#dev 2018-04-11

2018-04-11 UTC
#
@JmacDotOrg Jmac IndieWeb Rabbit Hole Progress Update: 🐇 https://github.com/jmacdotorg/webmention-perl#webmention 🕳 https://github.com/jmacdotorg/microformats2-perl#name All this is going into Plerd, destined to become live-fire-tested on Fogknife. My own vision-quest to help make the web better… somehow. i’ll explain later, okay? you’ll love it (twitter.com/_/status/983860417750544384)
eli_oat, dougbeal|mb1 and [eddie] joined the channel
#
[eddie] I think the term compatible is clearer to Gen 2/3/4 that a piece of software works with something else based on Desktop software paradigms
eli_oat and [jgmac1106] joined the channel
#
[jgmac1106] I still say you go with indieweb certified and use a validator. could include an svg certifcate or something
#
KartikPrabhu huh! there's no indieweb validator!
#
@oss_pl webmention-perl - Webmention libraries for Perl 5. https://github.com/jmacdotorg/webmention-perl (twitter.com/_/status/983865200414724099)
#
martymcgui.re uploaded /Special:Log/upload "uploaded "[[File:hwc-baltimore-2018-04-10.jpg]]" https://indieweb.org/File:hwc-baltimore-2018-04-10.jpg"
#
martymcgui.re edited /Special:Log/upload () "uploaded a new version of "[[File:hwc-baltimore-2018-04-10.jpg]]"" (view diff)
snarfed and renem joined the channel
#
[jgmac1106] @kartikPrabhu, there are a number of tools to check to see how "compatible a website is with indieweb yes? Couldn't that be scaled?
#
KartikPrabhu [jgmac1106]: not sure even what "compatible with indieweb" means
#
[jgmac1106] I also hate jumping into the dev channel by accident when not talking actual dev stuff, sorry
#
KartikPrabhu there is no one set of criteria that makes you "indieweb"
#
[eddie] KartikPrabhu, The context of that conversation is regarding Micropub/Microsub
#
[eddie] Although, I don’t think certification is needed. The key is trying to figure out a way to explain if a Micropub/Microsub client can work with a given service without using protocol names that can be confused for Gen 3/4 users
#
KartikPrabhu [eddie]: sure but then you are checking for specific things like micropub/microsub and not some "indieweb compatibility"
#
[eddie] So the term IndieWeb compatible is a term I tossed out there, on our journey to discover a Gen 3/4 friendly name. (We still haven’t come up with one yet)
#
[eddie] It was based on the analysis of tools like Ulysses and Byword using the term Wordpress compatible in place of XML-RPC
#
KartikPrabhu well then my site not "indieweb compatible" either
#
[eddie] the IndieWeb is composed of so many things, that it doesn’t match up apples to apples, hence the search for a suitable term
renem and [kaushal_modi] joined the channel
#
[kaushal_modi] [eddie] I don't think there's one term to encompass everything. Everyone uses one or more different Indieweb features (hope that's the right term now).
#
[kaushal_modi] For example, my site has proper microformats2 support, and displays the received Webmentions. I doesn't yet? deal with Micropub/sub/* etc (as I don't know anything about those).
leg and [eddie] joined the channel
#
[eddie] Yeah, that’s right. However with Micropub and Microsub being such integral features of a reader. As people shift towards IndieWeb from Twitter/Facebook, there needs to be an easier way to described a server compatible with Micropub/sub without using overly techy terms like Micropub/sub, especially considering they are one character apart 🙂
#
grantcodes pstuifzand: The setting is stored on the server now so no, will need to fix the actual bug causing it
#
grantcodes Check your console and open an issue if you can
eli_oat and iasai_ joined the channel
#
aaronpk There's a difference between "compatible" and "checks all the feature boxes and does everything"
#
KartikPrabhu aaronpk: sure, so what counts as compatible?
#
KartikPrabhu as in "indieweb comaptible"
#
aaronpk Does it work with other sites
#
KartikPrabhu in what sense?
#
aaronpk I'm not sure "indieweb compatible" is the right term
#
KartikPrabhu right, that was my point
#
aaronpk But we're talking about a specific thing regardless of the name
#
aaronpk And we need a name for it
#
aaronpk That isn't specs
#
aaronpk The closest analogy is ostatus since it isn't a spec itself but it's a collection of specs
#
aaronpk tho in the end people talk about whether things are compatible with specific products, like whether mastodon is compatible with gnusocial
#
KartikPrabhu right
#
KartikPrabhu maybe we should look at how RSS/Atom compatibility is talked about
#
aaronpk "Podcast feed" is an interesting example of a term that is not a spec and also not tied to a vendor that describes a pretty specific set of features and a format
[eddie] joined the channel
#
[eddie] That’s true, that is a pretty good example
#
[eddie] and definitely not locked on IndieWeb compatible for all the reasons in the chat. Haha
#
aaronpk OTOH "CD" is a spec that did end up as a term in common usage
#
[eddie] MP3, would be considered that as well?
#
[eddie] It is not AS widespread but has reached pretty far
#
aaronpk Yea
#
[eddie] I guess one questions is whether Micropub/sub should be a separate name group from mf2/webmentions for layman terminology.
#
aaronpk I'd say mp3 counts because you could go to a retail store and buy something labeled an MP3 player
#
[eddie] Since one is about client to server and the other server to server
#
aaronpk It's very possible those should be talked about separately since I could see buying products that support one set and not the other because the other just doesn't apply
#
[eddie] Yeah, that was mine thinking. They are only related because the same group of people use them, but their functioning and need fulfillment is completely different
#
aaronpk Like I can use Micropub and Microsub and never actually send a webmention or use Microformats
#
[eddie] Exactly. And people like KartikPrabhu support webmention and mf but don’t use Micropub/sub
KartikPrabhu and [jeremycherfas] joined the channel
#
[jeremycherfas] I wonder if there is a way to move away from IndieWeb as part of the label and towards something like “open”. So, maybe, “open friendly”.
iasai, KartikPrabhu and [kevinmarks] joined the channel
#
[kevinmarks] Everyone says open. Indieweb is more specific.
