wolftune, KevinMarks_, KevinMarks__, KevinMarks, Unifex, parzzix, tantek, j12t and gRegorLove joined the channel
j12t_ joined the channel
KevinMarks_ joined the channel
# 04:04 GWG Financement Qatar sounds particularly suspicious
j12t, nloadholtes, arlen, snarfed, csarven, gRegorLove and KevinMarks joined the channel
KevinMarks__, j12t, KevinMarks_, eschnou, snarfed, sammachin, nloadholtes, friedcell and fkooman joined the channel
KevinMarks_, friedcell, j12t, interactivist, AcidNerd, elf-pavlik and frzn joined the channel
interactivist joined the channel
friedcell and j12t joined the channel
glennjones joined the channel
interactivist, fkooman, eschnou, snarfed, nloadholtes, KartikPrabhu and glennjones joined the channel
nloadholtes and eschnou joined the channel
# 15:40 GWG pdurbin: Anything exciting going on there?
fkooman and tantek joined the channel
# 15:56 pdurbin GWG: actually, I've been saying for a while (especially to tantek and aaronpk ) that I think it's awesome when wikis are based on git. you get versioning for free. and it's distributed. there's a new wiki like this about REST I helped launch: https://trygvis.io/rest-wiki/contributing
# 16:02 GWG Checking out the wiki is an interesting concept.
snarfed and interactivist joined the channel
# 16:19 pdurbin GWG: right. to contribute I simply fork/clone the git repo backing the wiki, push my proposed changes into a branch, and ask for it to get merged in. In practice we use pull requests on GitHub but we aren't dependent on GitHub. The wiki is hosted on a personal domain.
# 16:19 GWG It seems like asking for a merge will create a longer update time.
# 16:24 pdurbin GWG: I could ask for "push" access on the upstream repo if that ever becomes a problem.
# 16:26 GWG I'm just playing devil's advocate, but the concern I think is that a new person would feel a barrier to contribution in that case.
# 16:27 GWG I'm wondering. Github I believe allows some editing from a web interface. Wondering if someone built that functionality into one of the open git front ends
loic_m and elima joined the channel
# 16:31 tantek GWG, indeed, github allows some web editing, but it's a bit slow / cumbersome. Mediawiki's editing UI is MUCH more responsive, thus much more conducive to simple drive-by incremental edits, thus more participation / contribution.
# 16:32 Loqi Ok, I'll tell them that when I see them next
# 16:34 pdurbin one could probably configure a git repo for drive by edits if one desired
# 16:35 tantek !tell elf-pavlik yes I tried that five years ago when I added PuSH support to tantek.com Atom ActivityStreams 1.0 feed, which Evan was able to realtime subscribe to from Statusnet. Support is still there.
# 16:35 Loqi Ok, I'll tell them that when I see them next
# 16:36 tantek pdurbin: sure, one could code nearly anything - but no one has - that's my point. A fast responsive UI is essential, without it - no amount of backend fanciness matters in anyway.
# 16:37 pdurbin tantek: my UI for editing the wiki is vim and I love it.
# 16:37 tantek Also it must be as simple as "Edit", "View Changes", "Save" in buttons in the UI. No command line mumbo jumbo.
# 16:38 elf-pavlik "Formats based on XML or other data serializations are out-of-scope."
# 16:38 pdurbin tantek: these are a bunch of REST nerds. the audience is not so broad
# 16:38 elf-pavlik which includes Atom ActivityStreams 1.0 feed
# 16:41 elf-pavlik tantek, how about doing it now and basing on what we *currently* work on in Social WG?
# 16:41 tantek elf-pavlik: and the sad thing is that AS2 is more complex than AS1 (for simple / common cases), and a one-off JSON format is also more work than incremental changes to Atom, and worse, JSON-LD adds needless complexity too.
# 16:42 tantek thus it is a much worse value proposition now, for me, or anyone on the indieweb, than AS1/PuSH was before
# 16:42 tantek before: AS1/PuSH was relatively easy to implement, and there were existing *public* deployed consuming implementations
# 16:43 tantek now: AS2/JSON-LD overly complex in many ways, ZERO existing public deployed consuming implementations
# 16:43 tantek as fars "what we currently work on" in Social WG, I'm hoping that we either GREATLY simplify AS2/JSON(LD), OR (more likely) come up with MUCH simpler counter proposals.
# 16:43 Loqi tantek meant to say: as far as "what we currently work on" in Social WG, I'm hoping that we either GREATLY simplify AS2/JSON(LD), OR (more likely) come up with MUCH simpler counter proposals.
# 16:44 elf-pavlik do you work on such simpler proposal by any chance? or know someone else who does?
# 16:44 pdurbin GWG: I'm not sure what other git front ends allow. The wiki software we're using (ikiwiki) does allow you to edit via a web interface. The edit goes into the git repo when you click save. I think this addresses the concern by tantek. I just appreciate that I can interact with the wiki using git and vim if I want.
