2017-06-17 UTC
kline joined the channel
# 01:07 tantek !tell Zegnat I like this new "get it done by Friday" policy for HWC pages / venues. Anything posted by Friday 17:00 Pacific Time I'll add to the Mozilla announcement for the next Monday, but otherwise no promises. In addition I think /next-hwc should always point to the next "regular" Wednesday evening HWC. Like usual fortnightly or a reliable regular city (e.g. Portland). # 01:07 Loqi Ok, I'll tell them that when I see them next eli_oat joined the channel
# 01:56 eli_oat oh, it is u-like-of, isn't it, not just like-of?! # 01:58 tantek for the mf2 property: microformats.org/wiki/h-entry # 01:59 eli_oat hmmm, I just double checked and I seem to be applying the right class in my markup # 01:59 eli_oat I'm trying to determine why my posts don't always appear in the woodwind # 02:00 eli_oat Would including an h-card within the h-entry throw things off? # 02:08 eli_oat hmmm, intriguing. I will continue to dig in. Thanks tantek # 02:08 tantek I just checked that pin13 link eli_oat and it seems fine - what's "not working" ? # 02:09 eli_oat just that all of my title-less posts come up blank on woodwind # 02:10 tantek however even if that is true, it highlights a need for fallbacks # 02:10 tantek which we have discussed in part in the past, just never fully formalized # 02:13 tantek e.g. I show and markup a plain text fallback as "summary # 02:14 eli_oat I wonder if that is how your posts are able to circumvent micro.blog's character limit? # 02:14 Loqi [Tantek Çelik] likes Aaron Parecki's post # 02:15 tantek interesting, had no idea I was breaking any micro.blog limits :) # 02:16 eli_oat (no way to link directly to it on micro.blog, it seems) # 02:17 tantek I think their micro.pub endpoint also has no length limit # 02:18 eli_oat yeah, I've been unable to get my endpoint to play nice with micro.blog # 02:18 eli_oat Or, I guess, I have been unable to get my endpoint to play nicely with their client would be more accurate # 02:19 eli_oat but my endpoint works on a hopes and dreams more than anything else, because I have no real clue what I'm doing there :P # 02:26 eli_oat I think you are right about their endpoing not limiting characters. I've been scrolling through manton's timeline, and it seems like the truncation is happening as a secondary process. tantek, ben_thatmustbeme, deathrow1, barpthewire, AngeloGladding, j12t, sebsel, KevinMarks, [miklb] and KevinMarks_ joined the channel
KevinMarks joined the channel
# 13:07 Loqi Ok, I'll tell them that when I see them next j12t joined the channel
[cleverdevil], [miklb] and KevinMarks joined the channel
j12t joined the channel
KevinMarks, KartikPrabhu, j12t, KevinMarks_ and ejd joined the channel
# 16:29 Zegnat KevinMarks, you were there for the first Virtual HWC right? On Hangouts? Could you add a point about Hangouts to /Virtual_HWC ? # 16:29 Loqi Zegnat: tantek left you a message 15 hours, 21 minutes ago: I like this new "get it done by Friday" policy for HWC pages / venues. Anything posted by Friday 17:00 Pacific Time I'll add to the Mozilla announcement for the next Monday, but otherwise no promises. In addition I think /next-hwc should always point to the next "regular" Wednesday evening HWC. Like usual fortnightly or a reliable regular city (e.g. Portland). [kevinmarks] and j12t joined the channel
KevinMarks joined the channel
# 17:37 GWG I have a mf2 parsing issue I need brainstorming help on. # 17:39 GWG Many WordPress themes include mf1 entry-content or entry-summary around the post content block. Many third party plugins add to the content block dynamically, as it is the only way to dynamically add things like metadata, ads, etc to WordPress posts. # 17:41 GWG This is the only way I can add things to posts as well, which results in properties nested inside e-content, such as reply-contexts. I have modified themes to work in such a way that this isn't an issue, but I can't modify every theme. I'm trying to think of an alternative solution. [miklb] joined