#
[kevinmarks] Facebook calls their rapidly shrinking api Open Graph.
#
[kevinmarks] Which is co-option of Brad's pitch
#
[kevinmarks] http://bradfitz.com/social-graph-problem/
#
Zegnat "more specific"? Surely anything following /principles is IndieWeb?
#
[kevinmarks] Right, whereas anything can be called "open"
cweiske and [jeremycherfas] joined the channel
#
[jeremycherfas] Yes, but there is also the fact that one can give a specific meaning to a word by using it consistently in a particular context. The fact that FB calls something Open Graph when it is not open, and gets away with it, shows that they understand this.
#
[jeremycherfas] So, OK, maybe no Open, but maybe not IndieWeb-something either?
jeremycherfas joined the channel
#
jeremycherfas In fact, rereading /principles maybe the thing to go for is Plurality. Pluraity-potent?
#
cweiske.de edited /comments (+49) "/* Comments Embeds */" (view diff)
#
jeremycherfas Apologies for my typing this morning
#
Zegnat Maybe, for plurality reasons, I should switch from mf2 to schema, ha
swentel joined the channel
#
swentel .
AngeloGladding, [pfefferle] and [kevinmarks] joined the channel
#
[kevinmarks] Does this work as an indieauth option? https://arstechnica.com/gadgets/2018/04/practical-passwordless-authentication-comes-a-step-closer-with-webauthn/
#
sknebel sure, you could build an endpoint that understands that
#
Zegnat I should really get a YubiKey at some point
#
Loqi I agree
#
Zegnat But yes. I think the endpoint could work with that. So with YubiKey (or equiv.) plugged in, you would enter your URL for login, it goes to your IndieAuth endpoint, your endpoint looks for the YubiKey and authenticates, if that succeeded the endpoint will instantly redirect you back to the origin website (e.g. the IndieWeb wiki) and you are logged in.
#
Zegnat No user action will be neccessary.
#
cweiske you still have to enter your domain name
#
Zegnat At the place where you want to login, e.g. the wiki, yes
#
Zegnat I don’t think Webauthn has active discovery we could use to take that away, surely?
#
jeremycherfas Woohoo. At long last, I now have a working PESOS of posts from reading.am. Not sure why the icon thing isn't working, but it will do for now.
#
Loqi 😃
#
Zegnat jeremycherfas++
#
Loqi jeremycherfas has 8 karma in this channel (21 overall)
#
Zegnat vHWC really needs to happen next week then, so you can demo! ;)
#
jeremycherfas Yeah, maybe.
#
jeremycherfas The problems now are definitely at the withknown end, because it strips various HTML entities and not others.
#
[kevinmarks] If you use webauthn does it provide any profile data?
#
[kevinmarks] Username?
#
cweiske I don't think so
#
cweiske I guess it's a token that identifies you
[jgmac1106] joined the channel
#
[jgmac1106] Sounds like folks are saying, "This is an Indieweb powered site or service by using MicroPub and Microsub protocols for publishing and/or microformats and webmentions for gathering shares, links , and comments from across the web. Still goes back to primary audience of the term devs or future users. No metaphor (compatible, powered, open graph) will be the perfect definition. If you want a perfect definition be literal. If you want to captu
#
[jgmac1106] mission speak in metaphors
#
[jgmac1106] I hate and/or stupid grammatical formation but trying to capture everything
#
jeremycherfas And/or is not only stupid, it is also meaningless. And yes, I agree that literal would be best, but that horse bolted long ago.
#
jeremycherfas To mix metaphors.
#
Zegnat I am not sure there is such a thing as “IndieWeb powered” at all, so I wouldn’t know where to start to define it either.
#
jeremycherfas I think the point is you don't need to define it. You just use it.
#
jeremycherfas It merans different things to different people, but there is a core there that is the essence of the thing
#
[jgmac1106] Exactly, Indieweb Compatible, IndieWeb Powered, Indieweb Plurality, the actual phrase chosen has very little to do with how it gets defined.
[kevinmarks] joined the channel
#
[kevinmarks] we have a badge
#
[kevinmarks] https://svgshare.com/s/5Hh
#
Loqi Indieweb svg badge https://svgshare.com/i/5Hh.svg
#
[jgmac1106] Maybe [kevinmarks] you want " abadgeable phrase" though I threw up in my mouth a little bit typing that
#
[kevinmarks] we also have indiemark and indiewebify.me
#
[jgmac1106] maybe you don't need a word just the badge
#
[kevinmarks] https://aaronparecki.com/2017/12/30/7/indieweb-badge
#
Loqi [Aaron Parecki] New Tiny IndieWeb Badge!
#
[jgmac1106] That is what I meant with a validator. Somebody does indiewebify.e and then they get a custom svg graphic to put on their site or service if it meets X conditions
#
jeremycherfas For me, the IndieWeb badge with a word like "powered" is a clear enough indication for Gen2-3 and 4
swentel joined the channel
#
[kevinmarks] i would say /indiemark is very Gen 1 focused
#
dgold I'm not sure that setting a 'validator' is significant
[xavierroy] joined the channel
#
dgold you don't _have_ to use mPub, mSub, mention &c
#
dgold you do you - if you are using a system that you control and you are happy with, and you are interfacing with the wider indieweb, then are you IW?