# 16:45 tantek elf-pavlik: this whole community is actively deploying and using *numerous* implementations/solutions to the high level user goals of Social Web WG
# 16:46 tantek elf-pavlik: if you want a JSON syntax, merely use any of the commonly available open source microformats2 parsers on any of our sites - and you'll get back a fairly simple usable JSON
# 16:47 elf-pavlik i understand you would like to see similar deployments based on AS2.0 before considering aligning your current work with it?
# 16:47 tantek (as far as the charter item regarding social activity streams syntax is concerned)
# 16:47 tantek elf-pavlik: note comparison to AS1/Atom above.
# 16:49 tantek I would actually drop my site's AS1/Atom/PuSH 0.3 support, except there are real world sites that still consume it (e.g. Quitter), thus I keep it for cross-site federation compatibility, even though it's not a format / standard that I actively work on or promote.
# 16:50 tantek elf-pavlik: have you made any progress with posting notes (tweets) on/from your own site instead of Twitter?
# 16:51 elf-pavlik also started converting W3C IRC logs to SIOC
# 16:52 tantek the backend doesn't matter as much as where it is on your site where people can read it.
# 16:52 tantek after all, the reason you send email or tweet is not for some backend plumbing format, but so that others *read* it.
# 16:52 elf-pavlik people can read it with whatever apps they want to use for it
snarfed joined the channel
# 16:53 elf-pavlik just as with email clients, irc clients, xmpp clients i don't want to impose my UI preferences on others
# 16:53 tantek that app is typically called a browser - it's up to you to provide the browser readable URLs that provide HTML, just as Twitter does
# 16:53 elf-pavlik many people use Twitter via their native mobile apps
# 16:53 elf-pavlik i would like to see some usage statistics
# 16:53 tantek except we're talking in the context of social *web*, not social *random apps*
# 16:54 tantek so you can't escape making your stuff work on web
# 16:54 elf-pavlik still i can provide some default app for those who just type URL in a browser
# 16:54 elf-pavlik what do you mean by 'work on the web'?
# 16:55 elf-pavlik URLs, preferably HTTPS ?
# 16:55 tantek elf-pavlik: already answered that for jasnell months ago but I suppose I can provide another permalink for you
# 16:55 elf-pavlik The Web != HTML
# 16:56 elf-pavlik did you write somewhere more about how you arrive at this opinion?
# 16:57 aaronpk tantek: pdurbin: fwiw the GitHub wiki is actually a git repo, but you don't need to know that to use the web UI that lets you edit it like mediawiki
wolftune joined the channel
# 16:58 Loqi The social web refers to the subset of the web that has social content, that is, content, like posts, which has obvious visible authorship (even if pseudonymous), and mentions other people or other social web content, via URL reference, not just name https://indiewebcamp.com/social_web
# 16:59 Loqi A public social website is a public website (at a URL you can curl), that has social content like posts which have obvious visible authorship (even if pseudonymous), and mentions other people or other social web content, via URL reference, not just name https://indiewebcamp.com/public_social_website
# 16:59 tantek elf-pavlik: not just my opinion - based on actual common human usage of the web - using browsers, that load URLs to HTML.
# 17:00 tantek elf-pavlik: plenty of folks have tried working on non-HTML replacements for the web and failed. I'm not interested in those discussions as they have too little chance of being relevant to be worth the time.
Erkan_Yilmaz joined the channel
# 17:03 elf-pavlik do you have examples of live updates with HTML payloads?
KevinMarks__ joined the channel
# 17:04 tantek there are more implementations / public sites of live updates with HTML payloads than there are AS2.0 implementations
# 17:05 aaronpk elf-pavlik: yes, Woodwind and the IRC logs push HTML to the browser in realtime, using eventsource and websockets
pdurbin_m joined the channel
# 17:06 snarfed ugh, i hate silos. bridgy would be so much easier if i could drop silo support altogether. :P
# 17:08 snarfed elf-pavlik: ben_thatmustbeme got that working recently
# 17:08 elf-pavlik tantek, why send HTML if sender can't know how client may want to render it?
# 17:09 tantek because it's the easiest and most dependable way to do so
# 17:09 elf-pavlik in most cases i just want to get the RAW data
# 17:09 tantek no in most cases users want to see the data nicely
# 17:09 tantek hence websites like twitter and facebook work hard on a good /ux
# 17:09 elf-pavlik snarfed, do you have link to the code which runs on browser, receives data pushed from server and renders it ?
