#dev 2018-03-04

2018-03-04 UTC
[eddie] joined the channel
#
[eddie] !tell [cleverdevil] I used https://gist.github.com/cleverdevil/f33530706d6e8dacd13a8bd8e8c15dba to output a json file and attempted to use https://gist.github.com/cleverdevil/c857695bb2de1e46686d720cad9d124c with that same file. It claims “ValueError: No JSON object could be decoded” Did you ever run into that?
#
Loqi Ok, I'll tell them that when I see them next
#
singpolyma.net edited /IPFS (+244) "IPFS is selfdogfooded" (view diff)
#
aaronpk I think that distinction is about whether the individual people behind IPFS use it for their own websites
apb, chrisaldrich and KartikPrabhu joined the channel
#
aaronpk post-type-discovery++
#
Loqi post-type-discovery has 1 karma
#
aaronpk just came in handy for https://aaronparecki.com/2018/03/03/16/
#
Loqi [Aaron Parecki] Here's a screenshot of Aperture's new channel filter UI. https://aaronparecki.com/2018/03/03/16/photo.png
ben_thatmustbeme and renem joined the channel
#
aaronpk yesssss channel filters are working
mblaney joined the channel
#
aaronpk oh no Zegnat i'm still getting a weird error from your site! "empty reply from server"
#
mblaney I'm getting a bit jealous of all the new reader functionality going on...
#
mblaney might be time to break out some reader dev work? didn't know other people's itches were contagious ;-)
[eddie] joined the channel
#
[eddie] aaronpk: Does the filter happen on the parsing side or the timeline side?
#
[eddie] As in, does it exclude posts from being stored or from being displayed?
tantek and KartikPrabhu joined the channel
#
aaronpk [eddie]: it stops it from adding to the channel but still stores it internally
#
aaronpk Works better that way if you have a feed in multiple channels with different filter options
tantek, Kyle-K and [kevinmarks] joined the channel
#
Loqi Ok, I'll tell them that when I see them next
#
Zegnat !tell aaronpk getting an empty reply from my site is a feature, not a bug. But it might be time to turn it off because people have a hard time figuring it out...
#
Zegnat !tell aaronpk Actually, I think I left you a sneaky PR with the fix somewhere . . . 😇
#
Loqi Ok, I'll tell them that when I see them next
barpthewire joined the channel; mblaney left the channel
#
loqi.me created /Gitter (+175) "prompted by Zegnat and dfn added by Zegnat" (view diff)
#
kaja.sknebel.net edited /Gitter (+2) "linkify ('… is <url>' pattern)" (view diff)
#
loqi.me edited /Gitter (+195) "Zegnat added "All Gitter rooms are available through an [https://github.com/gitterHQ/irc-bridge open-sourced] [https://irc.gitter.im/ IRC bridge] ([https://xkcd.com/1782/ of course it is])." to "See Also"" (view diff)
barpthewire joined the channel
#
dgold shakes tiny fists at people who's microformats markup fails to encompass author identity
#
sknebel heh
#
sknebel for usability, or because it broke code you wrote assuming there is one always?
#
Zegnat dgold, who would ever do such a thing?!
#
dgold sknebel: the former, of course! Would that I, a mere neophyte, would care about making an ass of u & me.
#
dgold yes, the latter.
#
dgold especially given that js is really unforgiving of such things
#
sknebel the authorship algorithm is a bit of complexity always, yeah
#
sknebel not sure if we have a JS-version of that yet
#
sknebel (I guess there is the one in sturdy-backbone, but that's fairly hacky and filled with code to create the visual display)
[gerwitz] joined the channel
#
[gerwitz] hello, lazyslack!