#
Zegnat I always like to just point at the /principles. If those fit, and it is a personal website, you are IndieWeb :)
#
dgold I'm regretting not embracing skippy's idea for plugins for nanopub, I think (on reflection) he's 100% on the right track
iasai joined the channel
#
[jgmac1106] @zegnat but for me the principles seem to be very similar to most open source or open web communities. I always thought of indieweb as the Open Web with adherence to very specific protocols, that is probably wrong, but thats the impression for those of us on the edges of the community
#
Zegnat Not to me. But different people will have different ideas of what the IndieWeb entails.
#
Zegnat There are definitely protocols that came out of IndieWeb, because people have been using those to communicate between sites (like webmentions), but that is not to say you are only IndieWeb if you use those specific protocols.
#
[jgmac1106] Then what is the difference between the indieweb community and the open web community? The principles seem to be the same. They have been around since I think 2010. Other open source communities do not rally around protocols the way indieweb does
#
Zegnat E.g. Post Type Discovery, that you can use to see if a document at a URL is a like (maybe after getting a webmention about it), isn’t bound to any specific markup. Doesn’t actually require the specific microformats that most people here are using c urrently.
#
[jgmac1106] I know when I talk about indieweb it is the POSSE model that resonates most with potential users. They can wrap their head around owning their data and choosing where to share it
#
Zegnat I don’t know other “open web” communities, so I can’t say what sets them apart from IndieWeb. IndieWeb is just a name for a group of people trying to get their websites to interconnect. But that is of course just my own personal view of it.
#
[jgmac1106] Even more telling if you mention indieweb to technical people in other open source communities they will say, "I love them. They are the microformats people, or negatively, they only want to push their solution"
#
[jgmac1106] just playing devil's advocate as someone who is a 1.5-2.0 user with limited technical skills
#
Zegnat I would love the people who say “they only want to push their solution” to come by and talk about their solutions. What we want is to be using the “best” solution. At least that is what I would want.
#
[jgmac1106] @zegnat didn't say it was right, just that it happens, and people who complain about solutions rarely have one of their own
#
[jgmac1106] I love the word indie but in the USA at least, it also brings the baggage (planned I am sure) of the business model from indie music of rejecting big publishers and trying to form your own label, or joining a small label, and doing it by yourself.
eli_oat joined the channel
#
[kevinmarks] The irony is that indie is the biggest genre now
#
[jgmac1106] yep, first time since 2011 physical media outsold digital mainly bc of the vinyl revival
#
[jgmac1106] there is a good metaphor there for an indieweb blog post
#
Zegnat “rejecting big publishers and trying to form your own label, or joining a small label, and doing it by yourself”, does sound IndieWeb
#
dgold [jgmac1106]: who/where are the OpenWeb people?
#
[jgmac1106] The music metaphor has always been strong. Right when indieweb was being introduced to the web when Kevin and Ben co-hosted TWIG to talk about Known. Reclaim Hosting and DO were really taking off. Reclaim used punk rock as a deliberate metaphor of indie in design aesthetic and philosophy
#
realize.be edited /Indigenous_for_Android (+163) "/* IndieWeb Usage */" (view diff)
#
[jgmac1106] [dgold] Open Web is more a loose term used in Open Source communities. These are often directly or peripherally connected to Mozilla. O'Reilly ran a conference going back to I think 2008 that used the term.
#
[jgmac1106] I think at one point MoFo board pushed "Open Web" as a top level goal and everything from around 2010-2013 (date range offish) had an open web mission style brand
#
[jgmac1106] @zegnat I always prefered Grateful Dead S my musical metaphor. Mainly personal preference and web history than anything. Barlow and manifesto, idea of tape trading as a model of give away product but cash in on service (concerts). But whether you are a vegan in a mosh pit or a vegan spinning in circles....principles Similar, but noise is different. Much like openweb and indieweb
#
[jgmac1106] I mean the first podcast Dave sent out on rss was Dead songs
#
[jgmac1106] Only problem with my metaphor is all monthly subscriptions must be sold "one for three and two for five"
[gerwitz] joined the channel
#
[gerwitz] it’s time for the Daily Dumb Question from [gerwitz]!
#
[gerwitz] The post type discovery algorithm on the wiki has a lot of “has property” tests.
#
[gerwitz] Are these generally understood to be deep searches? E.g. does a “in-reply-to” that’s nested in another item satisfy these heuristics?
#
Zegnat Parsed mf2 doesn’t have nested properties. So I am not sure I understand the question.
#
Zegnat Do you have an HTML example?
leg joined the channel
#
[gerwitz] no, you just exposed my ignorance of micropub json
#
[gerwitz] ah, mf2 maybe doesn’t, but a request from e.g. ownyourswarm might. The use cases I’ve found in the wild are simple: only h-cards with their own properties, but that means I have to decide whether to look deep
[kevinmarks] joined the channel
#
[kevinmarks] it can have nested formats - p author h-card in h-entry is common, and p-location h-card in h-event
#
[kevinmarks] if you look too deep it may be misleading - you may find cited comments for example
#
[gerwitz] it seems “unlikely but not impossible”
#
Zegnat Yes, but if you need to figure out the post type of an entry, the entry itself never has nested properties.
#
[gerwitz] well, that’s an answer then. “too deep” is “any depth at all” 😉
#
Zegnat If it does have nested properties, those aren’t the entry’s. They may be properties of nested comments instead, and must be ignored during PTD.
#
[gerwitz] well said, thanks
#
[gerwitz] (and yay for deleting code)
#
Loqi 😊
[xavierroy] joined the channel
#
[gerwitz] (I really ought to search for some aaronpk Ruby code to just copy)
#
jeremycherfas I've got an even dumber dumb question. I want to be able to share my little PHP thing for PESOS from reading.am to Known. But it needs the Known api key. What's the best way to leave it out of my working copy, but have it available when I want to run the program?
#
jeremycherfas I'm thinking, call it from the command line with an argument. Does that work?