KevinMarks_ joined the channel
# 17:10 snarfed elf-pavlik: nope, that's why i pointed you to ben_thatmustbeme
# 17:10 elf-pavlik tantek, which email client you consider THE NICEST?
# 17:10 elf-pavlik or which calendar client should everyone use?
# 17:11 elf-pavlik i wonder how many people here use different client apps to connect to this channel
# 17:11 snarfed elf-pavlik: ...but just as another data point, re HTML payloads in AJAX requests, it's actually a surprisingly common pattern for closed source webapps too. (surprising given all the /JSON/API/JS framework "rage")
# 17:11 snarfed (closed source so i don't have many citations, but i'll see if i can find one)
# 17:11 tantek you would publish your own indie /event posts on your own website, and /RSVP to others events using your own website
# 17:12 elf-pavlik so why someone should send me or receive HTML? you need RDFa or Microdata or Microformats to get structured data in it
# 17:12 elf-pavlik BTW neither micropub nor webmention sends HTML...
# 17:12 tantek elf-pavlik: same answer as before: in most cases users want to see the data nicely, hence websites like twitter and facebook work hard on a good /ux
# 17:12 elf-pavlik s/sends/receives/
# 17:12 Loqi elf-pavlik meant to say: BTW neither micropub nor webmention receives HTML...
# 17:13 tantek last time I checked, form-encoded is part of HTML ;)
KevinMarks__ joined the channel
# 17:13 tantek and you're right, they receive HTTP POST - different layer
# 17:14 elf-pavlik URLs to resources, which can support Accept: text/html and reply with ... RDFa :)
# 17:14 tantek right - tommorris has RDFa in addition to his microformats because he likes that
# 17:14 elf-pavlik acutally I don't mind serving you RDFa if you prefer valid HTML i can serialize to it automatically
# 17:15 elf-pavlik i don't want to send microformats but HTML
# 17:15 tantek I'd say check more with tommorris on that as he's figured out how to make it work for him
# 17:15 elf-pavlik Microformats != HTML
# 17:15 elf-pavlik one can put structured data in it with RDFa and Microdata as well
# 17:15 elf-pavlik just as RDFa
# 17:15 tantek microformats has worked with plain HTML since HTML4
# 17:16 tantek elf-pavlik: hence I referred you to tommorris is you're actually interested in building that
# 17:16 elf-pavlik can you get data from websites serving RDFa or Microdata - pure HTML
eschnou and KartikPrabhu joined the channel
# 17:16 tantek no you can't because both of those require extra attributes not define in HTML
# 17:16 Loqi tantek meant to say: no you can't because both of those require extra attributes not defined in HTML
# 17:16 elf-pavlik In dusseldorf I can add sending webmentions to HTML+RDFa so you can get data from it :)
# 17:17 tantek elf-pavlik: see how tommorris does it - he has webmentions working with other people
# 17:17 tantek yes, microformats are a format, but work with plain valid HTML
# 17:18 tantek that's the point, microformats did not need to change another spec in order to be compatible, valid etc.
# 17:18 elf-pavlik I'll make sure that my RDFa validates, which validator you would like me to check with?
# 17:18 elf-pavlik i can serve HTML why not, tools can automate serialization to it so whatever
# 17:18 pdurbin_m aaronpk: I wonder how hard it would be to add indieauth support.
# 17:18 tantek not sure, I don't care about consuming RDFa since it's harder than microformats
# 17:19 tantek I prefer to work on simpler formats / protocols first
# 17:19 tantek hence my point about the problems with AS2/JSON(LD)
# 17:19 elf-pavlik if someone writes JS serializer microformats JSON to microformats HTML i could even support that with profile parameter in Accept
# 17:19 tantek elf-pavlik: if you prefer to build more complex things - go ahead, no one will stop you - but it's not reasonable to ask others to work on more complex things, when simpler things will do
# 17:20 tantek I mean you can try - but it adds to the cost for everyone
# 17:20 tantek hence why it's better to always go with the simpler options first, and see if they can be made to work
# 17:21 elf-pavlik "find my friends and their friends in Paris who have accomodations to offer"
# 17:22 elf-pavlik or wait, something almost build with LDF and RDF open trip planer (multimodal)
# 17:22 elf-pavlik using Open Data
csarven joined the channel
# 17:24 elf-pavlik subjectivly higher complexity of RDFa over Microdata adds IMO not even comparable querying power
# 17:24 elf-pavlik and you still can use HTML if you want to
# 17:24 elf-pavlik and James and Evan can use JSON(-LD)
# 17:24 elf-pavlik and Henry and Melvin can use Turtle, no problem
# 17:25 elf-pavlik tantek, i agree no sense to talke more without live deployments which you can try yourself
# 17:25 tantek elf-pavlik: I don't understand how you leapt from aaronpk's metrics to "accomodations to offer" as a use-case so I'm afraid you're not making much sense.
