#microformats 2014-09-29
2014-09-29 UTC
KevinMarks_, tantek and Skud joined the channel
# 
Skud hi folks. can anyone tell me if there's a microformat or anything like it for expressing semantic equivalence of things? For instance, if I have a page on my site all about tomatoes, I want to link to some other sites saying "these are their pages about tomatoes"
# Skud like, "this remote page deals with the same subject matter as this one"
# Skud i'm looking at the wiki now
# Skud so it's be like <a href='http://en.wikipedia.org/wiki/Tomato' class='u-something'>
# Skud ummm wait. back up. we don't have a specific microformat for what we're doing.
# Skud it's a crop database
# Skud so i guess i was thinking more like a rel="something"
# Skud i'm trying to advocate for all the various open food projects to make an effort to link to some of the larger central data repositories, to assist with interoperability
# Skud we're not tagging something.
tantek joined the channel
# Skud tantek: were not tagging anything, and the h-entry docs say they're for blog posts etc, which our pages aren't.
# Skud tantek: here is the page in question http://growstuff.org/crops/tomato
# Skud the sidebar link under "More information", is what we want to mark up, to say "this wikipedia page addresses the same topic as our page"
# Skud yes but i don't want to say "this is a page about tomatoes"
# Skud i want to say "this is a page about the same thing as this page"
# Skud does that distinction make sense?
# Skud so it'll say <a href="http://en.wikipedia.org/wiki/tomato" rel="tag"> and that will suggest to the world at large that the tomato page on wikipedia and the tomato page on growstuff cover the same topic?
# 
tantek what do you think KevinMarks? shall we update the examples in http://microformats.org/wiki/rel-tag - especially since Technorati no longer supports arbitrary blog/page tagging like that?
# Skud well yes, sorry.
# Skud i must say this seems *extremely* odd to me.
# Skud well, having been using tags and folksonomies for about a decade, i thought i knew what they were
# Skud a small bit of text that describes the subject matter of some content
# Skud the examples on the rel-tag wiki page make perfect sense to me
# Skud but expanding it to work like you're suggesting no longer seems like a tag to me
# Skud wikipedia's tomato page is not a tag
# Skud it's a webpage full of content
# Skud it just seems a bit mindbending, in a semantic sense. the meaning of "tag" is being stretched to the point where i no longer recognise it.
# Skud ok, but here's another thing
# Skud some of the things we link to might not have that format of URL
# Skud for instance http://eol.org/pages/392557/overview
# Skud that's the encyclopedia of life's page for tomato
KartikPrabhu joined the channel
# Skud but the last segment of the URL is "overview" and before that "392557"
# Skud the other thing is, you can use tags for things that aren't identity equivalence. for instance i could tag something "funny" or "read later"
# Skud so when i use rel=tag it doesn't really imply that thing A is the same as thing B, just that thing B in some way describes thing A
# Skud now, we might want to *actually* have tags on our tomato page, tags like: fruit, full-sun, from-seedling, temperate-climate, etc
# Skud linking to rel=tag for those makes perfect sense to me
# Skud but i want something different for our links to external sites with content about the same entity
# Skud more like rel=alternate
# Skud but in a, like, semantic identity sense
# tantek here's a workaround for example: http://eol.org/pages/392557/overview?/tomato
# Skud here's a post I just wrote about what I'm getting at in a broader sense: http://talk.growstuff.org/t/open-food-interoperability-entities-unique-ids-and-semantic-equivalence/93
# Skud yes. we worked together.
# Skud tantek: we've met a couple of times i think, first at barcamp in 2007 IIRC
# Skud yeah
# Skud btw my post is largely in response to https://www.loomio.org/d/lKdgHqoJ/open-data-standards and a bunch of other projects in the space sort of failing around the topic
# Skud ha ha yeah, that's an issue
# Skud pragmatically speaking, for instance, a whole buttload of vegetables all have the same binomial/scientific names
# Skud to a scientist, it doesn't matter that 20+ types of leafy green are all "brassica oleracea var. acephela"
# Skud but to a veggie gardener, they really care about dinosaur kale vs russian red kale, say
# Skud yes
# Skud along with collard greens and heaps of others
Exploter joined the channel
# Skud check https://docs.google.com/spreadsheet/ccc?key=0AmnATvIrO73HdHRiMXBFbEk3dUJNTmFNazZZX0U0Ync#gid=3 (the brassicas tab)
# Skud all tomatoes are one species
# Skud all chilli peppers are 3 species
# Skud all squash (from cucumber to giant pumpkins to loofahs) are just a few species
# Skud they're just like "oh yeah there's some natural variation but they can interpollinate so they're the same species"
# Skud like dogs
# Skud they're all canis whatever, biologist don't distinguish between a labrador and a poodle
# Skud so basically, as gardeners, we are a bit "ehhh whatever" about what the scientists say
# Skud like it's sometimes handy to know the scientific name just to check whetehr something is what you think it is, but it's not a primary identifier
# Skud yes but still the same problem
# Skud as i said above, 20+ types of leafy green are all "brassica oleracea var. acephela"
# Skud sometimes it's var. or subsp. or group, whatever
# Skud yeah that's fine :)
# Skud the problem we have is that most of the biological databases only operate at the species level
# Skud but over on growstuff people are like "can you please add purple onions? can you add rainbow chard? can you add such-and-such a bean?"