#
jeremycherfas So far I have been calling it in the browser directly.
#
jeremycherfas Not sure how to pass a variable to the program from the browser
#
[gerwitz] environment variables would be the usual answer to this
#
Zegnat $_GET variables (?name=value at the end of the URL) work fine for passing from the browser, but those have the nasty side effect of ending up in your browser history.
#
[gerwitz] if I understand correctly, you don’t even need the browser to have the value?
#
[gerwitz] I don’t know if it’s PHP-ish, but this uses the technique I prefer in Ruby: https://github.com/vlucas/phpdotenv )
#
Loqi [vlucas] phpdotenv: Loads environment variables from `.env` to `getenv()`, `$_ENV` and `$_SERVER` automagically.
#
Zegnat That is PHP-ish enough. Frameworks like Laravel use it.
#
[gerwitz] be sure to add `.env` to your `.gitignore` if you’re using a shared git rep
#
aaronpk re: webauthn, indieauth.com (and the upcoming indielogin.com) won't use it because it requires pre-registration with the user. so you can use it with your own website, and you can use it with myindieauth.com since a user delegates their auth to that site so pre-registration is fine.
#
cweiske can't you add that token to your site, just as you add your gpg key?
#
jeremycherfas Thanks. So that would mean my thing would have a dependency. Not sure I am ready for that yet.
#
jeremycherfas Cweiske I have not had to add gpg to my Known site, so don't know how that would work.
#
aaronpk I don't think so cweiske but I might be wrong. I thought it was specific to the site consuming it. i've only glanced at the spec though
#
cweiske I didn't read it either
#
jeremycherfas Crossed lines. Apologies cweiske
#
aaronpk so far all i've been able to find is press coverage saying how great this will be, and the spec itself
#
Zegnat So it is a spec released without reference implementation, or any implementation at all?
#
aaronpk I mean chrome and firefox implemented it, I just can't find any tutorials or anything
#
aaronpk ok found a demo app https://github.com/google/webauthndemo
#
Loqi [google] webauthndemo: An example Java Relying Party implementation of the WebAuthn specification
#
aaronpk yeah I definitely need someone to write a tutorial version of this :)
#
@kaushalmodi ↩️ Dumb question.. but this needs one to have a Micropub endpoint running on their site right? I have set up my site only to receive Webmentions. (twitter.com/_/status/984067775680401409)
#
[gerwitz] In a micropub request, are duplicate property names legit?
#
aaronpk for things like category you might have category[]=one&category[]=two
#
[gerwitz] (a lot of the server code I’ve collected for inspiration uses key-value hashes that won’t allow it)
#
[gerwitz] ah, cool
#
aaronpk for form-encoded anyway
#
aaronpk in json everything is an array already
#
[kevinmarks] form-encoded should allow that without the [], but php and ruby are sad about it
#
aaronpk [kevinmarks]: micropub explicitly says [] are required
#
[gerwitz] barryf’s code is my leading inspiration, and makes everything an array Just In Case
#
aaronpk https://www.w3.org/TR/micropub/#form-encoded-and-multipart-requests
#
[kevinmarks] yep
#
[gerwitz] hmm, can I use `category[]` for a single value, if I’m a lazy client implementor?
#
aaronpk yea
#
aaronpk at least for the properties that support multiple values
#
aaronpk that's getting into vocab tho, which is defined in microformats
#
aaronpk e.g. h-entry doesn't support multiple values for "published" so servers will expect just one value for the "published" property
#
Zegnat well, h-entry the mf2 does support multiple values for published. Maybe jf2 does not?
#
Zegnat although you will be hard pressed for a micropub implementation that accepts multiple published times, probably :P
#
aaronpk Zegnat: does it really? I thought the description on http://microformats.org/wiki/h-entry meant only one value is supported
#
Loqi [Tantek Çelik] h-entry is a simple, open format for episodic or datestamped content on the web. h-entry is often used with content intended to be syndicated, e.g. blog posts. h-entry is one of several open microformat standards suitable for embedding data in HTML. ...
#
aaronpk "dt-published - when the entry was published" -> implies one value to me
#
aaronpk whereas "p-category - entry categories/tags" is plural
#
aaronpk and "p-photo - one or more photos"
#
aaronpk just because the json syntax allows it doesn't mean the vocab allows it
#
Zegnat But u-in-reply-to is also singular, while in the wild there are multiple examples of multiples
#
Zegnat But interesting, I never read those descriptions as being a canonical specifications on how many of them would be allowed!
#
aaronpk we have to define that somewhere, so if not h-entry then perhaps jf2 is the place to be more explicit about it
tantek_ joined the channel
#
tantek_ good morning #indieweb-dev
#
aaronpk darn it cweiske now you've got me trying to figure out of the public key from the webauthn registration can be put on my website and used by other relying parties
#
tantek.com created /validator (+28) "r" (view diff)
#
[kevinmarks] you would normally expect one dt-published and possibly multiple dt-updated
#
tantek.com created /indieweb_validator (+28) "r" (view diff)
#
[kevinmarks] though I have certainly written things that were published at different times in different places
#
tantek.com edited /Indiewebify.me (+26) "compatible see also friendly" (view diff)
#
tantek.com created /compatible (+22) "r" (view diff)
#
tantek.com created /compatible_with_indieweb (+22) "r" (view diff)
#
tantek.com created /indieweb_compatible (+22) "r" (view diff)
#
tantek.com edited /principles (+15) "/* See Also */ friendly" (view diff)
#
tantek.com created /powered (+22) "r" (view diff)
#
tantek.com created /indieweb_powered (+22) "r" (view diff)
#
tantek_ aaronpk, all microformats properties are syntactically plural
#
aaronpk syntactically yes, but I thought the vocabularies made distinctions
#
tantek_ mf2 in particular that is. if/when a vocabulary says anything about "a" or "the" value of a property implying single-value, it means it's only defining a semantic for the first value
#
tantek_ nicely allows future expansion
#
[kevinmarks] I'd read 'the' as first value and 'a' as any value.