# 17:25 elf-pavlik otherwise it turns into some religious evengelizing
# 17:25 tantek elf-pavlik: or I try to put it as "philosophy" - which has less potential to be offensive.
# 17:25 elf-pavlik tantek, how about cars available to share, or bikes available to share?
# 17:26 tantek elf-pavlik: there are existing websites for those, perhaps you can document research about how they work.
# 17:26 tantek documenting the UX of existing solutions is how you can figure out how to build a UX of an alternative.
# 17:26 tantek instead of jumping into plumbing and protocols.
# 17:27 tantek which rarely produce a UX, or a usable one at that.
# 17:27 elf-pavlik but you one sholdn't cuple UI with data
# 17:27 tantek the only data you can depend on is when its presented in a UI
# 17:27 elf-pavlik and this way try to impose same UX on people with different preferences
# 17:28 elf-pavlik especially UI of your choice, which you understand and feel comfortable with
# 17:28 tantek having one UI to start with, is more important than worrying about UI of choice (multiple UIs)
# 17:28 elf-pavlik tantek, i agree with many of your points, and till you see some running apps using RDF in powerful way we can't go beyond e-waving hands
# 17:29 elf-pavlik Facebook++ ;)
# 17:29 elf-pavlik diversity--
# 17:30 tantek elf-pavlik: since this is in the context of the social web working group, I am more interested in public social websites than random apps
# 17:30 elf-pavlik public *data* which appears on static website as well feeds modern dynamic apps
# 17:31 elf-pavlik "This form data set encoding is in many ways an aberrant monstrosity, the result of many years of implementation accidents and compromises leading to a set of requirements necessary for interoperability, but in no way representing good design practices. In particular, readers are cautioned to pay close attention to the twisted details involving repeated (and in some cases nested) conversions between character encodings and byte sequences."
# 17:32 elf-pavlik sounds exciting to build on it \o/
# 17:36 elf-pavlik tantek, i go back on adding SSE to two websites i run, and just because i argue about some things doesn't mean that i disagreew with everything yous say :)
# 17:38 pdurbin_m aaronpk: yes, the standard GitHub wiki (gollum) is backed by git but it's a flat list of pages. I like how ikiwiki lets you organize your content into subfolders of arbitrary depth.
# 17:40 pdurbin_m and unlike Mediawiki, there is no database tax :)
gRegorLove joined the channel
# 17:41 KartikPrabhu elf-pavlik: I suppose the gist of the above discussion (caught up on logs) is that you are free to use anything you want for your data. So far HTML is the most widely consumed/read form of data and so it is better to build on it than reinvent the wheel
# 17:45 tantek elf-pavlik: that comparison is very out of date, biased, and is superseded largely by microformats2
# 17:47 tantek elf-pavlik: again, it comes back to simpler solutions are better, and dealing with (and asking others to deal with) the more complex solutions is bad because it adds cost for everyone - something that manu's and others "comparisons" don't understand
# 17:47 tantek more features does not mean better necessarily, that's the problem
# 17:48 tantek elf-pavlik: for generic comparisons like that, also #microformats is probably more ontopic
# 17:48 tantek here let's focus on helping you improve your personal site
# 17:49 elf-pavlik hope to deoploy tomorrow
# 17:50 tantek KartikPrabhu: indeed, I'm trying to refocus the discussion to be more about being productive on one's own indieweb site
# 17:50 tantek if you could capture what you're working on there
# 17:50 tantek if you have a new personal website, you can sign in with it and create a new user page
# 17:51 elf-pavlik holding fiends baby for few minutes - afk
# 17:51 elf-pavlik tantek, why do you think it changed?
# 17:52 tantek because you said so above *my website* with a different domain
# 17:53 elf-pavlik maybe wrong use of gramar - subject predicate object got mixed ...
# 17:54 elf-pavlik i start using OPO ontology for current activities / real time presence
# 17:56 tantek elf-pavlik too hard to keep track of your different working on things as just lines in IRC - hence suggesting you collecting them on your user page
# 17:56 elf-pavlik how can i use that vocab in microformats html? very easy thing in RDFa!
# 17:56 tantek that page is some abstract archicture thing - not a simple list of things you're working on
# 17:57 elf-pavlik it has vocab to publish in comon way what i curently do
# 17:57 tantek if you capture a list of all the things you just stated even today that you're working on - then there's a beter chance that someone may be able to help you answer such questions about them
# 17:58 elf-pavlik i prefer to publish it on my own site
# 17:58 tantek also it helps to see what you think is a higher priority for you (of all the things you are working on)
# 17:58 elf-pavlik recently added prujects section
# 17:58 tantek great - publish a "working on" page on your own site, and then link to it from your user page
# 17:59 tantek I suggest the wiki because it is the simplest / fastest way to do so - but if you have something better, go for it