#
[gerwitz] I am presently implementing an indieauth client, and find it very curious that I’m supposed to send the user with a domain for “me” (hans.gerwitz.com) but when I verify the code will receive back a URL (https://hans.gerwitz.com/)
#
[gerwitz] I’d like to understand why so I can figure out whether the protocol part of that returned URL is something I need to pay attention to
#
Zegnat I am not understanding, what do you mean with “supposed to send the user with a domain”? You always use full URLs in IndieAuth
#
[gerwitz] Ah, perhaps I am misled by https://indieauth.com/developers
#
Loqi IndieAuth.com
#
[gerwitz] > placeholder=“yourdomain.com”
#
Zegnat I think IndieAuth.com might handle normalisation for you, so people do not have to enter everything
#
Zegnat I would refer to https://indieauth.spec.indieweb.org/ if you are building anything yourself
#
Loqi [Aaron Parecki] IndieAuth
#
[gerwitz] (it’s interesting that the spec calls for canonicalization for URLs but prepending “a scheme” rather than a specific default… my recent transition from http: to https: has caused me a lot of confusion thanks to this sort of ambiguity.)
#
Zegnat I think I have discussed that, the reason it doesn’t specify is because it should be up to the client implementation to decide if they want to enfore HTTPS default or not
#
Zegnat See https://github.com/indieweb/indieauth/issues/8
#
Loqi [Zegnat] #8 Should URL Canonicalization specify a default scheme?
#
[gerwitz] thanks for that issue reference. I prefer to be either convinced or at least understand what I’m disagreeing with 😉
#
[gerwitz] Any tips for development workflows would be nice. I’m finding myself deploying to a test instance on Heroku quite often, but don’t want to set up a copy of indieauth.com just for dev 😕
#
aaronpk if you're making an indieauth client you shouldn't have any reason to set up a copy of indieauth.com
#
Loqi aaronpk: Zegnat left you a message 6 hours, 3 minutes ago: getting an empty reply from my site is a feature, not a bug. But it might be time to turn it off because people have a hard time figuring it out...
#
Loqi aaronpk: Zegnat left you a message 5 hours, 59 minutes ago: Actually, I think I left you a sneaky PR with the fix somewhere . . . 😇
#
aaronpk Zegnat: ?? I don't see a PR on XRay which is the only place I can think of that would be able to change that behavior
#
[gerwitz] Maybe my terms are poorly calibrated. I’m verifying domain ownership and will eventually implement an indieauth endpoint, but for now am using indieauth.com (the site, not the project) as a consumer.
#
[gerwitz] …which makes testing pretty shallow until the code’s deployed where it can expose a callback URL to indieauth.com.
#
aaronpk okay, so a few things. in this case, indieauth.com is meant to be "prove domain ownership as a service" if you're building something where you want people to log in with their domain name.
#
aaronpk for example, the IWC wiki uses it, because I didn't want to build that code into the wiki itself
#
aaronpk indieauth.com allows the user to use multiple methods to prove they own a domain name. RelMeAuth (twitter/github), PGP, email, and IndieAuth authentication.
#
aaronpk the redirect URL involved with indieauth.com doesn't actually need to be reachable by indieauth.com, it only needs to be reachable by the computer you're logging in on. so you can definitely use a localhost redirect URL for testing
#
[gerwitz] oh, christ, duh.
#
[gerwitz] thank you
#
[gerwitz] I was confused because of the need to share my callback URL when verifying an auth code. I suppose indieauth.com uses that value for verification against the original request only?
#
[gerwitz] (the more params the merrier, I guess)
#
aaronpk yeah exactly
#
[gerwitz] speaking of auth code verification, is it most “best practice” to trust the returned “me”, verify it against my expectations, or ignore it?
#
aaronpk https://indieauth.spec.indieweb.org/#authorization-code-verification "The resulting profile URL MAY be different from what the user initially entered, but MUST be on the same domain."
#
aaronpk you should use the `me` returned, but only if it is on the same domain as the user originally entered
#
Loqi I agree
#
aaronpk for example Known will return URLs like "https://known.kevinmarks.com/profile/kevinmarks" even if he entered "http://known.kevinmarks.com/" initially
#
[gerwitz] 🤔
#
[gerwitz] scheme change is normal, then?