the channel
# 17:46 GWG [miklb]: Want to weigh in on this? # 17:47 sknebel GWG: so you can not influence the outer markup, since it is part of the theme, correct? # 17:48 GWG In my theme, I can address that. But not in 100% of themes. # 17:48 [miklb] I haven’t given it much thought, but should considering what I’m working on. # 17:49 [miklb] but ultimately, I’m not sure ever going to fully solve it. # 17:49 GWG I wonder if this is a microformats parsing question. # 17:50 sknebel since value-class-pattern doesn't apply for e- properties, I have no idea how you'd hide stuff inside e-content # 17:50 GWG If you have an h-cite nested inside e-content by necessity... # 17:50 Zegnat No reason you can’t have .u-in-reply-to within .e-content # 17:50 GWG I have to put a full reply-context in there. # 17:51 GWG It parses correctly, but the parsed content html and plaintext also includes it. # 17:51 Zegnat Yes, people who parse your site and want to display your comment might have issues. That’s not really inherit to mf2 parsing # 17:51 GWG I meant, is there a way to mark it up that it wouldn't be an issue? # 17:51 GWG Without moving the in-reply-to outside of the e-content? # 17:52 Zegnat You can hide it, but then people reading the wordpress blog might not understand it to be a reply # 17:52 Zegnat <data class="u-in-reply-to" value="URL">My reply right here!</data> # 17:52 GWG I'm talking about hiding it only from the parser, not visually. # 17:53 GWG The only way I can think of is something in the markup that tells the parser to ignore mf1 and only parse mf2. # 17:53 GWG Then it would ignore the entry-content in favor of the e-content. # 17:55 Zegnat To be fair, if plugins can only put stuff inside entry-content I imagine a lot of plugins exist that add stuff that isn’t supposed to be content either. Which then means it is a bug on the themes to have entry-content in that location in the first place :/ But I see how that would be hard to fix. # 17:56 sknebel if it is mf2 parsing (as in, the root element parsed is an h-entry) entry-content should be ignored. if it is mf1 parsing, then e-content doesn't do anything # 17:57 sknebel but I guess the issue is that all templates have "hentry" and you can't force it to add "h-entry"? # 17:59 GWG This is going to be a significant usability issue in the long-term for onboarding new users. # 17:59 GWG sknebel: You can, because that property is filterable. # 17:59 sknebel if it has "h-entry" and i'm reading the parsing algo right, entry-content should be totally ignored # 18:02 Zegnat On an element with .h-entry.hentry (both), the .hentry must be ignored and “backcompat properties” are ignored as well # 18:02 Zegnat So, yes, GWG, adding h-entry next to the existing hentry (which you probably want to keep for possible CSS reasons) should do exactly what you want and ignore entry-content # 18:02 GWG Time for a new wordpress-uf2 then # 18:03 GWG [miklb]: No. But pfefferle isn't actively maintaining it. # 18:04 GWG He authorized me as a secondary maintainer. It's years old and could use a serious refresh. # 18:05 [miklb] either way, wondering if working on that would give more bang for the buck before IWS # 18:05 Zegnat GWG, my example shows separate e-content and u-in-reply-to packed within entry-content # 18:07 Zegnat PHP, Go, Python, and Ruby, all the parsers agree on the interpretation! (sknebel, good news, no bugs there :p) # 18:07 sknebel oh, apropos, maybe the comparison site we talked aobut would be a nice IWS hackday project? # 18:09 GWG pin13.net/mf2 when I try that shows correctly, but Semantic Linkbacks doesn't...even though they both use php-mf2 # 18:11 Zegnat That could be a fun hackday project sknebel! Something you would want to work on together? eli_oat joined the channel
# 18:12 Zegnat Wasn’t there an issue filed somewhere for doing comparisons? j12t joined the channel
# 18:16 GWG And that compare site showed all parsers getting it right. # 18:16 GWG Except it still isn't parsing it correctly. # 18:18 sknebel doesn't it see the content attribute and falls back to name? # 18:20 GWG I'm looking at the output in the debugger I installed to output the parsed version to the screen # 18:20 sknebel "isset( $node[ $key ][0])" - is that true for the "content":[{"html":..., "value":...