# 17:59 elf-pavlik thanks to OPO you or your app can 'follow nose'
# 18:00 elf-pavlik via RDF properties ~! link relations
# 18:00 tantek you can't follow something that doesn't have URLs
# 18:01 tantek saying so doesn't make it so - add a URL to your user page to prove it
# 18:01 elf-pavlik wiki -> my homepage -> portable linked profiles project (for now)
# 18:01 tantek otherwise it sounds / looks like you're interested in a theoretical discussion not anything practical,
# 18:02 tantek can you provide a specific link to your "Working On" the way KartikPrabhu did above?
# 18:03 tantek it's not reasonable to ask people to dig through your personal site
# 18:03 elf-pavlik links to rel="foaf:currentProject"
# 18:03 tantek and that's also not any of the things you mentioned above
# 18:03 tantek so this all seems very disconnected elf-pavlik
# 18:04 tantek hence a request that you just provide a simple flat list
# 18:04 elf-pavlik i'll add more there right now!
snarfed joined the channel
# 18:06 tantek ok I see your projects heading - but it doesn't mention any of the other things you said you're working on above
# 18:06 tantek nor does it have an ID so I'll use a fragmention
# 18:08 elf-pavlik tantek, what do you see missing ID ?
# 18:08 tantek on your Projects heading - so I used a fragmention to link to it instead
# 18:09 elf-pavlik oh *fragment*
# 18:09 tantek do you have any projects that you are working on for your personal site? or just group / collabrorative efforts
# 18:10 tantek which is what is missing. it's only called a "Fragment_identifier" when in a URL
# 18:11 elf-pavlik I dont wont to use any other IDs than IRIs
# 18:11 elf-pavlik preferably https: or http:
# 18:11 elf-pavlik to avoid silos :)
# 18:12 GWG I just had a nice long walk. I'm thinking of doing something complex to my work again.
# 18:12 tantek ok that doesn't make sense since silos also use http https and URLs
# 18:12 elf-pavlik if you use *just* fragment for ID, don't you put data in silo?
# 18:13 tantek elf-pavlik: "work on a model instead" sounds like plumbing before presentation - which is generally acknowledged to be a flawed methodology here.
# 18:13 elf-pavlik i will add to my homepage identifier for projects !
# 18:14 elf-pavlik and use in HTML 'about' or 'resource' attribute when embedded partially on my homepage...
# 18:17 elf-pavlik tantek, what do you use in 'plain' HTML5 to identify nested resources with full URL ?
# 18:19 kylewm hmm, I'm intrigued by this secret project from benwerd and erinjo
# 18:20 elf-pavlik if i want to list on my homeage a violent person's and identify her with URL website (~ h-card) but don't want to put clickable link since i do not want people to visit that website
# 18:20 elf-pavlik let's say in 'stay away from those folks' sub page
# 18:21 tantek elf-pavlik: could you document that use-case somewhere either on the wiki or your personal website? I have not heard anyone mention it here before
# 18:21 zachdonovan maybe this is irrelevant, and I certainly understand from context, but
# 18:22 elf-pavlik tantek, we can discuss later clar way of identify the subject (entity) of statements we publish, i really need to get back on coding now :)
# 18:22 Loqi architecture astronomy is the practice of analyzing problems, seeing patterns, and then generalizing to higher and higher level abstractions on top of those patterns to the point where the abstractions become so general, so vague, so detached from the original problems being analyzed, that they don't mean anything at all https://indiewebcamp.com/architecture_astronomy
# 18:22 GWG !tell acegiak I had an idea. Need to bounce it off someone.
# 18:22 Loqi Ok, I'll tell them that when I see them next
fkooman, tvn, benwerd, frzn, KevinMarks_, eschnou and interactivist joined the channel
tvn joined the channel
tantek and tvn_ joined the channel
# 19:58 KevinMarks_ Thinking about inline embedding of tweets and an indieweb equivalent
# 19:59 tantek KevinMarks: you need a start and end fragmention for that - or a fragmention of the entire quote
# 20:00 GWG KevinMarks_: Why not a display library that can be ported into different languages. Isn't there something like that for comments display?
j12t joined the channel
# 20:03 KevinMarks_ The way twitter.com embeds tweets inside another tweet is much more lightweight in ux terms than the tweet embed that is generated by the js decorator
# 20:04 GWG KevinMarks_: That was why I asked about the broader implications
# 20:14 tantek KevinMarks: I have my doubts about a fragmention of the entire quote - due to problems with long URLs (already documented in why short URLs).
# 20:15 tantek KevinMarks: if you believe "fragmention of the entire quote seems like a good idea", go ahead and start publishing some notes with such fragmentions on your own site, and report back with permalinks and your experience doing so.
nloadholtes joined the channel
# 20:17 tantek !tell aaronpk,kylewm,KartikPrabhu,GWG I am interested in your review and opinions of https://indiewebcamp.com/proof_of_work - both criticisms and suggested improvements in the hopes of continuing to focus this channel on more practically productive conversations, as quickly as possible.
# 20:17 Loqi Ok, I'll tell them that when I see them next
LCyrin and tvn_ joined the channel
friedcell joined the channel
elima joined the channel
# 20:30 KartikPrabhu !tell tantek I think a "Why" section would be good. Answering "why a proof of work" is demanded/good .
# 20:30 Loqi Ok, I'll tell him that when I see him next
# 20:37 Loqi slack/tantek: The why is an interesting challenge in itself, as it's anywhere from efficiently getting new folks to be productive for themselves, to defending the community from academics.