# Skud because that's how gardeners work
# Skud openfarm are a new project just starting up and they're keenly loading thousands of crops from a big biological database
# Skud and they are very keen on open data standards and interoperability but have never done any work in that area before so they are flying kind of blind
# Skud oh if only they were subject matter experts ;)
# Skud i mean i'm not really either
# Skud yeah, most of us are enthusiasts i'd say
# Skud but one of the things i learned at freebase etc is that it's better to model your ontology on what normal people use and talk about, than to go for some abstract and complex model that nobody really understands
# Skud yeah
# Skud the other thing is that i see groups like openfarm deciding to use eg. USDA planting zones, which are US-only *and* are changing rapidly at prsent (50% of the US has shifted growing zone since the 1990s)
# Skud ehhh. they'll either figure out something different, or they'll be effectively US-only :-/
# Skud tantek: that's what growstuff is doing. we crowdsource data on what people are planting/harvesting and their geolocation, and aggregate that data and infer planting advice from it.
# Skud openfarm are hoping to get enthusiasts to write "growing guides" for their locations and type of garden (eg. hydroponic, permaculture, container gardening) which is a different approach but not inherently a bad one.
# Skud tantek: neither, it's actually largely based on the huge knitting/crochet site Ravelry (http://ravelry.com)
# Skud you can't edit other people's info. you just record your own stuff (like a garden journal) and then that information is aggregated.
# Skud http://growstuff.org/crops/tomato.json may be instructive
# Skud how do you mean?
# Skud heh.
# Skud they're not recording information about "tomatoes", they're recording information about "the tomatoes i planted in my backyard on saturday, from seed that i got from such-and-such a shop"
# Skud our "crop" model really has nothing but a name to it.
# Skud all the actual data is on "plantings" and "harvests" and suchlike
# Skud that's a work in progress: http://talk.growstuff.org/t/tracking-the-source-of-seeds/92
# Skud not really. most people just buy their seed from a catalogue.
# Skud and don't conceive of it in that way. trying to get gardeners to think in github terms would be pretty impossible i'm afraid.
# Skud what you're suggesting sounds a little more like what the open source seed initiative are doing http://www.opensourceseedinitiative.org/
# tantek digging around I found this http://www.wikiplantatlas.org/plants/earth/index.php (sorry, pardon the pun)
# Skud there are WAY too many garden puns in this project
# Skud hmmmm i'm not so sure. there are some projects like that but they mostly deal with native species etc.
# Skud there was one like that developed at australia's govhack event recently
# Skud native species i mean
# Skud yup. plus in australia almost all our food species are non-native.
# Skud yeah :)
# Skud yes. send them our way :)
# Skud we had a hack night in SF back in June but it was quite small.
# Skud yup. don't i know it.
# Skud and i was only passing through for a little while so it was hard to get the word out.
# Skud no, not personally. i've looked into it a bit.
# Skud if you mean https://github.com/fictivekin/openrecipes it seems to be fairly moribund
# Skud oh yeah taht was the other place we crossed paths IIRC
# Skud i wanted to stay for indiewebcamp but couldn't.
# Skud i think we were in some OSB session together or something. i probably didn't say hi though :-/
# Skud anyway, back to the open food interoperability thing... a vast number of us are doing Rails projects with JSON REST APIs
# Skud but we're almost all taking different approaches to 1) what crops/varieties/species we cover, and 2) what data we collect and store
# tantek filed an issue which I thought would be pretty easy to fix: https://github.com/fictivekin/openrecipes/issues/198
# Skud for instance, wrt what crops we cover, Growstuff's scope is "edible plants", while PracticalPlants is "useful plants" and openfarm is "all plants"
# Skud http://openfoodfoundation.org/document/open-source-food-projects-list-wiki/target-audience-functionality-web-based-openfood is a good place to start
# Skud though they are starting at the "farmers trying to distribute their produce" end of things and haven't ventured as far as eg. openrecipes wrt ingredients and how you use them
# Skud oooh would you mind dropping a link to that on Growstuff Talk?
# Skud brb just microwaving some lunch
# tantek Skud, can you start by editing http://microformats.org/wiki/plant#Real-World_Examples and adding each of those resources/groups as subsections?