}] ? # 18:22 GWG But the output from php-mf2 is coming in differently. # 18:25 GWG If I turn off support for parsing classic microformats markup, it renders correctly. # 18:27 [miklb] I thought the parser should do that automatically # 18:27 Loqi [gRegorLove] #111 Improve backcompat parsing # 18:30 GWG I have to have it ignore classic microformats to parse correctly. KevinMarks joined the channel
# 18:31 [miklb] oh, I thought you meant it was an option in Semantic Linkbacks # 18:33 sknebel the parser code looks right, as in having that only as a switch for backcompat # 18:35 Zegnat can’t really read much more code, jumping around on phone is not working very well :p # 18:36 GWG I've introduced a temporary situation...a stock WordPress theme, and wrapping the appropriate information in e-content inside entry-content. # 18:39 sknebel and php.microformats.io *also* has backcompat enabled # 18:40 [miklb] and it parsed that scenario correctly, e-content inside entry-content inside wrapper with h-entry & hentry right? # 18:40 Zegnat if php.microformats.io has it enabled and runs the same version as SemanticLinkback then it must be the HTML mark-up... # 18:40 sknebel GWG, can you post the HTML as a gist or something? # 18:42 Zegnat Or put the entire HTML in php.microformats.io # 18:42 GWG This is the page I used. I changed the theme to a stock one and enabled wordpress-uf2 to auto-add e-content. # 18:43 GWG pin13.net parses content correctly. SL doesn't. Both using the same parser. # 18:43 Zegnat I was just going to say, that parses correctly on php.microformats.io … really odd # 18:44 GWG But it has inspired me to update wordpress-uf2. # 18:44 GWG Zegnat: If I toggle the classic to false, it does it correctly # 18:44 Zegnat You want u-like-of there btw, not p-like-of as it has now # 18:44 [miklb] GWG I just forked it myself. if nothing else, will help me better understand for my theme. Spinning up a test site for it now # 18:45 Zegnat But that prefix change doesn’t fix your issue of course # 18:45 GWG Zegnat: You can like things without a URL. But that's another issue. # 18:45 GWG I will make a note of that. I can add the logic to swap between them. # 18:47 Zegnat Need to finish the dishes, then I’ll have another look at the parser, brb tommorris, KartikPrabhu and gRegorLove joined the channel
# 18:57 sknebel good morning gRegorLove. For some reason, semantic-linkback (WP plugin) uses mf1 when available, but on php.microformats.io it parses correctly (preferring the also-existing mf2, as it should) # 18:58 gRegorLove Ah. Which particular fork / commit on github are we looking at for semantic linkbacks? # 18:59 GWG gRegorLove: I'm looking at the Github master version # 18:59 GWG Same thing happened on stable though. # 19:03 GWG I modded it to embed e-content inside entry-content to see how classic backcompat worked. # 19:07 GWG gRegorLove: It was parsing entry-content instead of e-content. # 19:07 GWG Giving the whole context as a result. # 19:10 GWG I just started working on updates to wordpress-uf2. # 19:11 GWG I've decided to move that up ahead of other things. Might be able to write it more effectively. It's been idle for 2 years. # 19:12 GWG [miklb]: Watch the dev branch in the current repo # 19:13 [miklb] yeah, I forked from pfefferle’s repo so I’ll keep an eye out. # 19:14 GWG I have the ability to add branches, but merging to the master branch requires review # 19:14 GWG I thought he gave me that permission. # 19:14 GWG But I'll worry about that when I have something worth merging. # 19:15 GWG I just gave it a proper README, language file, etc. The same structure we use for all the plugins. I've matched him on it for a while now. # 19:16 GWG gRegorLove: Yes. I built a