#
tantek.com edited /friendly (+105) "re-organize a bit, friendly, in-progress, brainstorming, previously, micro.blog is friendly" (view diff)
[jgmac1106] joined the channel
#
Zegnat I like that distinction, [kevinmarks]! Though I can totally see why people coming to read about microformats in the first place will not automatically do the same. I am with aaronpk that this could use some explicit note somewhere easily discoverable.
#
[kevinmarks] remember that the vocabularies are summaries of what we have found in the wild - you can mark up multiple of any property, but the page should be based on what we see
#
[kevinmarks] so the h-book conversation this morning implied to me that we may find author as an important field on h-product
#
[kevinmarks] but if you do make an h-book, use the same property names for isbn etc
#
Zegnat https://zegnat.net/tmp/bookshelf.html has h-book, but as the URL suggests, is a temporary file and the HTML is volatile.
[eddie] joined the channel
#
[eddie] Regarding all the discussion above regarding the “IndieWeb compatibility” thing, I think got confused. The discussion was never about if a website meets certain baselines. It’s how to describe the ability to a Gen 2/3/4 person that a specific web software or service supports the use of Micropub/Microsub. Not All IndieWeb abilities.
#
[eddie] I completely agree that there are too many IndieWeb type capabilities none of which are required
#
[eddie] But if a new user say downloads Indigenous off the App Store and they want to know if they can use it with their service or what new services to sign up for to be able to use an IndieWeb reader, how do they know that?
#
jeremycherfas.net edited /friendly (+75) "/* Pandoc */ status of link" (view diff)
#
[eddie] For XML-RPC, in the wild, we’ve found that it has used Wordpress compatibility as the easy to understand thing over protocol language. In podcast feeds, it’s a generic statement and MP3 players uses the actual protocol name
#
skippy but "WordPress compatible" is a coopting of the original term "MetaWeblog", which WP implemented to be compatible.
#
skippy pedantic, sorry. not strictly relevant to your point [eddie]
#
skippy IndieWeb is an umbrella term that ultimately signifies the preferences and priorities of the individual user. Some users want to create new content; some want to dynamically interact; some want to consume only. Indieweb focuses on what the user wants, not what a specific service provider wants.
#
jeremycherfas So, if that's what you want (which is maybe why you were interested in Indigenous) then IndieWeb-friendly fits the bill, no?
#
skippy http://s2.quickmeme.com/img/e5/e5995c0636a3ebbd79539d491d044426a62ad1d5a298cefce4a035789385e99c.jpg
#
@cleverdevil ↩️ @benwerd FWIW, I've already been able to make a serverless IndieAuth Token and Auth endpoint work on Lambda... https://cleverdevil.io/2018/going-serverless-with-python-wsgi-apps (twitter.com/_/status/984088714782609408)
#
jeremycherfas Users always sounds to me like some kind of drug-addict -- which in some ways may be appropriate.
#
[eddie] Skippy that’s a good point. But people use the term Wordpress over MetaWeblog because that is a more understandable and relatable word
#
[eddie] The issue we ran into was when I wrote up some onboarding language for Indigenous to help new users sign in or get registered for a compatible service, having Micropub on one screen and Microsub on another is not incredibly clear or friendly. So the idea was finding a more clear/friendly term that could denote services that could work with a Reader client. I would shy away from IndieWeb friendly because that can denote anything (not clear), it c
#
[eddie] services that support Webmention, mf2 or any other IndieWeb tech
#
[eddie] * it covers sites and services that support Webmention, mf2, or any other IndieWeb tech
klandwehr joined the channel
#
schmarty [eddie]: similar to WordPress vs MetaWeblog, I don't think there is a "right" term that folks will understand unless it relates directly to their website and the software or service provider that runs it.
#
[eddie] That’s true, schmarty
#
schmarty e.g. "Indigenous can post to your website if it is powered by Known, a hosted Micro.blog, WordPress with the Micropub plugin, or if you know that your site supports Micropub"
#
[eddie] schmarty++
#
Loqi schmarty has 12 karma in this channel (76 overall)
#
[eddie] That’s a great direction!
#
schmarty "You can use Indigenous as a feed reader if you have an account on Aperture account on your own or someone else's server."
#
jeremycherfas Yup, that first one is good, schmarty
#
[eddie] Ahh that actually helps because I’m also going to support micro.blog’s api for reading. So I’m able to say a micro.blog or Aperture account. Which prevents the issue of making it sound like micro.blog supports Microsub
#
[eddie] As more public Microsub servers go up I can update the language to include them, because the app itself will likely never push someone to make a Microsub server. (Too technical)
#
schmarty it's high-touch to keep that kind of thing up-to-date, but I expect Indigenous will continue to change in the near term as the available software and services evolve. :}
#
schmarty haha, thanks. i've been thinking about this since the discussion last night. while I agree that it doesn't make sense for later generation users to understand Micropub vs Microsub, there is a concrete list of software that support those specs.
#
schmarty and nobody's website runs on nothing :}
#
schmarty [eddie]: it might even make sense to say "Micro.blog or Aperture account, or an account on any Microsub-enabled service"
#
schmarty [eddie]: yep that makes sense
#
[eddie] Yeah that seems good. And like you said, eventually I’ll leave the list as it with the top 2 or 3. Likely the end goal would be to list hosted services and then have “or any Microx-enabled service”
#
[eddie] Because if someone knows enough to host something themselves Micropub or Microsub enabled service should provide enough for them to find software to install. But if someone doesn’t know how to install software the hosted services are really their best options, hence displaying those are options.