# Skud which resources/groups do you mean?
# Skud oh you mean like growstuff etc?
# Skud i'll add growstuff and a link to some of the lists of other gardening projects... i don't have time to add them all
# Skud man, captchas are getting creepier now it's more and more obvious you're transcribing people's house numbers :-/
# Skud yeah we use it too. i just used to feel so much better about it when it was transcribing old books and things.
# Skud here you go: http://microformats.org/wiki/plant-examples#Growstuff
# Skud i mentioned a couple of problems i foresee
# Skud including the species issue
# Skud i don't think any of the ones you listed are open source/open data are they?
# Skud oh yeah they don't need to be
# Skud i just mean, i have mostly been looking at the other open projects
# Skud and of the ones you list, none are Australian, so i hadn't actually heard of any of them (well except for BBC gardening)
# Skud ok, time for me to head out and get some stuff done.
# Skud good talking to you tantek
# Skud i'll lurk in here for a bit. ping me if anything relevant is being discussed.
# Skud or dive in on that thread on Growstuff Talk
# Skud yeah, i'd love to hear his thoughts.
# Skud yeah.
# Skud maybe i can catch up with him when i'm in the UK soon.
Aranjedeath, KevinMarks_ and Exploter joined the channel
# 
KevinMarks Fascinating conversation on plants
# KevinMarks It was only when gardening myself that I realised that food plants are all clones
# KevinMarks Or almost all - when you buy fruit trees they are rootstock with a graft from an existing tree
# KevinMarks So all Meyer lemons are clones from one source tree
# KevinMarks They may mutate a bit but they don't breed
tantek, Exploter, caseorganic, ChiefRA, kez and pfefferle joined the channel
# Skud tommorris: i'm already pretty aware of wikidata (i was involved in its early days) and am going to be meeting with some of those folks in berlin soon (i'm visiting london and berlin for a couple of weeks in Oct)
# Skud i'm hoping that lydia pintscher can connect me with any wikidata-ians (what is the word for that?) who are particularly into this sort of stuff.
# Skud KevinMarks: fruit trees can and do breed from seed, but the desirable fruit characteristics (sweetness, etc) are usually recessive genes, so seed-bred fruit tend to be not as good as grafted/cloned varieties. however, chance-grown fruit from seeds is how we got some of our common varieties, eg. granny smith, which just kind of randomly happened.
KevinMarks__ joined the channel
# KevinMarks__ Skud - I loved this article you posted before: http://www.motherjones.com/environment/2013/04/heritage-apples-john-bunker-maine?page=1
# Skud tommorris: thanks :) any interest in a pint or something while i'm in town btw?
# Skud there's also this event: http://talk.growstuff.org/t/london-working-bee-code-sprint-october-18-19-2014/79/6
# Skud KevinMarks__: that dude sounds so great. i would love to meet him. there's a guy in my town who's "the apple guy" but he's not as much as an expert as that dude. or at least, he couldn't tell me what the apple in my backyard was.
# KevinMarks__ i wonder at what point gene sequencing becomes common enough that it's done routinely
# Skud hmmm that would be cool
# Skud tommorris: wait i think i just remembered you don't drink. coffee? bubble tea? vegetarian lunch?
alanpearce and krendil joined the channel
# Skud tommorris: i'm not sure i have your email address handy but if you drop me a line at skud@growstuff.org we can probably sort something out
jschweinsberg, robmorrissey and eschnou joined the channel
# 
@HostVergelijken Ervaring #flexwebhosting: "...doe het niet!" http://www.hostingvergelijken.eu/flexwebhosting-hosting-recensie/?wpcrp=1#hreview-115 #hostingvergelijken (twitter.com/_/status/516576347477803008)
globbot, TallTed, eschnou, danielfilho, robmorrissey, alanpearce_ and gRegor` joined the channel
# 
@seitics Habe jetzt so eine O2 Multipa^H^Hcard http://www.youtube.com/ Einziger Anbieter mit DE-weiter MSC, d.h. nicht per SS7 geolokalisierbar? (twitter.com/_/status/516620185466580993)
# 
@astockma RT @seitics: Habe jetzt so eine O2 Multipa^H^Hcard http://www.youtube.com/ Einziger Anbieter mit DE-weiter MSC, d.h. nicht per SS7 geolokali… (twitter.com/_/status/516621349096542208)
tantek, alanpearce, gordonb, caseorganic, Exploter, netweb, encolpe, eschnou, KartikPrabhu, KevinMarks_, Aranjedeath, KevinMarks and krendil joined the channel
# gRegor` In an h-card, should a department go in the p-extended-address? I don't see a property for 'department' on the wiki.
# gRegor` It's a member-only directory of college graduate programs.
# gRegor` So it's behind a login.
# gRegor` But basically listing university name, department, address
# gRegor` University: Oregon State University
# gRegor` Department: School of Biological and Population Health Sciences
# gRegor` The rarely-used p-organization-unit?
caseorganic, Exploter, KevinMarks_ and KartikPrabhu joined the channel