#
aaronpk yeah since the site knows which scheme is canonical but the user typically enters a bare domain or just http
[kevinmarks] joined the channel
#
[kevinmarks] Actually, known.kevinmarks.com is http only, but upgrading is good practice. I should probably work out how to fix that
#
[kevinmarks] (slack notified me)
#
[gerwitz] should I silently update a user’s URL if they re-authenticate later and I receive a new one?
#
aaronpk that's trickier. it's probably safe to consider http->https an "updated" URL for the user
#
aaronpk but not a change like "known.example.com" to "known.example.com/profile/user1" since then user2 on that site could steal the login
#
[gerwitz] Unrelated to my effort today: I should share that as a new user there are some cases where this gets confusing. (I think I had my site redirecting to force https:, but my github profile linking to http:, and an auth endpoint at indieauth.com … and some-but-not-all micropub clients threw errors telling me that `https:` != `https:`)
#
[gerwitz] as for updating, I think I’m doubling down on an earlier decision to use domain alone as my canonical user ID. I’ll update their URL quietly (since I’m not really using it) as long as the domain portion matches, and balk if it doesn’t.
#
aaronpk that will fail for multi-user sites, but if that's a limitation you're okay with then that's fine
#
[gerwitz] It definitely is
#
[gerwitz] (I’m building a tool for two sites, but over-engineering for the experience 😉 )
[miklb], barpthewire, j12t_ and tantek joined the channel
#
tantek.com edited /iCloud (+90) "expand dfn, silo of sorts" (view diff)
[cleverdevil] joined the channel
#
[cleverdevil] Good morning 🙂
#
[cleverdevil] whoa, nice job on the Aperture channel filtering aaronpk++!
tantek joined the channel
#
aaronpk 🎉
#
[cleverdevil] !tell eddie that's because I changed how I did it to utilize individual JSON files per post, rather than a single large monolithic file - updated gist here -> https://gist.github.com/cleverdevil/f33530706d6e8dacd13a8bd8e8c15dba
#
Loqi Ok, I'll tell them that when I see them next
#
[cleverdevil] Suggestion for you [aaronpk] on Microsub. It'd be nice to be able to either per-channel or per-source request no unread tracking.
#
aaronpk I was thinking about that too
#
[cleverdevil] Currently, I have Twitter feeding into Aperture, and honestly I don't enjoy it.
#
aaronpk same for me with instagram. there's *always* new photos so it's not meaningful to have the unread counter
#
[cleverdevil] Part of it is that it makes my unread count enormous, and I don't view Twitter as a "make sure to read it all."
#
[cleverdevil] Same.
#
aaronpk I think I will start per channel, because per source seems like it'll be too hard to see what's going on
#
[cleverdevil] That makes sense to me.
AngeloGladding joined the channel
#
[cleverdevil] I do think, eventually, there will need to be more granularity within channels to support more traditional feed reader behaviors, for example viewing a channel as a "folder" with feeds inside.
#
[cleverdevil] Where you can drill down into specific feeds.
#
aaronpk I agree, I think that's a natural extension
#
aaronpk the other thing i was thinking was being able to group channels together, like put channels in folders
#
[kevinmarks] twitter should never have added unread counts
#
aaronpk But for now can you open an issue on Aperture requesting the per-channel unread count feature?
leg joined the channel
#
[cleverdevil] Sure!
#
[cleverdevil] On Aperture, or Microsub?
#
aaronpk Aperture
#
[cleverdevil] :thumbsup:
#
aaronpk I don't think it needs to be a Microsub feature yet
#
aaronpk If you want to be able to toggle that setting from any app then it'd be a Microsub feature but I'd like to hold off on adding the kitchen sink to the spec until we demonstrate how useful these features are
#
[cleverdevil] Done -> https://github.com/aaronpk/Aperture/issues/28
#
Loqi [cleverdevil] #28 Per-channel unread tracking
#
[cleverdevil] Makes logical sense 🙂
#
aaronpk For example I could totally see someone building a Microsub server that doesn't do any read tracking at all intentionally
#
[cleverdevil] Sure.