#
[eddie] 👏 break-through!
gRegorLove joined the channel
#
schmarty yep, i believe that will all evolve. trying to settle on good Gen3/4 naming now feels like a placing a bet on a roulette table that is still being assembled. ;}
#
aaronpk "Microx" 😂
#
skippy alternately, evolving an agreed-upon vocabulary now makes it harder for one org to coopt it later. See also: WordPress.
#
[eddie] Haha, can you tell I got tired of typing the 16+ characters around “Micropub or Microsub”?
#
[eddie] I was like, forget it, I’m just gonna put an X to mark the spot
#
schmarty that said, Indigenous is really pushing the envelope on usable IndieWeb experiences and i love how much thought and care you are putting into it!
#
schmarty eddie++
#
Loqi eddie has 29 karma in this channel (53 overall)
#
schmarty the terminology is evolving in realtime :D
#
skippy maybe shift the thiniking from protocols to actions, or intents. We already talk about "syndication". I should think some beginnger friendly words could be found to adequately encapsulate "creating content", "sharing content", "engaging in discussions" etc
#
[eddie] Yeah I definitely think that can help things, skippy!
#
Zegnat Shorthands: Micro[ps]ub, M6b, …
#
aaronpk Mi5b
#
skippy what's "micro" about micropub or microsub?
#
[eddie] Thanks schmarty! I figure if it is gonna sit in the App Store and be potentially downloaded by anyone, we wanna make sure it’s one of the friendliest IndieWeb apps available
#
schmarty M[ps]b
#
[eddie] 😆
#
aaronpk skippy: 1) they are meant to be minimal protocols and easy to understand, and 2) the vocabularies and syntaxes are based on microformats
#
skippy thanks. that makes sense to me. but as more people want to dip their toes in, they wont want to care about microformats.
#
skippy so the continued use of the micro prefix seems a sticking point, in some ways.
#
aaronpk why?
#
aaronpk "micro.blog" does not call itself "micro" because of microformats
#
jeremycherfas Because it doesn't tell you anything unless you already know.
#
aaronpk it's the name of a spec, I don't really see that as a problem
#
aaronpk "micropub" a micro publishing spec
#
skippy no, the name of the spec is not a problem. i'm just observing that if that nomenclature carries forward to more user-facing elements with less-technical users, it will be confusing. What IS micro about Indigenous, for example?
#
aaronpk "microsub" a micro subscribing spec
#
aaronpk I thought we were all in agreement that the spec names are not intended to be user-facing terms
#
schmarty i think the issue here is that there are user-facing tools like Indigenous which user-facing terms that don't exist yet.
#
schmarty *which need
[kevinmarks] joined the channel
#
[kevinmarks] µP, µS, µF
#
Zegnat Has any name ever survived using the µ sign?
#
[kevinmarks] not µTP, they switched to uTP
[jgmac1106] joined the channel
#
[kevinmarks] I was being facetious, those are SI units too
#
schmarty had not heard of the "poise" unit of viscosity!
#
schmarty aww, it's non-SI :/
sebsel joined the channel
#
jeremycherfas What is gotalk?
#
Loqi It looks like we don't have a page for "gotalk" yet. Would you like to create it? (Or just say "gotalk is ____", a sentence describing the term)
eli_oat and snarfed joined the channel
#
dgold um... how does one install phpunit, again? (asking for a friend)?
#
aaronpk i usually install it via composer
#
aaronpk in the project
#
aaronpk then run ./vendor/bin/phpunit
#
skippy https://phpunit.de/getting-started/phpunit-7.html
#
dgold aaronpk: its a compass install, and your README saysYou'll need to make sure the following PHP extensions are installed. Typically these are installed using the package manager of your operating system.
#
aaronpk um, that's weird
#
aaronpk that shouldn't be listed ther
#
aaronpk i don't even think there are any tests in compass haha
#
aaronpk yeah just ignore that
#
dgold lol!
#
dgold thank you, I was, i mean _my friend_ was going mad
swentel joined the channel
#
benharr.is edited /IRC_People (+100) "add benharri" (view diff)
benharri and KartikPrabhu joined the channel
#
skippy Zegnat: not sure if you saw this last night: https://github.com/J7mbo/twitter-api-php a single-script PHP library, something you asked about a while ago.
#
Loqi [J7mbo] twitter-api-php: The simplest PHP Wrapper for Twitter API v1.1 calls
#
Zegnat snarfed, when I try to ID to Bridgy with https://vanderven.se/martijn/ it correctly sends me to my endpoint.
#
Loqi Martijn van der Ven
#
Zegnat But when I authenticate, getting back to Bridgy it gives me a 400 error:
#
Zegnat HTTP Error 400: 400 Bad Request The server could not comply with the request since it is either malformed or otherwise incorrect. Verification response missing required "me" field
#
aaronpk "verification response" sounds like that's the response from the HTTP request bridgy made?
#
Zegnat skippy, I saw it, but haven’t had time to check it out yet! Thanks for digging it up though!
#
snarfed thx Zegnat. disappointing, this worked recently.
#
snarfed will investigate
#
snarfed i suspect rate limiting :(
#
snarfed (ah, the error on my attempt is different)
#
Zegnat Hmm, that would surprise me aaronpk. selfauth has been used in many different places without issue. But of course it could be an error on my part, so let me know if I can help testing in any way, snarfed!
#
snarfed ahh Zegnat you're not using indieauth. ok
#
aaronpk https://github.com/snarfed/oauth-dropins/blob/master/oauth_dropins/indieauth.py#L173
#
aaronpk oops ignore me
#
snarfed i swear i'm ready to drop instagram entirely
#
aaronpk yeah https://github.com/snarfed/oauth-dropins/blob/master/oauth_dropins/indieauth.py#L171 looks like the auth endpoint didn't return the "me" parameter?
