2022-02-02 UTC
cygnoir[d], jacky, [jgmac1106], darkkirb, [fluffy], [chrisaldrich], [Aaron_Klemm], [benatwork], ur5us, [chrisaldrich]1, [benatwork]1, [schmarty], hoenir, aaronpk[d], Myst[d], mossymaker[d], MarkJR84[d], sarahd[d], grantcodes[d], marksuth[d], KartikPrabhu, [jackjamieson], phae, [aciccarello], bg2, zachburau[d], angelo, Seirdy, P1000[d], Christian_Olivie, [manton], hans63us[d], Ramon[d], corenominal[d], capjamesg[d], sayanarijit[d], daiyi[d], Zegnat[d], GWG, klez, barpthewire_, m5zs7k, saptaks_znc, rektide, ben_thatmustbeme, Saphire, sebbu, willnorris, yolk, Kaja, aaronpk, jamietanna, sknebel, aspenmayer[d], balupton[d], [tw2113_Slack_], Seb[d], rattroupe[d], [KevinMarks], IWSlackGateway, Jeremiah[d], Zegnat, JSharp, [jeremycherfas] and gRegor joined the channel
[tonz], jacky and KartikPrabhu joined the channel
# 17:16 [tantek] capjamesg[d] your updated page is getting some views ^ 😂 # 17:17 [tantek] You could reply saying hi thank you and that was your first MDN update and feedback & suggestions welcome! KartikPrabhu and Asaf_Agranat[d] joined the channel
# 19:41 [tantek] aaronpk have we considered adding IndieAuth login to the microformats wiki also? would there be a way to preserve existing user/pw logins while allowing new IndieAuth logins? # 19:42 [tantek] if we're able to make that work, I can think of a few more MediaWiki wikis that would likely also adopt a similar setup # 19:42 aaronpk i suspect any time that was brought up in the past i said we would have to wait until the upgrade to revisit it :) # 19:43 [tantek] I figure especially with capjamesg[d]'s success in setting up a new MediaWiki with IndieAuth, it feels like time antrdnv[d] joined the channel
# 19:46 GWG aaronpk: How many accounts does the microformats wiki have? # 19:47 [tantek] GWG, lots have been created & banned due to spam holes of the past # 19:47 [tantek] at this point it's too much work to moderate any new "public" wiki that an individual or small group would setup # 19:48 [tantek] hence the iteration on getting IndieAuth sign-in working for MediaWikis in general, and getting it adopted by more wikis (so it can become a "normal" thing for any new MediaWiki) # 19:49 aaronpk I do think if we expect it to be used by other mediawikis it makes more sense to build the whole indieauth bit into the plugin itself # 19:49 aaronpk that would remove the setup step capjamesg[d] had to do with Vouch-Proxy # 19:49 aaronpk the main advantage of vouch-proxy is that you can share the login across a bunch of different subdomains, which is useful for indieweb.org # 19:50 sknebel (also, it goes through indielogin and thus doesnt just do indieauth, right?) # 19:51 aaronpk but then that also brings up the question of whether it should be just indieauth, or also support relmeauth to twitter/etc, which then adds the additional setup step of getting api keys # 19:51 aaronpk correct, vouch-proxy doesn't do indieauth itself, it does oauth/oidc to a single configured provider # 19:51 sknebel which is indielogin, which gives us indieauth and relmeauth? # 19:52 [tantek] "share the login across a bunch of different subdomains" is super useful yeah # 19:53 aaronpk but yeah frankly i wouldn't recommend the indieweb wiki setup as is for other mediawikis ur5us and gRegor joined the channel
# 20:13 capjamesg[d] I am so happy that Vouch Proxy supports indieauth. But the documentation for actually getting set up was quite crowded. # 20:13 capjamesg[d] Not on the indie auth part, the vouch proxy installation and setup part. # 20:14 capjamesg[d] I think a mediawiki plugin might make adoption a bit easier to those who haven’t used the IndieWeb wiki much. # 20:15 GWG Would it be a IndieAuth plugin or a websignin plugin? # 20:16 aaronpk the problem with full web sign-in is needing API keys, and at this point really only github is easy to get api keys from # 20:20 GWG aaronpk: What about just the pgp key, I meant # 20:27 aaronpk i should check the stats to see how much usage pgp actually gets jacky, barpthewire_, ben_thatmustbeme, IWSlackGateway, rattroupe[d], Zegnat, [KevinMarks], sebbu, Seb[d], gRegor, Jeremiah[d], darkkirb, corenominal[d], aspenmayer[d], [tw2113_Slack_], aaronpk, [fluffy], sknebel, [Aaron_Klemm], ur5us, yolk, hans63us[d], balupton[d], [tonz], [jgmac1106], klez, GWG, KartikPrabhu, [schmarty], hoenir, Myst[d], aaronpk[d], mossymaker[d], MarkJR84[d], sarahd[d], Kaja, willnorris, daiyi[d], antrdnv[d], capjamesg[d], rektide, Ramon[d], sayanarijit[d], Christian_Olivie, Seirdy, P1000[d], [jackjamieson], m5zs7k, Zegnat[d] and JSharp joined the channel
[aciccarello], angelo, [manton], phae and saptaks_znc joined the channel
