Cookiwiki:Community Portal
From Cookiwiki
Contents |
General organization: by country, or using categories only?
The issue is: should we retain the organization by country (hardcoded, not through categories) in the Cookiwiki?
Pros
- It would be nice to have them as a place where in time one can add histrorical and ethnic information about cooking in various countries.
- It allows organizing the recipes of a country in a way that makes sense for that country (for instance, in the order in which the various categories of food are eaten in that country during a meal).
- In some time, many countries will be created, so it won't be very much of a burden to users to add recipes there.
- The content of this site excludes copyrighted materials, so it's fair to assume that many of the recipes here will have some kind of an ethnic flair to them, or at the very least have been handed down from one generation to the next. As such, it makes sense to organize things based on how the recipes were transmitted to people.
Cons
- If recipes are added with the "Add a recipe" button, then they won't automatically be added to the right country page.
What is your opinion?
For any wiki to be useful, each page needs to point to another page, and be pointed to by some page. Otherwise each recipe will be a dead end, only arrived at by a lucky search. I wouldn't expect every user to link each page to a higher order one, but community regulars could certainly fill that function. As users become better equipped with the site, they too will take on those roles. Also, its important that we actually put content on the Ethnicity pages such as trends in a countries cuisine, links to region specific cooking techniques & products, etc. otherwise they'd better be served by a search function. Next, it might be useful to outline how we differ from the WikiBook Cookbook. Finally, always keep in mind that we can change things at any time, especially as a larger and more diverse community develops. kneague
How do you think we should deal with English to Metric conversions?
What about putting up a conversion table, says LucaDeAlfaro? Do we need more?
I'm a bit nervous about people converting incorrectly. Its too bad there's no macro to automatically converr to/from metric in the toolbat. --kneague 23:34, 4 December 2005 (PST)
WikiBook Cookbook
How do we compare to the WikiBook Cookbook?
LucaDeAlfaro says: I am not sure. When I started this wiki, I could not find the Wiki cookbook; I found it only after the fact. Then, I started browsing it. To my eye, much of it has been written in a somewhat superficial way. For instance, there is no note in Fettuccine Alfredo saying that this is not an Italian recipe (try ordering it in Italy!), but rather, is a foreign interpretation of Italian cuisine (not to imply lower quality: Chicago-style pizza can be excellent). Or look at mozzarella; "it has become practical to substitute it to parmesan" - practical in what sense, that it can be done? Same can be done with gruyere, but there must be a difference between what one wants to do, and what one can do!
I would like a wiki that contains also information about the history of recipes, the various types of ingredients, the culinary tradition of the country where the recipe originated, particular tricks for preparing it known only to people who really do prepare those recipes. A wiki that is not necessarily encyclopedic in coverage, but where people write about the food they know well.
However, I guess the real question is: why starting a new wiki, rather than going to the Cookbook and start editing it?
- Yeah, I don't know-- the Wikibooks are more fact based than opinion and experimental based. I'm not sure if they would cultivate the kind of a non-authoratative community you (and I) seem to want. I'll look through their site over the next few days. However, they certainly have a strong infrastructure already. We should think about some permament limitations of the Wikibook and see if they outweigh devoting our resources here.
- On second thought, (1) it certainly would be a lot less work to just edit the Cookbook. I mean, the infrastructure and basic knowledge is there already, but it seems to be missing that non-sterile touch you desire, (2) its a stable site that doesn't require your time to maintain, (3) it might be more constructive to add to the Cookbook than make a competing site, especially since any visions on this site can be implemented on the Cookbook.