#
snarfed scraping--
#
Loqi scraping has -4 karma in this channel (-7 overall)
#
aaronpk oh noooo snarfed
#
snarfed rate limiting means signup now sometimes (often) fails
#
aaronpk wow
#
snarfed scraping--
#
snarfed scraping--
#
aaronpk that would mean i'd have to add backfeed to ownyourgram
#
snarfed scraping--
#
aaronpk haha speaking of rate limiting
#
snarfed aaronpk: and deal with the exact same problems? :P
#
snarfed filed https://github.com/snarfed/bridgy/issues/809
#
Loqi [snarfed] #809 instagram signup is buggy
iasai joined the channel
#
swentel .
singpolyma and iasai joined the channel
#
snarfed i guess i've never really done a deep dive on our auth stuff. ok so...
#
snarfed i've skimmed http://microformats.org/wiki/RelMeAuth, https://indieweb.org/RelMeAuth, https://indieauth.com/faq, https://indieweb.org/IndieAuth
#
Loqi RelMeAuth is a proposed open standard for using rel-me links to profiles on OAuth supporting services to authenticate via either those profiles or your own site. RelMeAuth is the technology behind web-sign-in. Editor Tantek Çelik (http://tant...
#
snarfed iiuc, indieauth extends relmeauth to add oauth style code/token minting, verification, and scopes. is that right?
#
snarfed i also haven't yet found formal specs for either indieauth or relmeauth. just the pages above. so it seems like aaronpk and Zegnat disagree (maybe?) on whether the indieauth code verification request is supposed to return a me param, but i don't know where to go to determine who's right
#
Zegnat snarfed: https://indieauth.spec.indieweb.org/
#
Loqi [Aaron Parecki] IndieAuth
#
snarfed aha. guess i should have just googled 'indieauth spec' instead of looking for links
#
snarfed thx
#
aaronpk taht *should* be linked from a bunch of pages
#
Zegnat https://indieweb.org/IndieAuth-spec on the wiki, except that is split up in several pages. The actual spec sub domain is better.
#
Zegnat After all the bickering on minute details of the spec, I don’t *think* aaronpk and I disagree on anything anymore :P
#
snarfed lol
#
aaronpk yeah i think this is just a weird misunderstanding of what's going on
#
aaronpk the GET response should not return the `me`, the POST response should
#
Zegnat me should not be in the redirect, me should be in the verification response. If my verification response is wrong I should fix that.
#
snarfed oh i know. i'm looking at the code verification POST response
#
snarfed maybe zegnat's returned an error instead, and i ignored it
#
snarfed Zegnat: mind trying bridgy again? i added logging
#
Zegnat Done
#
Zegnat Also opening up my endpoint code
#
snarfed ahhh i'm blindly expecting form-encoded, and ignoring content type, while zegnat and the spec say it's json
#
Zegnat I actually return form-encoded if that is what you are requesting
#
Zegnat And it should be defaulting to form-encoded
#
Zegnat What am I returning, did you log it?
#
@Inkwater_Masha Publishing is a choice. Here's an interesting perspective of the challenges and reality out there. http://ow.ly/23Qt30jjmqs #publishing #self-publish #traditional #amwriting #indieauth #author (twitter.com/_/status/984131498470232064)
#
snarfed Zegnat: json. {"me":"https:\/\/vanderven.se\/martijn\/"}
#
snarfed i don't send Accept or do conneg
#
Zegnat That’s ... odd :/
#
snarfed i expect you just default to json
#
Zegnat No: https://gist.github.com/Zegnat/4ad87603bcabbf8e095363df99845e50
#
Zegnat Default to form-encoded, only send JSON when the q-value for JSON is bigger than for form-encoded
#
Zegnat Is your HTTP library putting some sort of values in the Accept header without you specifying?
#
snarfed huh ok
#
Zegnat If Accept was empty, I would have expected you got a 406 back from me.
#
aaronpk wow that's pretty thorough conneg support Zegnat! I always cheat
#
Zegnat I wrote this for selfauth
#
Zegnat I wrote get_q_value() too. Selfauth is CC0, so feel free to take it if you need it, aaronpk.
#
Zegnat hacks in some Accept header logging and hopes not to break his endpoint
#
snarfed i haven't yet found anything in my code path or libs that should be adding Accept
#
snarfed thanks for logging!
#
Zegnat You are sending `Accept: */*`
#
Zegnat So my endpoint assumes you also know how to handle JSON in that case, and gives you JSON
#
Zegnat (That is, you accept JSON as much as you accept form-encoded, and JSON wins by spec.)
#
Zegnat Very dangerous header to send. The kind of header that makes rhiaro return JSON when you accept HTML content ;)
#
KartikPrabhu snarfed: it might be the default accept header
#
snarfed KartikPrabhu: yeah but default *where*
#
snarfed i wonder if app engine adds it
#
snarfed odd
#
KartikPrabhu snarfed: are you using the requests lib?
#
snarfed yup
#
KartikPrabhu is looking through its documentation
#
aaronpk i've definitely written code that checks the first character of the response and if it's { then i json_decode otherwise form-decode
#
snarfed sigh
#
Zegnat That’s definitely also a good way to do it, aaronpk.
#
aaronpk "good"
#
KartikPrabhu snarfed: from this example http://docs.python-requests.org/en/master/user/advanced/#request-and-response-objects it seems */* is default request Accept header
#
Zegnat For IndieAuth, I would recommend something like `Accept: application/json,*/*;q=0.9`. That says you would accept whatever the server feels like responding with, but only if the server couldn’t specifically respond with JSON.
#
Zegnat Keeping the */* is good for servers that do not really do conneg
#
snarfed if Accept: */* is dangerous, lots of people are living dangerously. curl and wget send it by default
#
Zegnat Completely right though for those tools, snarfed. Because they want to download whatever is at the URL.