debugger function that echos it to the screen to check URLs without having to actually mention something j12t joined the channel
# 19:17 gRegorLove Only thing I can think is that Mf2\Parser is included elsewhere before this plugin's files load, so the check at the beginning fails to include the php-mf2 included with SL # 19:19 GWG That is possible. Post Kinds also uses it. I can update that version and check. # 19:19 GWG Wordpress is horrible with dependency management # 19:21 sknebel so my horribly hacked carcass of that file finds the right content # 19:22 sknebel to check if it uses the right parser you temporary could rename the class to something else, to make sure you get the right one? # 19:25 [miklb] how would someone new to IndieWeb know to look for a plugin that adds “uf2”. Just thinking through the name # 19:26 [miklb] also realizes it can eventually be added to the IW plugin loader, but still. # 19:26 sknebel gRegorLove: version-check method sounds like a good idea for these situations # 19:29 sknebel and really, letters in plugin-names aren't expensive # 19:29 sknebel "wordpress-plugin-that-injects-mf2-into-pages" would be easier to understand :P # 19:29 Loqi [pfefferle] wordpress-uf2: add microformats2 classes to your WordPress theme # 19:30 [miklb] wordpress-mf2-theme-helper would be the direction I would go j12t joined the channel
# 19:37 [miklb] ok, well I opened an issue to discuss renaming, especially if possibly moving to the indieweb repo and pushed to wp.org repo. # 19:38 Loqi [gRegorLove] #133 Consider adding a version check method # 19:39 GWG I am going to update the plugin and then we can argue the name # 19:40 GWG The Post Kinds version is 4 months old gRegorLove_ joined the channel
# 19:56 Zegnat Do we have confirmation that it is dependency hell and not a parser bug? # 19:59 GWG I will push an update. And look at mf2 conflicts between plugins # 20:25 Loqi gwg has 10 karma in this channel (233 overall) # 20:26 sknebel Zegnat, did you implement actual vouch logic, or "just" the data collection for it? # 20:28 Zegnat I have logic, but don’t yet do anything with it because I haven’t thought of how I am going to keep a record of who is allowed to vouch yet # 20:30 Zegnat I started with a list of hostnames, but that wouldn’t do for sites hosted in subfolders, so I need to change my thinking # 20:30 GWG Okay. Indieweb Post Kinds php-mf2 updated....problem fixed. # 20:30 Loqi gwg has 11 karma in this channel (234 overall) # 20:31 Zegnat (This was an interesting piece of debugging, makes me believe WP dev work should continue in -dev and not be split off to a different channel) # 20:32 [miklb] imagine though someone editing the wiki at the same time # 20:40 [miklb] GWG do you provide a markdown and txt version of the README? AngeloGladding joined the channel
# 20:49 GWG [miklb]: I took pfefferle's approach of writing a txt version and using Grunt to convert it # 20:53 [miklb] ah, ok. Not sure a plugin like the press-this needs a grunt task. D # 20:55 GWG I'm increasing the complexity of the uf2 plugin right now. # 20:55 GWG Dividing the different sections into files so the logic of loading them can be better enhanced. # 20:58 [miklb] whenever I read “increasing complexity” I get itchy ? # 21:10 Zegnat in my experience, refactoring is a lot of moving bits and bobs around, creating new files whenever you think it is better. But it doesn’t sound as bad as “increasing complexity” # 21:13 [miklb] interesting. I’ve always thought refactor meant to simplify and clean up. # 21:13 GWG [miklb]: Sometimes it is about making it easier to work with # 21:15 [miklb] sure. I’m just asking general questions. Not questioning decisions. # 21:17 Zegnat A lot of refactoring I did at my previous job involved writing new libraries next to legacy code. With all the extra files and documentation it felt like I was making the project more complex. But overall it became easier to find where code was. # 21:17 [miklb] but I don’t think of making easier to work