# 20:48 Loqi slack/tantek: Also, as I think ben_thatmustbeme mentioned indirectly, increasing the signal to noise ratio of the channel archives.
gRegorLove joined the channel
j12t and interactivist joined the channel
KevinMarks__, KevinMarks___ and KevinMarks_ joined the channel
KevinMarks__ joined the channel
loic_m joined the channel
snarfed and benwerd joined the channel
j12t and tantek joined the channel
# 22:12 elf-pavlik if someone adds it to Accept header i could also render microformats HTML
# 22:14 KartikPrabhu elf-pavlik: what does that document mean? sorry not familiar with w3c jargon
# 22:15 tantek profile relations are largely obsolete, as they were a theoretical demand that nobody actually bothered implementing or needing.
# 22:16 tantek AFAIK profile stuff is academic waste of time
# 22:16 elf-pavlik i don't talk about HTML but HTTP headers
# 22:16 tantek no actual use-cases need it, nor did anyone bother to implement any of it
# 22:17 tantek KartikPrabhu: links to profiles like that were requested by various Sem Web folks, but except for a little bit of GRIDDL work (also obsolete), no one bothered with them
# 22:17 KartikPrabhu also one does not need a new HTML link header to get microformats data. Get the HTML page and parse the microformats in it
# 22:17 tantek so I'd wait until someone deploys something that consumes them before wasting any time on it
# 22:17 elf-pavlik when someone makes request to my homepage, by default i return RDFa but i requestor asks for Microformats or Microdata I could also respond with that
# 22:18 tantek that AS2 issue also looks irrelevant as its not blocking any implementations
wolftune joined the channel
# 22:18 elf-pavlik RDFa gives you valid HTML
KevinMarks_ joined the channel
# 22:19 elf-pavlik I don't want to mix RDFa, Microdata and Microformats in the same response
# 22:19 tommorris just use microformats and RDFa and forget microdata ever existed
# 22:19 elf-pavlik but could adjust embeding structured data in HTML based on requested profile
# 22:20 tommorris microformats because it works now and RDFa because it can represent stuff that's not in microformats
# 22:20 elf-pavlik RDFa works now as well
# 22:20 elf-pavlik different people have different preferences and Accept HTTP header allows to state them clearly
# 22:21 tommorris the point of RDFa is to embed statements in HTML. text/html is the right MIME type.
# 22:21 elf-pavlik yes but one can add profile to stay more precise
# 22:22 Loqi Ok, I'll tell them that when I see them next
KevinMarks__ joined the channel
# 22:23 tommorris elf-pavlik: there isn't an accept-profile header. just send both. is there any use case for someone wanting to specify the type of data-in-HTML format?
# 22:23 tommorris like, if you get an HTML document with microformats and you don't want to parse them, don't parse them. same for RDFa
# 22:24 elf-pavlik i don't want to mix templating for those two
# 22:25 elf-pavlik i can have nice and clean JSON-LD -> RDFa rendering
# 22:25 elf-pavlik but on demand i can also send Microformats
# 22:25 tommorris elf-pavlik: if you are publishing JSON-LD, why publish RDFa?
# 22:25 elf-pavlik if someone *expects* them in response i see no problem with making it explicit in request
# 22:26 elf-pavlik i also publish HTML and could of course embed JSON-LD in HTML
# 22:26 elf-pavlik i don't care about GRDDL
# 22:27 elf-pavlik what is content negotiation?
# 22:29 elf-pavlik what is conneg
# 22:29 Loqi conneg is short for HTTP Content Negotiation, a method by which a browser or other web client can request content of various types from a web server, and depending on what is requested, and what the server supports, it tries to provide the best it can https://indiewebcamp.com/conneg
# 22:30 elf-pavlik trying to find it now...
tantek joined the channel
# 22:35 tommorris elf-pavlik: I still don't know what practical point there is to content negotiating whether you get text/html with RDFa in it vs. RDFa + microformats and/or microdata, but I guess if you wanted a URL to arbitrarily represent microformats, why not http://microformats.org/
# 22:35 elf-pavlik i don't think us two agreeing here will do ;)
# 22:36 tommorris elf-pavlik: I mean, the URL is arbitrary anyway because nothing is ever going to be triggered from it. you may as well use urn:microformats
# 22:36 elf-pavlik not full media type but just profile
# 22:36 tantek elf-pavlik: it seems like you are trying to make more work for yourself, so it's not clear what your motivation is
j12t and snarfed joined the channel
# 22:37 tantek tommorris's suggestion of just use both microformats + RDFa seems reasonable, and passes the /selfdogfood test - he does so himself already, on his public website
# 22:37 elf-pavlik 1) i don't want to mix RDFa and Microformats in my responses but could adjust rendering to use one depending on request
# 22:37 tantek elf-pavlik: making up extra work for people (whether others or yourself) tends to be frowned up in the indieweb community, because it is harmful to everyone's productivity.
# 22:38 KartikPrabhu tommorris: any specific reason to use RDFa? is it some kind of data that mf2 can't represent? If so would it be useful to have an mf2 version of it?
