#tantekedited /hcard (+0) "re-order n subproperties in common properties summary by order commonly published" (view diff)
Soopaman and manu1 joined the channel
#tantekedited /hcard (+2328) "/* Properties */ longer example hCard demonstrating list of common properties" (view diff)
#tantekedited /hcard (+121) "fix some alignment, simpler intro, shorter descriptions, nowrap so they align even when narrower, float both to avoid text overlapping" (view diff)
#tantekplease take a look and see if that helps with the introduction of common properties (which I think is the 2nd most common thing people will look for after the very simple examples at the top) http://microformats.org/wiki/hcard - cc: Phae aaronpk hober mkowens singpolyma
#tantekedited /hcard (-1) "/* Properties */ left left floats for more predictable layout" (view diff)
#tantekok with that minor styling tweak the new common "properties" section and line-by-line example should be good to go
#tantekgood morning Phae, would appreciate your review/thoughts of the usability/readability of the new section I added to the hCard spec, in terms of how it reads from the beginning: microformats.org/wiki/hcard
#tantekmorning neuro - take a look at microformats.org/wiki/hcard and let me know what you think of the new list of common properties and example next to it
#barnabywaltersIMO that is a significant improvement — much clearer and more useful
#neuro`I like it. Would have probably put the code part on the left though
#neuro`Plus no vertical separator between the code and its explanation makes it a little harder to read.
#tantekneuro - I've heard the opposite (about left vs right) - I think having the more english readable list on the left is a nicer reference for most folks
#tantekand the code on the right provides a good learning reminder for newer folks
#tantekwhen you say no vertical separator - are you saying the two columns are next to each other?
#barnabywaltersanything that makes microformats easier to implement and understand is good in my book
#neuro`tantek: just did some tests, and a visual separator doesn't add any readability, so forget about it. And yes, for normal people, English first, code second makes sense.
adactio and barnabywalters joined the channel
#Phaehey tantek. i'll get to it. first day back after time away, so i'm a bit swamped this morn. should have some time this evening.
elf-pavlik, globbot and comkee joined the channel
#ChiefRAtantek: a) I would somehow combine the "Property list" section with "Properties" one-on-one (making only one section) b) I would of placed "Properties" 1st, then "Property list" 2nd and then "Property notes" sections first and then the "Example" one for usability.
#tantekThe vast majority of authors using hCard don't care about *all* the properties
#tantekIn addition, I've had the suggestion from folks that they want/need to see very simple a markup example of each property as they're reading the simple list, and as such the side-by-side layout is the most compact I could come up with (being able to see as much of that on the screen at once also improves usability)
#tantekFinally, the decision to do the markup example of each property next to the list of common properties led to realizing it should all make sense as part of a single composite example which you could copy/paste (which people *will* do because that's what they do with example) and thus it made sense to reduce the common properties set down to what people actually *do* commonly publish.
#tantekThus a) we still need a "property list" section with the full list of properties *somewhere else* (that I'm still figuring out how to best design), and b) keeping the example of the common properties adjacent to the list of common properties themselves is *better* for usability, than separating them into different sections where people would have to scroll back/forth to look at a property definition vs. use in markup.
#tantekAlternatively, if you think you've figured out a better content design for that section, try a sample edit of it on your user page, and show some folks in comparison to what's on hCard right now and see what feedback you get.
#tantekEvery one of these recent content design iterations on the hCard spec has been based directly on feedback from multiple individuals.
#tantekPhae, welcome back, and whenever you can get to it would be great. Hopefully it won't take more than 5-10 min to review and get a feeling for.
#tantekhmm, I wonder if Loqi has a pattern match for verb all the nouns and then turns it into an image.
#mkowenstantek: I feel like there's probably two sets of properties, sort of like "Core" and "Extended" properties. I don't exactly know how to delineate them, but the current list seems fine as of right now, in my opinion (at least with the limitations of the wiki format).
#tantekmkowens - yeah, in theory there shouldn't need to be such a delineation, due to the process and 80/20 rule, but in practice with hCard and hCalendar I took all of vCard and iCalendar and that was probably a mistake.
#tantekthat being said, even some of the presumably 80/20 properties from research end up not being used as often as the data would have led us to expect
#tantekbtw I despise the term "core" because it's almost always just a replacement for a "pure judgment call"
#mkowensActually, I don't think that was a mistake.
#tantek"common" to me implies at least there's some data/experience with actual usage in the wild
#tantekmkowens, fair enough. including all of vCard and iCalendar EVENT object was perhaps not so much a mistake as a "best first guess" that I should have iterated further on and dropped various properties accordingly.
#mkowensAnd yes, "Core" is sometimes a "pure judgement call" but "Common" and "Uncommon" doesn't feel right for defining spec. :-\
#mkowensYeah. I'm not a fan. And you're probably right regarding the use of "Common" versus "Core." I just feel like the connotation of "Core" and "Common" are both different. "Common" feels—to me, at least—not definitive, but rather entirely observational.
#tantekobservational tends to be better than a priori
#tantekI've yet to see someone good at defining "core"
patcito joined the channel
#mkowenstantek: what about the DOM Core? Doesn't that represent a pretty good definition of "core"?
#tantekso yes, any time you see someone refer to "core", in a spec, ask them, what makes it "core"? what are the practical distinctions between "core" and other parts?
danbri and Soopaman joined the channel
#tantekedited /hcard (-44) "/* Properties */ tighten up the columns so they're closer together" (view diff)
#tantekbarnabywalters - good point. he could use hCard and rel=me both on his "vcard style site" as well as on his blog home page where he links to several other profiles: http://visualidiot.com/
#barnabywaltersARGH!! TOO MANY PROFILE PAGES! How does he justify the expenses for all the random domains he manages?!
#Loqi[http://twitter.com/idiot] @BarnabyWalters Trouble is, though, I try to give as little info away as possible, which means it probably fails hCard validation.
#barnabywaltersI hadn't heard of them before he mentioned them
#singpolymahCard validation :P part of the problem with microformats listing "required attributes" is people get the dumb idea that if they don't have those attributes, they shouldn't use the microformat
#tanteksingpolyma - agreed - hence why everything is optional in microformats-2 - we've learned our lesson.
#singpolymathat was a fight I used to have with people over hAtom all the time
#tantekalso - note that the "required" properties in hCard/hCalendar/hAtom came directly from the requirements in vCard/iCalendar/Atom - and carrying that "required" semantic forward was a mistake (I'll take responsibility for that)
#tantekwhereas what I should have simply done is describe how to generate any such required other-format properties from the respective microformats
#singpolymainvalid in that I was either missing a required attribute (I think author is "required" in hAtom, or was). or the wrong format (I used to publish hAtom with wrong-formatted dates all the time)
#tantekwell the author requirement is going away - so you're fine there (and helped shape the evolution/iteration)
#singpolymayes, I know :) I'm not complaining, just saying
#barnabywaltersI initially published Atom feeds with normal datetimes
#singpolymaI publish valid <time> these days :) Back in the day I would just dump whatever the default format on my Blogger system was, because that used to be the only option.
#tantekbarnabywalters - perhaps we could help him out with some example markup
#barnabywaltersSomehow I don't think it's worth it. But we can but try
#tantekah, he doesn't like the hCard generator - interesting
#singpolymabecause my hAtom parsers can happily handle any date format (and that's all I ever parsed it with) and also because I consider some semantics ("this is the date") better than none ("this is some text")
#barnabywaltersfor me the key is making one that you can customise to fit with your code layout style
#singpolymaI don't know if something like the hCard creator can be made very good, but I'm willing to be proven wrong :) Usually CMS plugins work well IME