with increasing complexity ¯\_(ツ)_/¯ # 21:18 Zegnat I expect GWG referred to the same sort of thing when he said increasing complexity. # 21:19 Zegnat The number of people who equal “more files” to “more complexity” is staggering, in my experience. # 21:20 GWG It is hard to have a plugin try and do what that plugin does. It's very simple and needs more. # 21:22 Zegnat sknebel, re: vouch, I also tried to work rel scanning into my vouch flow. If you say trusted.example.com linked to you, my backend will verify that said link does not have nofollow on it. # 21:24 Zegnat So people adding rel="nofollow" to dynamic content like comments will help keep vouches spam free :) # 21:24 Zegnat Someone who commented to a blog I trust can’t suddenly pass my vouch by claiming the blog linked to them. # 21:25 Zegnat Though I am not sure if Indieweb comment displayers add rel="nofollow". Need to do more research on that. # 21:26 sknebel for indieweb sites, it makes sense to mf2-parse and check sites that way. everything that's not a property or in the content can be ignored # 21:26 sknebel hm, didn't we have proposals for signed webmentions somewhere? # 21:27 Zegnat I’d rather keep it general, sknebel, keep mf2 parsing out of my vouch verification # 21:28 Zegnat [miklb], I seem to recall that it does. I think most CMS will. It is people pulling in comments from webmentions on their indie sites that I am uncertain about. # 21:29 sknebel ^^^ another example of a vouch implementation on there # 21:35 [miklb] so what is the argument for code in the indieweb repo to not be copyleft? # 21:39 Loqi [Zegnat] #39 Normalise licensing for projects in the IndieWeb GitHub org. # 21:40 sknebel Zegnat: copyleft is GPL-style, so nothing is copyleft there # 21:41 Zegnat Oh. My reason for not going GPL-style is virility. If you want the microformat and indieweb specs to be pickedup everywhere and parsers in the indieweb repo to be leading implementations, no reason to force users into specific licensing. That’s why I would like to see more CC0. # 21:45 sknebel v2 is compatible with GPLv3 in terms of patent rules, maybe you remembered news about that? # 21:49 [miklb] TIL the MIT license isn’t really the MIT license rather X11 or at least that’s what gnu says. tantek joined the channel
# 23:06 schmarty is excited about IWS! is also getting over a head cold. :| # 23:23 schmarty i had been abusing in-reply-to because likes/replies/etc would show previews but plain mentions would not. # 23:24 schmarty so i now auto link preview any URL on its own line, using the same mechanisms i use to show reply contexts. # 23:26 tantek and yes, while similar in components / building blocks, a reply-context is very different from a link-preview! # 23:28 schmarty oops. i just noticed that page has no "display" section. # 23:28 schmarty i display link previews similarly to reply-contexts, as a nested h-cite. # 23:29 schmarty possibly wrong or confusing but i am going to sit with it for a bit # 23:30 schmarty haha, even my explanation just now was confusing. i meant that unlike a "u-in-reply-to h-cite", it's just an "h-cite". there's still a "u-url" inside w/ the url of the cite. [eddie] joined the channel
# 23:30 [eddie] schmarty: That sounds pretty cool! I don’t fully understand what you are talking about though ? Do you have an example? # 23:31 Loqi [Marty McGuire] Thanks to our guest Jeffrey Toppe for showing us the money in this week’s We Have to Ask Podcast! We Have to Ask PodcastWe Have to Ask Podca... KartikPrabhu joined the channel
# 23:31 [eddie] ohhh So your post is the text at the top and the embed below is the link preview? # 23:32 schmarty yep, and the link previews can go anywhere in a post, they're denoted by a URL on its own line. # 23:33 [eddie] ohhhh, that was my next question. So the link preview is not actually extracted from the text above, it’s specifically you putting a URL and nothing else that it knows you want to “embed” that post # 23:33 [eddie] I was wondering how you kept un-wanted