- Yes, we could definitely give it a try. We could try to add recipes there and add the information we wish to have there. If we get edited out too often because the style of things we want to do is not accepted there (it would be unlikely, I guess), we could come back to this wiki. In the next couple of days I will try to add some recipes in the Italian section, just to see what happens. LucaDeAlfaro
- Agreed, I'll do the same. --kneague 22:47, 7 December 2005 (PST)
re:wikibooks and wiki cookbooks
If you want a truly community-driven site, I respectfully suggest you *not* use a wiki. We have discussed at length the problems with using wikis on the wikibooks Talk:Cookbook page and its archives. There are many features lacking from mediawiki that inhibit certain possibilities with a Cookbook (e.g. category intersection, listing several categories to a page), and many software extentions that would be necessary to make a useful community site are Cookbook-specific. On wikibooks, there seems to be little chance of these ever being deployed, and since most mediawiki development is for wikipedia, the changes that we most need are not the top priority. Furthermore, some of these software extentions would be difficult to apply to a wiki environment (e.g. rating systems, automatic calorie calculation, measurement conversion, or automatically linking recipes with a certain ingredient to a category for that ingredient). This leads me to believe that a wiki cookbook in general is a dead-ended project, even if it's fun. If you have software engineering skills and would want to build a recipe site, recipezaar is a good one to look at for ideas. I might also be interested in working on that project, though I'm not quite sure what the difference would be between it and other recipe sites... wb:User:Kellen Talk 85.178.242.176 13:54, 11 December 2005 (PST)
having been combing the archives over there for a few days, i now agree with kellen's sentiments. in addition, i initially thought it a good idea to allow a recipe to evolve. by that i mean cooking a recipe from the wiki, and then altering its page to improve it based on your experience. the problem with such a procedure is that different users will have different skills, different cooking equipment, and markedly different tastes. i could imagine the pizza sauce page being a constant battle between people who prefer an authentic sauce to the sweetened crap I'm forced to eat in California. who wins in that case if both users legitimately feel their version is superior? and if the recipes are split into two different pages, as a novice user, which recipe do i trust as being better suited to my tastes? in that sense, the non-authorative nature of wikis might actually be a bad thing. in any case, there are an abundance of great ideas for cooking sites out there, each with their own pros and cons, and i do think a site can be developed to combine those ideas. ideally, it would be a combination of ease of use, community input, and a bunch of the useful scripting functions mentioned by Kellen. i think it would be best to have those scripts in place first prior to making a site go live. otherwise, there would be an endless amount of restructuring each time those features are added by a crafty and dedicated programmer. kneague 16:53, 11 December 2005 (PST)
LucaDeAlfaro (this wiki maintainer) says:
I am interested: I would like to know more about which extensions are needed. I am a good programmer, even if I have limited time, and if some of these extensions are not too difficult, I could very well add them.
What I cannot promise is that I would add them in a Mediawiki wiki. The reason is that Mediawiki is written mostly in php, and uses an sql backend; I know neither php nor sql, and I am not dying to learn them (well, sql if necessary). Also, the code is not very modular. However, at work I am maintaining some Wikis by MoinMoin. MoinMoin is written in Python, which is a very nice programming language, very readable, and it has a very simple way to add macros. In the two weeks I have been playing with Moinmoin, I have already been able to add useful capabilities we needed at work.
In Moinmoin, I think it would be quite easy to save, as user preferences for the users who login (anybody can login, just as here), a metric/imperial flag. And it could be quite easy to enter measurements as [[Quant(250, g)]] that would then be displayed in the units of the reader convenience. In fact, I could even provide somewhere a button or something that toggled the setting. Of course, using [[Quant(250, g)]] calls for more sophistication on the part of the person editing the recipe.
Let me comment on other aspects:
- category intersection: categories are very easily done in Moinmoin, and I think this is feasible.
- listing several categories to a page: if I understand correctly, this is already there in Moinmoin.
- automatic calorie calculation: this I would give up. I don't think there is any way you can do this. The results would anyway be bogus. The ingredients are not standard, nor the amounts of lots of things like oil, butter, etc, which contain calories as well. For instance, when Italian pasta sauces call for "pancetta", living in California I buy bacon, then slice by slice I cut out most fat, to achieve something more similar to the initial meat/fat balance. How do you account for that (which is not even said in the ingredients)? My idea anyway would be to cater to interesting cuisine, rather than dietary needs; everyone will have to figure out independently the dietary implications of recipes.
- rating systems. As it is easy to write Moinmon macros, one could perhaps write some sort of pagerank-like algorithms, that automatically compute a "reliability" factor for each contributor. Then, based on the reliability factor of the people who have already edited a recipe, people with low reliability factor could be prevented from editing it further. This is not unlike slashdot, where only people with some "trust factor" can moderate. We could also have a system of invitations, so that we could invite people with high reliability factor (suppose we invite people at a renouned cooking school to contribute).
So, I would be very interested in hearing the concrete suggestions you have, regarding making a cooking wiki site truly working. It might not be altogether unfeasible. In particular, I would be interested in looking into what makes a contributor "authoritative", in the context of a wiki, and on how to gracefully gain this information from wiki content and history, or from some other input (e.g., in slashdot style, asking people who wish a higher authority to rank other contributors, then try to measure how good they are at that).
Maybe also glance at MoinMoin, and let me know if you think it could work to use that as the basis; I'd much rather be modifying Python code than php code.