#
[cleverdevil] If we're going to get into that world, though, I'd prefer to see these sorts of things documented in the standard, but marked as optional, with some sort of capability detection mechanism.
#
[cleverdevil] But, we're getting ahead of ourselves methinks 🙂
#
[cleverdevil] (Or, I am getting ahead of myself I should say).
#
aaronpk it only needs to be in the standard if it's part of how clients interact with servers. if it's just server behavior it doesn't need to be in the spec
#
aaronpk another fun one i'm planning on is dynamic/virtual channels, where aperture will add things to a channel based on filtering other content from other channels. that doesn't need to be in the spec because as far as clients are concerned it'll work just like a normal channel
#
[cleverdevil] That's a cool feature 🙂
#
www.boffosocko.com edited /podcast (+352) "Silo examples: Anchor.fm and Cast; highlighted microcast examples" (view diff)
[jeremycherfas], KartikPrabhu and [eddie] joined the channel
#
[eddie] !tell cleverdevil: Ahh, awesome. I’ll try out the new gist. Thanks 😄
#
Loqi Ok, I'll tell them that when I see them next
#
Loqi [eddie]: [cleverdevil] left you a message 1 hour ago: that's because I changed how I did it to utilize individual JSON files per post, rather than a single large monolithic file - updated gist here -> https://gist.github.com/cleverdevil/f33530706d6e8dacd13a8bd8e8c15dba
eli_oat joined the channel
#
[eddie] aaronpk: Awesome, I’m gonna have to update my Aperture today to get the filters 😄
#
aaronpk if you wait an hour or so you'll get per-channel read tracking mode settings too :D
[miklb] joined the channel
#
[miklb] what is aperture
#
Loqi Aperture is an open source Microsub server developed by Aaron Parecki https://indieweb.org/Aperture
#
aaronpk [eddie]: am I going to break Indigenous if suddenly `unread` is a boolean instead of an integer?
#
[eddie] hmmmmm 1 of 2 things will happen: It will break or it will just stop showing unread indicator for that channel
#
aaronpk let's find out! 🎉
#
[eddie] haha sounds good.
#
Loqi awesome
#
aaronpk it already handles the case when no `unread` property is present, right?
#
[eddie] It’s supposed to, but I haven’t tested it yet and I don’t remember if that code is in TestFlight or on my local machine
#
[eddie] That said I should probably push a new point release
#
[eddie] So if it doesn’t then the new point release will handle it
#
aaronpk cool
#
[eddie] I’ll push that later today once I verify it handles unread=true okay
#
dgold https://media.ascraeus.org/file-20180304190410-6716.png
#
[eddie] So if anything Indigenous should only break for people for like, a day
#
dgold vv basic websub reader
#
aaronpk dgold++ nice
#
Loqi dgold has 14 karma in this channel (34 overall)
#
[eddie] Awesome, dgold!
#
aaronpk [eddie]: the benefit of having very few users right? :D
#
[eddie] aaronpk: Exactly
#
[eddie] Especially since the microsub users on Indigenous are even less
#
dgold how many of us are there, eddie?
#
[eddie] I’m pretty sure in microsub land we have: aaronpk, [cleverdevil], dgold and I
#
[eddie] There are 40 on the beta for micropub (but they don’t all login often, obviously)
#
[cleverdevil] There is also [grantcodes] and I think [schmarty]
#
Loqi [cleverdevil]: [eddie] left you a message 10 minutes ago: Ahh, awesome. I’ll try out the new gist. Thanks 😄
#
[eddie] oh yeah!
#
[cleverdevil] But yeah it’s a small group 😀
#
dgold that's *loads*
#
[cleverdevil] 😉
#
[eddie] That’s right! I forgot about schmarty joining the microsub land. I can’t remember if grantcodes is on Indigenous or not, but he’ll probably have to fix Together after the change 😉
#
[cleverdevil] Have you considered using Together as a starting point for your Electron client, [dgold]?