previews from showing up when you mentioned something off-hand ? # 23:34 [eddie] But that makes sense. Almost like a Twitter card embed # 23:34 schmarty [eddie]: yep, i chose that because i think i'd be unlikely to do it by accident # 23:34 schmarty yeah, it brings posting to my own site much more inline with how i post to twitter or FB. # 23:34 Loqi schmarty has 4 karma in this channel (16 overall) # 23:35 schmarty haha, thanks! i have been letting my site improvements drag a bit recently and was glad to get back to it before IWS next weekend. # 23:37 schmarty i need to reload my itches for my site and tools like /Screech into my head and try to prioritize them [miklb] joined the channel
# 23:42 [miklb] schmarty what did you use to parse the content for the preview? eli_oat joined the channel
# 23:44 schmarty in general, the ruby microformats gem. i have special cases for soundcloud and youtube (to embed a player) and twitter (using XRay as a service). i fall back to simply displaying a link if none of those produce something usable. # 23:46 Loqi XRay is an open source API that returns structured data for a URL by parsing microformats and following other indieweb algorithms, and is part of the p3k suite of applications https://indieweb.org/XRay gRegorLove joined the channel
# 23:51 tantek schmarty - yes, you want to make sure that any link-preview you put in your posts, doesn't show up when someone else shows a link-preview of YOUR post. # 23:52 tantek e.g. when Loqi does a "link-preview" of something here in IRC like you saw :) # 23:57 gRegorLove You could close the p-name before the link preview, but that's not very flexible if you have [content] [link preview] [more content] # 23:57 tantek schmarty - perhaps take a look at existing examples of link-previews of posts with link-previews inside themselves! # 23:57 tantek e.g. silo examples, and see how they are doing it # 23:57 tantek e.g. xray applied to tweet permalinks, where the tweet permalink itself has its own embedded link-preview
tantek.com edited /2017/Leaders (+86) "/* Schedule */ lunch after, then afternoon activities outside" (view diff)
tantek !tell Zegnat I like this new "get it done by Friday" policy for HWC pages / venues. Anything posted by Friday 17:00 Pacific Time I'll add to the Mozilla announcement for the next Monday, but otherwise no promises. In addition I think /next-hwc should always point to the next "regular" Wednesday evening HWC. Like usual fortnightly or a reliable regular city (e.g. Portland).
eli_oat this post was a doooozy: http://tantek.com/2017/153/t1/sunrise-fog-summits-npsf
Loqi [Tantek Çelik] a jpg. a jpg. a jpg. a jpg. a jpg. a jpg. a jpg. a jpg. a jpg. ? #sunrise vs #fog on 4 summits before 6:40, then #NPSF #hillsforbreakfast. 1552' in 4.7mi. 51d til #TSFM2017, my longest race so far @THESFMARATHON.
Photos from / by / of:
1. Twin Pe... https://igx.4sqi.net/img/general/original/476_XKESZJ1fzJwrMcsMWNR82TQhNpWIQ1nhzoIK2n9V0NE.jpg
vanderven.se martijn edited /2017/Leaders (+918) "/* HWC Organizing */ more notes, thoughts, and expressing interest" (view diff)
@SocialMedia547 Improving the webmentions directory http://business-commerce.InfignosMedia.com/socialopine/index.cfm?a=T837083&subcat=71 #SocialMedia ? (twitter.com/_/status/876074508251258880)
Zegnat KevinMarks, you were there for the first Virtual HWC right? On Hangouts? Could you add a point about Hangouts to /Virtual_HWC ?
gRegorLove Howdy. I saw talk of php-mf2 backcompat?
GWG gRegorLove: I'm testing the theory with this page. https://tiny.n9n.us/2016/09/11/like-of-hydrox/
loqi.me created /WordPress_uf2 (+155) "prompted by gRegorLove and dfn added by gRegorLove" (view diff)
gregorlove.com edited /WordPress/Plugins (+20) "/* Other Indieweb Plugins */ +WordPress uf2" (view diff)
martymcgui.re edited /User:Martymcgui.re (+25) "/* What Works */ now displaying link previews for some mentions \o/" (view diff)
schmarty thanks! i used some of my downtime today to implement /link-preview display on my site