#
Zegnat And they accept everything, as they simply grab the stream and do something with it.
#
snarfed fair
#
snarfed and i have special cases for rhiaro in multiple projects 😅
#
snarfed https://github.com/snarfed/bridgy/blob/master/util.py#L50 , http://www.indiemap.org/docs.html#rhiaro
#
Zegnat In reality, most of your projects should probably just state in the accept header that you prefer to get HTML back
#
Zegnat Known breaks on it?!
#
KartikPrabhu snarfed: http://docs.python-requests.org/en/latest/user/quickstart/#custom-headers for custom headers
#
snarfed i have maybe a dozen projects that total hundreds to thousands of different http request code paths, with a variety of libraries, stacks, and semantics. honestly i'm not going to go through and figure out conneg for all of them
#
KartikPrabhu ha!
#
aaronpk that's why i just check the actual content
#
Zegnat Fair snarfed.
#
aaronpk and parse json if it starts with {
#
aaronpk #poormansconneg
#
Zegnat You would expect JSON or form-encoded back from IndieAuth, so content sniffing is fine
#
KartikPrabhu aaronpk: shouldn't the response contain infor about what it is in some Response header?
#
aaronpk you'd think that would always be accurate wouldn't you
#
KartikPrabhu <sigh>
#
aaronpk ;-)
#
swentel hehe
#
Zegnat That’s like thinking the Accept header on the request side is always going to be accurate ;)
#
Zegnat Funny enough I think some projects might not be sending Accept headers at all. Because I know for a fact defaulting to form-encoded has helped me log in to a few places. And it would only default if no header was sent.
#
Zegnat Should maybe special case it for */* only...
#
snarfed conneg--
#
Loqi conneg has -1 karma in this channel (1 overall)
#
singpolyma conneg++
#
Loqi conneg has 0 karma in this channel (2 overall)
#
singpolyma ;)
#
snarfed i understand why it exists, but for the vast majority of use cases, it's overkill and harmful
#
snarfed heh
#
Zegnat It sometimes feels like it modifies the meaning behind a URL a bit. Because you actually get multiple different files at the same URL, switching on MIME type
#
KartikPrabhu yup
#
aaronparecki.com edited /Micropub-extensions (+329) "/* Query */" (view diff)
#
jalcine agh missed that convo about `Accept:` and `curl` :(
#
aaronpk you never really miss conversations in IRC
#
jalcine ha truly but I don't wanna be that 10-day post reviver
#
Zegnat Surely it was the last subject discussed? That’s especially fine to continue on :)
barpthewire and cweiske joined the channel
#
cweiske.de edited /search (+92) "/* Software */ tntsearch" (view diff)
iasai_, [kevinmarks], KartikPrabhu, [jgmac1106] and snarfed joined the channel
#
dgold aaronpk: moving compass data from one instance to another -- is that a complex task?
#
aaronpk hm, no should be simple
#
aaronpk move the storage folder then make sure there's an entry in the mysql database with the same name
#
dgold excellent
#
aaronpk in theory :) i haven't actually tried this before
#
dgold my hyper-cheap vps has turned out to be worth pretty much what I paid for it
#
aaronpk haha
#
dgold hmmm, someone's asking for me to detail installation of nanopub in greater detail, I may have to do that
#
aaronpk docs++
#
Loqi docs has 1 karma
#
aaronpk documentation++
#
Loqi documentation has 3 karma in this channel (11 overall)
snarfed and [eddie] joined the channel
#
dgold how do I link to another document in github docs?
#
aaronpk in the same repo?
#
aaronpk you can use markdown link syntax if you want to make a link to wherever
#
aaronpk i think there's a wiki-like syntax for linking to files within the same repo
#
aaronpk oh just a relative link works, like [read this too](READTHISTOO.md) works from README.md
singpolyma, dgold and kaushalmodi joined the channel
#
kaushalmodi dgold: Yes, that's me! (asking for nanopub installation instructions for PHP-newb)
#
dgold hey kaushal! THat issue is just about to get a [WIP] flag!
plindner joined the channel
#
dgold i'm afraid you will need composer installed
#
kaushalmodi oh
#
kaushalmodi about document linking: https://help.github.com/articles/about-readmes/#relative-links-and-image-paths-in-readme-files
#
kaushalmodi I hope there isn't any need to know/learn PHP to get nanopub to work.. I can deal with setting variables in config.php.. (will wait for the updated instructions). Thanks!
#
aaronpk dgold: you might consider zipping up a complete install of it and publishing it in the "releases" section on github
#
kaushalmodi dgold++
#
Loqi dgold has 17 karma in this channel (42 overall)
#
aaronpk with the vendor folder
#
aaronpk i was doing that with XRay, tho I haven't done it in a month or two since i was doing a release every week for a while
#
dgold I will do that, but I'm about to eat, unfortunately
#
kaushalmodi If what aaronpk says is possible, awesome! I can deal with "cp this to your static/ dir, set this in config, and you are set" :)
iasai_, snarfed, iasai, [kevinmarks] and eli_oat joined the channel
#
snarfed hah, seeing a new error in bridgy's outgoing webmentions: "ERROR: please fill the required fields (GDPR checkbox)."
#
snarfed looks like a wordpress site. probably a plugin for commenters to consent to GDPR conflicting with the webmention plugin.
#
snarfed cc GWG pfefferle
#
snarfed example: https://brid.gy/log?start_time=1523486573&key=aglzfmJyaWQtZ3lyNAsSCFJlc3BvbnNlIiZ0YWc6ZmFjZWJvb2suY29tLDIwMTM6MjAyOTAwMzYyNzM3MjkzNAw
#
aaronpk wow
klandwehr joined the channel