# 22:38 elf-pavlik what extra work? if one expects to ger microformats (or RDFa or Microdata) why not to ask for it?
# 22:38 tantek elf-pavlik: in that case, assume microformats are the default, because they don't require any elements or attributes beyond what is explicitly specified in HTML5 for text/html
# 22:39 tantek elf-pavlik: do you use that dataset? at what URL on your personal site?
# 22:39 KartikPrabhu I can. But I haven't found use for them. Hence my question to tommorris if he has found use for such things
# 22:39 tantek if not, it fails the selfdogfood test, and you shouldn't bother recommending it here in #indiewebcamp
# 22:39 tommorris elf-pavlik: no microformats parsers implement support for specifically asking for microformats, because nobody publishes it, because that's too much work
# 22:40 tantek asking for "specifically asking for microformats" is extra work, and is unnecessary, therefore it should be ignored
# 22:40 tommorris for me to implement it in mf2py would require you to document like five or ten people using it, otherwise it's pointless overhead that serves no purpose. I'm reasonably sure barnaby and other implementers would want to see real-world usage before putting in an HTTP header to implement an unspecced requirement that serves no real purpose.
# 22:41 tantek tommorris: is there a general approach in the Sem Web community to filter out these kinds of theoretical requests?
# 22:41 tommorris because semantic web pipe dreams are fun, but when they start taking time away from me drinking mimosas and require me to write code, I need actual use cases
# 22:42 tommorris eventually the w3c closes the working group and the world moves on as if nothing happened. :)
# 22:42 elf-pavlik vocab="http://schema.org/"
# 22:42 tantek already did today, hence request for ID attribute
# 22:43 tantek that looks like a lot of unnecessary extra work
# 22:44 tommorris elf-pavlik: why not just add microformats2 to that? nothing wrong with using both?
# 22:44 KartikPrabhu elf-pavlik: I would very much like to know any functionality RDFa gives you other than just embedding data for the sake of embedding data. I had RDfa on my site but removed all traces since it was not useful to do anything
# 22:45 KartikPrabhu I kept microformats since it gives you webmention and nice rendering of comments and all that
# 22:45 tommorris "Alice is browsing Bobs social profile on the web and finds an interesting note, Alice navigates to the notes permalink and decides she wants to inspect the Activity Streams representation of the note"
# 22:45 tommorris Bob responds: it's just HTML, use that. why add JSON? ;-)
# 22:45 tantek.com edited /RSS (+134) "/* Causes Plumbing Misfocus */ linky linky, conclude with method of correcting for plumbing misfocus"
# 22:46 KartikPrabhu Alice is browsing Bobs social profile on the web and finds an interesting note, Alice navigates to the notes permalink and decides she wants to inspect the Holographic Nonlinear Doohickey representation of the note"
# 22:47 gRegorLove I don't intend to send an HTTP header to request mf2. I'll just request the page and if I find mf2, cool.
# 22:47 KartikPrabhu tommorris: you should blog that user story. If not I will and add attribution link to logs :P
# 22:48 tommorris elf-pavlik: everyone uses JSON because everyone uses JSON. five or ten years ago it was XML because everyone read that XML was the thing. technology is driven by fashion as much as it is by sound technical reasoning.
# 22:49 tommorris and because everyone feels they need "an API" and can't grasp that HTML as an API solves like 80% of use cases of read-only APIs
# 22:49 elf-pavlik tommorris, didn't you say that you use RDFa where you miss microformats terms?
KevinMarks_ joined the channel
# 22:49 tommorris elf-pavlik: yep. and because I like to show that it's possible
# 22:49 elf-pavlik HTML =! microformats
# 22:50 Loqi elf-pavlik meant to say: HTML != microformats
# 22:50 tommorris <footer class="vcard h-card" typeof="schema:Person foaf:Person geo:SpatialThing gnd:Person
# 22:50 tommorris yago:LivingPeople yago:SoftwareDeveloper yago:LGBTPeopleFromEngland yago:Adult109605289
# 22:50 tommorris yago:AlumniOfTheUniversityOfLondon yago:PeopleFromSurrey" about="https://tommorris.org/"
# 22:53 tommorris elf-pavlik: microformats goes in HTML and is perfectly compatible with HTML.
# 22:54 tommorris elf-pavlik: all I can say is, if you implement microformats in such a way that you have to send weird custom headers to get the microformats, nobody will read the microformats and any benefit will be lost.
# 22:55 Loqi slack/tantek: No extra header is need because microformats are *just* HTML.
# 22:55 tommorris I mean, let's say you convinced me and barnaby and glennjones and the other mf2 parser devs to add that support, then what about in-browser parsing? what about if Operator came back from the dead and started parsing microformats2 - convince Firefox to send magic headers?
# 22:56 elf-pavlik browser extensions can modify headers AFAIK
# 22:56 Loqi slack/tantek: Whereas RDFa requires extra attributes beyond HTML5, thus could require explicit profile support.