# 21:16 [tantek] Feels like a data graphic of some kind would be useful here # 21:16 [tantek] What's the time period for those stats aaronpk? Daily? Past year? sebbu joined the channel
Saphire and jacky joined the channel
# 22:16 [schmarty] haha i love the PGP option but i am also pretty much always signed into github so i just use that when i use indieauth.com 😅 # 22:17 [schmarty] 1-click vs click, copy, paste into command line script, type GPG password, return to browser, paste, click # 22:17 aaronpk oh the trick is to set up a hot key to sign the text in place :) # 22:18 aaronpk pgptools on the mac can do it, it shows up in the "services" menu when you select any text in any app # 22:34 jacky but oddly enough, the structure lightly resembles pseudo code which makes it good for a scan # 22:42 jacky > if that child element itself has a microformat ("h-*" or backcompat roots) and is a property element, add it into the array of values for that property # 22:43 jacky like if I have a <a href="/foo" class="u-author h-card u-url">Footster</a>, the properties there would be `u-author` and `u-url` # 22:44 jacky unless "else use the parsed property value per p-*,u-*,dt-* parsing respectively" means "use the property element as the value to be added to the property class that was found" # 22:44 jacky actually those are two separate things (which I kinda blame on the formatting of this page) # 22:57 sknebel also, multiple property classes on an element mean all of them get parsed # 23:01 sknebel i.e. the "is a property element, ..." implies "do this for each property" # 23:01 jacky I think I'm confused b/c of this phrase "add it into the array of values for that property " # 23:02 jacky because in my head, that means with "u-author h-card u-url", I'd be adding it (the parsed h-card) to the 'author' and 'url' properties of its parent, which is _wrong_ to me # 23:03 sknebel I dont think that specific example would make sense very often, but thats how it works # 23:04 jacky but it's probably b/c my example is very not likely when it's been semantically written # 23:05 sknebel nested elements really work the same as any other property value # 23:05 sknebel i.e. "u-url u-uid" is a common pattern, resolves in two properties with the value # 23:05 sknebel if that happens to be a h-* root instead of a plain u- value, it doesnt care # 23:05 jacky I guess now what's tripping me up is resolving one's "value" # 23:07 jacky b/c tbh it makes _perfect_ sense now (the value changes depending on who's looking at the property) # 23:08 sknebel so it probably makes sense to parse the child once, and then for each property its "assigned" to calculate the value that fits for the case # 23:08 Loqi sknebel has 3 karma in this channel over the last year (35 in all channels) # 23:09 sknebel (technically for each property type, but without measuring I can't tell you if thats actually faster in practice to make that distinction. the python parser does it that way, but not everything is python and I don't remember if that was done that way as a result of profiling or not) # 23:11 sknebel so, I'm off to bed now, but happy to help tomorrow if theres other questions # 23:11 jacky this was a considerable road block for me :) # 23:21 [tantek] jacky, there's a challenging balance in specs between precision and conciseness to prevent ambiguity. shorter and fewer words usually means fewer places for unintended "bugs" in the prose to creep in # 23:22 [tantek] specs have their own "specese" (pronounced like legalese) as a result # 23:29 jacky I could have imagined this being a bit tricky to implement the first few times around # 23:29 jacky I think potentially some inline examples _could_ help # 23:29 jacky but that's what the test suite is for, I guess ur5us joined the channel
capjamesg[d] Yes 😂
[schmarty] haha i love the PGP option but i am also pretty much always signed into github so i just use that when i use indieauth.com 😅
jacky I was actually coming in to say that I kind of wish that https://microformats.org/wiki/microformats2-parsing had either a micro-glossary or was re-worded a bit to be less sparse in some parts