#
dgold I have, [cleverdevil], but I need to understand how electron works first, hence getting basic functionality working
#
dgold its been an education
#
[cleverdevil] Cool that makes sense.
#
[cleverdevil] I’d be totally open to including an Electron based thing integrated directly into Together. Especially if it can give us a mobile app 😀
#
[eddie] [cleverdevil] I’m sure you hit “published” on some posts during this process and they didn’t publish correctly. Do you just move the file from the published folder back into the main mf2 folder?
#
[eddie] 😊
#
[cleverdevil] Yup. This is a very manual process at the moment 😀
#
[eddie] Oh yeah the manual is fine. Just wanted to make sure moving the file wouldn’t mess anything up 🙂
#
[cleverdevil] For the most part I skipped anything that was more than a note.
#
[eddie] Yeah turns out my site doesn’t like really old notes
#
[eddie] 😆
#
[eddie] It’s conflicting with my location/weather stuff, so I need to tweak that
#
[cleverdevil] I’m thinking of doing all of the photos via my media endpoint and manually crafting “stories” using very similar Markup to [manton]’s Sunlit.
#
[eddie] Ohhh interesting
#
[cleverdevil] What would be even cooler would be if [manton] added Facebook as an input source to Sunlit so that I could use the excellent app instead of hacking it together myself.
#
[eddie] Ohhhh that would be interesting
#
KartikPrabhu likes how [eddie] adds "h"'s to the "Oh"
#
[eddie] lol that’s my deep thought response
#
[eddie] [cleverdevil] Something I just realized, granary seems to generate the facebook URL as a url property rather than a syndication attribute. Do you manually make that change in your known backend or do you not carry the syndication link into your system?
#
[cleverdevil] I didn’t carry the syndication link :/
#
[eddie] Now that I think about it, if I might be deleting these from fb, I’m not even sure if I WANT to keep the syndication url around in my system
#
[cleverdevil] I’m likely going to go back and add them in, or delete all of the old content and republish it once I figure out how to get Known to accept it.
#
[eddie] ohhh okay
#
[cleverdevil] I’d like Bridgy to crawl it all before I delete it from Facebook.
#
[eddie] Bridgy only does the recent posts, but I wonder if I import the syndication links and I install a local copy of bridgy temporarily if I can just change the code to not have a limit and scan my whole facebook 🤔
#
[eddie] I might try that, since I’m wanting to learn a bit more about Bridgy’s inner workings for https://github.com/snarfed/bridgy/issues/796
#
Loqi [snarfed] #796 micropub API for bridgy publish
#
aaronparecki.com edited /Instagram (+87) "/* username takeovers */" (view diff)
#
[eddie] [cleverdevil] Not sure if you’ll want it or not but I got it to provide the syndication url: https://gist.github.com/EdwardHinkle/45b5828b43987eb89610e3f22a15eb58
#
[cleverdevil] Nice, thanks!
#
[cleverdevil] I may just go through and delete/re-publish my old migrated content.
#
[eddie] :thumbsup:
#
[cleverdevil] So, the content filtering doesn't seem to work for my Twitter channel, [aaronpk]
#
aaronpk I just noticed that and i'm fixing it :)
#
[cleverdevil] I am using granary.io with HTML format.
#
[cleverdevil] Ah, cool 🙂
#
aaronpk that's what I get for not using a real enum
#
aaronpk "repost" vs "reposts"
#
KartikPrabhu ha! that is similar to what caused me to almost break mf2py "prop" vs "props" for variable name
#
aaronpk [cleverdevil]: pushed and fixed. you'll need to update the settings on each channel again
eli_oat joined the channel
#
[cleverdevil] Cool
#
[cleverdevil] I assume it will only apply to new items?
tantek and eli_oat joined the channel
#
aaronpk Correct
snarfed, tantek and [snarfed] joined the channel
#
[snarfed] [eddie] you can make bridgy backfill an individual post by plugging it into the "Resend for post"box on your user page
#
[snarfed] not sure how many you want to do though :P
#
[eddie] ohhhh, I didn’t realize that snarfed. Good to know
#
[eddie] lol and understood 😉
#
[snarfed] definitely scriptable, just put a POST to that form in a loop
#
[cleverdevil] Ah, much better. Now, I've got a trimmed down Twitter list that I follow in Together, with read tracking off, and RTs filtered out.