# 22:57 tommorris eh, not sure about that. RDFa works just fine in HTML5. you just ignore validators, like everyone does already. :)
# 22:57 Loqi slack/tantek: Microdata is a mixed bag because it was abandoned by w3c, uses attributes beyond HTML5, yet attributes are in WHATWG HTML.
# 22:57 tommorris I know hsivonen's validator checks RDFa Lite 1.1, not sure about RDFa Core though
# 22:58 Loqi slack/tantek: Tommorris sure, "works fine" is a good metric :) however if anyone is proposing requiring explicit profiles for anything, they should start with requiring it to get back RDFa, since it uses attributes beyond what is specified in HTML5 and thua text/html.
# 23:00 Loqi slack/tantek: Anyway this is all a /plumbing centric distraction with no uses cases to back it up.
# 23:01 tommorris plumbing rule: unless human excrement is coming out of the pipes, I do no more plumbing than needed.
# 23:02 Loqi slack/tantek: Should we debunk the assumed relevance of conneg?
# 23:02 Loqi slack/tantek: To discourage distractions with it?
# 23:05 tommorris conneg is nifty, but I'm leaning strongly in the other direction - HTML as the One True Format, and using conneg to return the data as the result of the parsed HTML (if you request JSON, you get back data from mf2py; if you request RDF, you get back the RDFa parsed into whatever RDF format you asked for)
# 23:05 tommorris HTML is the format, everything else is just a parsing thereof.
# 23:05 KartikPrabhu tommorris: in that case I would just return HTML and the requester could do whatever the hell parsing they want
# 23:06 tommorris KartikPrabhu: yep, but all I'm doing is providing a layer of convenience for them.
benwerd joined the channel
# 23:06 Loqi slack/tantek: That way we can at least capture your wisdom for others to reuse
KevinMarks_ joined the channel
# 23:07 Loqi slack/tantek: Elf-Pavlov no need for the play by play - as no one else here has expressed any interest in profiles.
# 23:08 KartikPrabhu now if only someone makes that for HTML+mf2 -> Atom , I'll ditch my DB -> Atom code
# 23:08 Loqi slack/tantek: Kartik what about barnaby's converter?
# 23:08 tommorris KartikPrabhu: I am tempted to add iCalendar, Atom and vCard output to mf2py at some point
# 23:09 KartikPrabhu tantek was thinking of self-hosting with python instead of over-loading barnabywalter's server
# 23:09 KartikPrabhu tommorris: adding to mf2py seems excessive. But I'd be for a shim that takes mf2py output and converts to these things
# 23:09 tommorris KartikPrabhu: if you write an mf2py-to-Atom converter, I'd be happy to merge it into mf2py
# 23:10 KartikPrabhu because then you'd have to add JSON-LD, Holographic Quantum Data format and all those things
# 23:10 tommorris nope, because microformats aren't tied to those data formats
# 23:10 Loqi slack/tantek: Kartik - a I'm away from laptop now - could you add a /proof_of_work#Why section with perhaps some of the thoughts I expressed above?
# 23:11 tommorris and benevolent dictatorship means I can play favourites. :)
# 23:12 Loqi slack/tantek: Thanks Kartik, appreciated.
tmro joined the channel
# 23:15 Loqi slack/tantek: Also where was that spot in the logs the past couple of days where ben_thatmustbeme was complaining about the logs being too long to catch up on? Or was it too noisy?
# 23:15 Loqi slack/tantek: We need to cite that as another real world reason why we need to do active /proof_of_work challenges in the channel
webhat joined the channel
snarfed joined the channel
# 23:18 KartikPrabhu snarfed: does it need a full install of sorts to go between mf2 and Atom? i.e. with the API keys and all that?
# 23:19 KartikPrabhu snarfed: sweet! I'll try to make up something with this then :) thanks
# 23:19 Loqi slack/tantek: goes trawling through logs
# 23:19 snarfed KartikPrabhu: enjoy! the atom part isn't documented in the readme well but the code has good docstrings. just use the top-level methods in atom.py and microformats2.py
# 23:22 KartikPrabhu blah! JS is so fragile! my fragmentioner.js only works on pages that marginalia.js works on, since I combined both codes into one JS file!
# 23:23 Loqi slack/tantek: Couldn't find it. Maybe it was a private message or unofficial
wolftune and KevinMarks_ joined the channel
# 23:38 Loqi proof of work is a way of responding to questions or requests with a challenge to the requester to do some work before questions are answered; in the context of the IndieWeb such work includes asking what is their personal site, what is the next thing they want to improve on it, all as ways to refocus abstract or plumbing questions on practical personal site selfdogfooding https://indiewebcamp.com/proof_of_work
snarfed joined the channel
KevinMarks_ joined the channel
# 23:59 GWG I'm thinking of trying to get microformats2 support into WordPress by submitting patches. I'm trying to figure out if this is a good idea.