- Sorry I guess I wasn't clear enough. I don't think a wiki is the right software for a recipe site at all. Bolting on all that other stuff is very hacky, and doesn't solve some of the fundamental problems (consistent layout, encoding of all the measurements, etc). More appropriate (unfortunately) is to write a recipe site from scratch. IMO, most of the wiki stuff actually just gets in the way of the recipe (e.g. allowing users to add pictures whereever generally just fucks up a consistent layout). The data for recipes is highly structured when compared to the rest of the junk that goes into a wiki, and all of the fun features of recipe sites come from manipulating the data with knowledge of the structure (recipe scaling, unit conversion, etc). Putting that data into a wiki makes it more difficult to get at the underlying structure and is in many ways just a waste of effort. wb:User:Kellen Talk 85.178.254.174 01:19, 12 December 2005 (PST)
I think it really important to survey the most popular recipe sites out there and to list their positives and negatives. By doing so, you can establish what your goals are for this site, and to specifically identify how this one differs. Moreover, the goals should also address why someone with a library of cookbooks would consult this site as opposed to their authoritative references.
Here's some things I can think of that I personally would want out of a cooking site. Carefully consider the use of these ideas-- as a whole they sound kind of like Homer Simpson's car, which had all these amazing features, but what so overdone no one wanted to touch it! Maybe we should move this conversation over the to the Wikicookbook, asking folks about the limitations they see on their site and what a different site could offer.
- The ability to list a few ingredients that I have in my refrigerator, and turn up recipes that build around them.
- A feature allowing me to identify recipes using items already in my cubbard-- the No Cooking tag does this nicely already.
- Cross-categorization that is fueled by the user individually adding tags to the recipe. e.g., Category:Breakfast, Category:Recipe, Category:German would make this recipe show up on the following pages: Breakfasts Recipes, German Recipes, German Breakfast Recipes.
- An optional, user-specific preference that hides all recipes containing a certain list of ingredients or categories of ingredients. This list could be set up in a user's preference file and would apply to every search performed by the user. Thus, people who hate asparagus need not worry about ever seeing the word asparagus again.
- An automatic labeling for Kosher, Vegetarian, Vegan, Gluten-free, etc. recipes. This can be achieved by having a new-ingredient template with tags for these kind of categories. So, butter would be listed as Vegetarian, Kosher, and Gluten-Free, unbleached flour would be listed as Vegetarian, Vegan, and Kosher, beef would be listed as Kosher and Gluten-free, and Pork would be listed as Gluten-free. A recipe for beef-shnitzel with ingredients flour and beef would probe its ingredient pages, and automatically deduce that this meal is only okay for people following a Kosher lifestyle. A recipe for pork-shnitzel, with pork and flour would show up as something no one with a dietary restriction could eat. Why do it this way? In the long run it requires less of the user, it doesn't require the user to know what constitutes something like vegan, it opens up more dietary possibilities to people with restricted diets ("oh wow, i never have thought of X as vegan, but i suppose it is!")
- A combination of the last two would instantly turn the cooking site into an excellent resource for people with special dietary needs or preferences.
- A recent changes module that encompasses the whole site and nothing but the site. This is something thats probably never going to be implemented at the Cookbook, but is already in place at the Cookiwiki.
- The English/metric conversion scheme you mentioned sounds great.
- The nutritional information thing scares me-- I think most cooking hobbyists are more concerned about taste and freshness than nutrition per se. Were one to program it ... one possible implementation it to have a standard ingredient tagging style. For example, [[Ing(Onions), Quant(2, medium)]]. Useful yes, user friendly-- not really for the average user. And I don't know how you'd deal with things such as "1/2 cup onions chopped" vs. "1/2 cup chopped onions" or the bacon example you gave above.
- A ratings system based on collective input, but that is guided by voted rather than a wiki-like editing scheme. A Netflix style one that matches your own preferences to others' votes would be even better. That said, such a scheme would cripple the site because these procedures are computationally heavy.
- A scientific and historical analysis of each recipe would be great, but short of plagiarizing the New Best Recipe or Harold McGee's On Food and Cooking, I'm not sure how this could feasibly be done.
kneague 12:14, 13 December 2005 (PST)
- The nutritional summary should be approximate based on usual use of items. it's pretty easy to get measurements for *most* items already, and there are recipe sites which already calculate these values. I would also add to the list:
- mutliple-category based filtering (think of it as viewing the set intersection of certain tags) as on recipezaar
- 85.178.201.35 15:26, 13 December 2005 (PST)
Current Status
I am thinking and working at making some of the above suggestions possible. I also have some other ideas, that I hope will contribute to make this an interesting site for sharing recipes, while avoiding too much "fight" about which version of a recipe is the correct one.
I need some time to implement what I want to do, but I hope to have something in a couple of months. LucaDeAlfaro, 23:39, 19 January 2006 (PST)