#
[cleverdevil] One thing I'm missing from my regular Twitter client is threading and oEmbed.
#
[snarfed] [cleverdevil] did you filter out RTs manually yourself?
#
[cleverdevil] [snarfed], nope, a new feature of Aperture!
#
[cleverdevil] :)_
#
[snarfed] nice! beat me to it in granary
#
GWG I have just started a new repository to hook a bunch of sensor data to my site.
#
loqi.me created /weaning_yourself_off_Facebook (+29) "prompted by tantek and redirect added by tantek" (view diff)
#
tantek.com edited /Facebook (+101) "/* How to wean yourself from */ delete the app" (view diff)
#
tantek.com edited /Facebook (+53) "/* How to wean yourself from */ 7 weeks somehow haven't checked notifications" (view diff)
[manton] joined the channel
#
[manton] [tantek] [aaronpk] Catching up on that discussion about your Atom feed... If I remember correctly, it's because Tantek includes the full text in the title field too. Micro.blog has special support for detecting this and trying to do the right thing, but there are some whitespace comparison issues that trip it up. The best solution is really to leave the title blank, although I'd like to fix this as well so that it falls back to the body text.
#
Loqi [manton]: sknebel left you a message 2 days, 2 hours ago: since you already recommend dnsimple for micro.blog (http://help.micro.blog/2015/custom-domains/), you should also add it to their service templates so people have an easier time configuring it! see https://github.com/dnsimple/dnsimple-services#fields
tantek, [kevinmarks] and [eddie] joined the channel
#
tantek wb [manton] !
#
[cleverdevil] [eddie] I've got photo albums converting now... it uploads all of the images in the album to the micropub media endpoint, and then constructs simple markup lifted from Sunlit (hi, [manton]) to generate a MF2-JSON file that I then publish manually.
#
tantek [manton]: not sure if you noticed, but I dropped replies from my Atom feed
#
tantek I decided that no one who consumed Atom was showing the necessary context for my replies to make sense, so they likely looked mostly like noise to those readers
#
tantek [manton] while you're here, I noticed my photos don't show up in micro.blog
tomwolf joined the channel
#
[manton] [tantek] I think the photo issue is because it's using your feed item title instead of the post HTML. I'll look at it again, but that was my memory from when I looked at this a few months ago. (It could also be that if it has to truncate the post, it loses the formatting and photos.)
#
tantek oh yeah I drop the photo URLs from the item title because they were just noise to users
#
[manton] [cleverdevil] That's cool.
#
tantek but the photos are the first thing in the entry contents so they shouldn't be truncated
#
[manton] I'll have to look at the feed again... But pretty sure it's using the title and linking back to your site, instead of using the contents.
#
[cleverdevil] https://cleverdevil.io/2012/hawaii-2012
#
[cleverdevil] 😄
#
Loqi [Jonathan LaCour] Hawaii 2012
#
[cleverdevil] This is so fun!
#
[cleverdevil] I need to do the same with Instagram backfill.
#
[cleverdevil] I assume I can just use OwnYourGram for that.
#
[eddie] Hmm I just realized something. My old posts are from Hawaii time…. lol So my site is going to show them according to maryland time. Not something to solve today but importing old posts has now created the time zone dilemma 😕
#
[eddie] I had two posts that were on the same day but on my site is seen as different days lol
#
aaronpk [cleverdevil]: OYG doesn't do backfill yet
#
[cleverdevil] [aaronpk] awww... consider this a formal feature request 🙂
#
[cleverdevil] I'd be awesome to just say "please backfill everything from this date backward"
#
aaronpk I do need to do that at some point, cause i'm still missing all my IG photos from before I launched OYG
#
[cleverdevil] :)_
#
[cleverdevil] Well, I am now officially done migrating my Facebook albums and notes.
