many vegetarians are young people who are also building their own cooking styles from scratch. So VCfE goes through the basics - like what is cubing and what is dicing - with illustrations. While it has complex recipes, it also includes almost absurdly simple ones, like preparing beans with olive oil and salt. But think of them like "kata"; basic exercises, that you can build up to more complex forms.

A recent book which I've recently found extremely useful is Salt, Fat, Acid, Heat. It is a wonderfully readable introduction to the elements of cooking and how they combine, and it is great for complete newbies as well as experienced home cooks. It manages to be both accessible, practical and a little bit technical, perfect for the techie who wants to get better at cooking. It taught me things simple dishes like roasted chicken, which I had prepared like a hundred times before I read it, which have changed how I cook forever.

It can easily have the same recipes, and it doesn't need to compete, in the same way Wikipedia doesn't with traditional encyclopedias: when you take away commercial incentive, something can just exist.

I don't want to eat the exact same macaroni and cheese dish served on one side of the country that I can on the other.

Great! Looks like there are a few dozen options, and there are alternative recipes in some of these pages:

https://en.wikibooks.org/wiki/Special:Search?search=macaroni...

I think what I am trying to say is that cooking can never be standardized, nor should it.

Wikis are descriptive, not prescriptive. This book seems to be so, as well, making no attempt at standardization. If you have thousands of Reuben recipes you want to input, why not?

You cannot standardize the world's recipes into a monolithic volume like Wikimedia has done with facts.

Wikipedia doesn't standardize anything. You totally can collect them; there are a bunch of cooking sites and applications on the internet, just as there are a bunch of sites that collect facts.

I mean you could, but the 7,000 variations on every chef's take on a Reuben sandwich would be a chore to sift through.

That's what curation is for, and there are plenty of people who search through Wikipedia, find interesting articles, and share them.

Facts rarely change, tastes often do.

This makes a wiki the best possible place for recipes, then. Not only is there no hard-limit on the amount of content (storage is very nearly free, especially given Wikimedia's fundraising numbers), there's also version control history, and HTML has subheadings for variations on the same recipe that might occur over time.

As for recipes, I think a neutral recipe can be a good starting point for your learning journey. It would probably be better if they had more references to variations though, similar to Wikipedia's further reading.

For example, if the Macaroni and Cheese Cookbook page turned into an index for dozens of M+C recipes, named for the cultural influence or ingredient influence, that could certainly be useful.

With the current state of recipes on the internet (most are extremely low quality click-bait designed around ad word integration for profit, for example, searching "instant pot recipes" returns just loads and loads of terrible monetized results) a source like this doesn't have to try terribly hard to be a high quality resource

Even so it feels like it would be more useful as an encyclopedia of flavor than it would be for help making dinner tonight. Learning about the evolution of a dish would be very interesting, but I wouldn't trust an anonymous wiki editor very much in terms of actual cooking

I think the biggest problem to me is that recipes are very much driven by personality. People learn a few chefs/sources that have produced good results for them, and they stick to those sources. They go buy the cookbooks from those people and stay in a safe walled garden where they can trust each recipe

I think another big problem is the low quality of the recipe source. Traditionally listing ingredients with some steps along the way is sufficient for the experienced cook, but good modern recipes (to me) do so much more than list ingredients and steps.

I follow a chef J. Kenji Lopez-Alt and when he breaks down a recipe (for Serious Eats or his books like The Food Lab) he goes so much more into WHY each decision was made, which empowers you to not just understand the decision but also change it.

For example: The All-American Beef Stew -(https://www.seriouseats.com/2016/01/food-lab-follow-the-rule...) (and the short version https://www.seriouseats.com/recipes/2016/01/all-american-bee...) Every aspect of the dish is analyzed and discussed!

Or Fuschia Dunlop's translation of Sichuan cooking for a western audience, (e.g. https://andrewzimmern.com/2013/03/28/fuchsia-dunlops-fish-fr...) she helps distill Chinese culture and the cooking of this region with a lot more than an ingredient list and a step. That information, delivered by a trusted expert, is vital to understanding and mastering a dish from a culture very different than my own.

Why would I want some anonymously written/edited list of ingredients and steps when I can have a professionally produced, multi-media (huge useful step by step images AND video), "scientifically-inspired" discussion of the recipe and why it does what it does?

That's the kicker for me, I'll always go back to Kenji or Alton or Fuschia or someone I trust because I know they have delivered good results and researched the recipe deeply enough to help me succeed not just at their version but at my own version too

E.g. https://en.wikipedia.org/wiki/Martini_(cocktail)

Simple example, ever seen a Chicago style pizza? It would be impossible to create a standard “pizza” recipe that covers all pizzas ... but you can have a specific style with a name. Culturally, there are places in the world (for example France) that do treat food recipes this way. The US is too individualistic for this to be found “acceptable” culturally.

Not that this is what is being attempted here, but rather than one or more recipes that each represents a particular set of curated choices, you could instead create a "recipe as multidimensional model" that captures the entirety of the configuration-space for "the food that is called X", where you can both adjust inputs and see how output parameters (e.g. saltiness, acidity, viscosity) will change, or constrain output parameters to find optimal input parameters.

In other words, something like https://hur.st/bloomfilter/, but for tuning your chocolate-chip cookies.

If you look up a particular recipe online, you end up finding those 7,000 variations scattered across numerous blogs and websites. I don't see any requirement on the number of variations of a recipe that can be added to the Wikimedia Cookbook, and it's no worse to have some of them hosted here.

Not sure the mediawiki format is great for this however.

I have no idea if this is in use anywhere or if any actual recipe service exposes this information, but there it is.

https://www.vintag.es/2018/11/honeywell-kitchen-computer.htm...

"It was advertised as a machine for storing recipes and helping housewives in their daily domestic tasks. However, reading and introducing a recipe was a difficult if not impossible task as the computer had no display and no keyboard. It required a two-week course in order to learn how to use the machine."

Also we can be pretty critical here on HN but I see no reason why Wikibooks must be judged without reference to its potential--not only is it already helpful to some, for various reasons all along the long tail, but perhaps it's just a few tweaks away from even greater functionality. Critiques too often impart the idea that the current experience is lacking _and therefore_ needs complete replacement by e.g. "a website that..." when the existing efforts at least show potential in terms of humans being willing to work together productively toward building a generous resource. That's really something.

Good point. In fact, with the risk of vandalism, this could be a recipe for disaster. On Wikipedia, vandalism may lead me to believe that the president of Russia is Taylor Swift, something usually harmless. Here, someone may vandalize a recipe to add 1 cup of arsenic. Okay, more seriously, I wonder if there are subtle ways to mix normal ingredients into poison. I suppose this is already a risk on WikiHow, where someone could tell you to clean your house by mixing bleach and ammonia.

What about a web of trust for some wikis? Instead of letting everyone edit a page, begin with the wiki's founder, who can delegate editor rights to people he knows, then they can grant the right to their friends, and so on. I have always loved the wiki format and prefer them at work over more draconian tools with complicated workflows. So I am surprised to hear myself saying this. But recently at work I have also been dealing with server certificates. So a web of trust is on my mind.