#
[cleverdevil] Now, all I really have is individual photos.
#
tantek wow
#
[eddie] That’s awesome, [cleverdevil] Very inspiring. I’ve had getting FB data into my site sitting on a list for the past year, but never spent any time thinking or working on it until you took the plunge 🙂
#
[cleverdevil] 😄
#
[cleverdevil] Now, I am struggling to figure out how to get some of the individual photo data out of the exports.
#
[cleverdevil] I have a few "albums" like "iOS Uploads" and "Mobile Uploads" that are just very very large albums of unrelated photos that were uploaded from my phone.
#
[eddie] Ohhh yeah, me too
#
[cleverdevil] I have all of the individual photos, which is good, but it doesn't seem to have any of the metadata from the individual items 😕
#
[eddie] Ohhhhh, that’s junk
#
[eddie] I wonder if the node.js app just needs to be tweaked. I remember it mentioning that there are more fields available
#
[cleverdevil] Likely.
#
aaronpk sneak peek: https://media.aaronpk.com/Screen-Shot-2018-03-04-13-24-01-QwgDTRLBnb.jpg
#
[cleverdevil] Nice 🙂
#
aaronpk pretty sure I'm gonna build this webhook receiver into aperture
#
[eddie] I’m confused, what happened here?
#
aaronpk I want my github emails in aperture cause github web UI is too confusing to find them
#
[eddie] That makes sense
#
tantek I do feel a bit lost with github notifications as well so I end up ignoring them :/
#
[eddie] I use an iPhone app called GitHawk and it’s been working well for now. Leaves a notification there until I’m done with it and has useful share actions to I can use Indigenous to reply to stuff
#
[eddie] But I would like to get GH into Aperture at some point as well.
#
tantek what is GitHawk
#
Loqi It looks like we don't have a page for "GitHawk" yet. Would you like to create it? (Or just say "GitHawk is ____", a sentence describing the term)
[jeremycherfas] joined the channel
#
[eddie] GitHawk is an iOS app that allows a user to triage their GitHub notifications. While primarily silo-focused, it’s usage of the Share Sheet for every PR, issue and comment allows for IndieWeb usage with apps that support Micropub.
#
Loqi ok
#
loqi.me created /GitHawk (+258) "prompted by tantek and dfn added by [eddie]" (view diff)
#
[eddie] Feel free to tweak that
#
aaronpk [eddie]++
#
Loqi eddie has 26 karma in this channel (46 overall)
#
[eddie] GitHawk < https://itunes.apple.com/us/app/githawk-for-github/id1252320249?mt=8
#
[eddie] I’m not sure that worked. haha
#
Loqi rofl
#
[eddie] Thanks, loqi. Thanks
#
aaronpk needs 2 <<
#
Loqi you're welcome, [eddie]
#
[eddie] ah, gotcha
#
[eddie] GitHawk << https://itunes.apple.com/us/app/githawk-for-github/id1252320249?mt=8
#
Loqi ok, I added "https://itunes.apple.com/us/app/githawk-for-github/id1252320249?mt=8" to a brand new "See Also" section of /GitHawk
#
loqi.me edited /GitHawk (+88) "[eddie] added "https://itunes.apple.com/us/app/githawk-for-github/id1252320249?mt=8" to "See Also"" (view diff)
#
Zegnat I have my email client handle GitHub notifications these days: https://i.imgur.com/p0Ir8e7.png
#
aaronpk folders per repo?
#
Zegnat Those are auto-generated subfolders to a smart folder
#
aaronpk fancy
#
Zegnat MailMate settings, if anyone wants to copy this behaviour: https://imgur.com/a/q0DQ0
#
[cleverdevil] So, in theory, I should be able to backfill from Instagram to my website with granary.
#
[cleverdevil] Hey, [snarfed], does granary support paging for Instagram feeds?
#
[cleverdevil] I'd like to fetch my entire set of photos.
tantek, AngeloGladding and [snarfed] joined the channel
#
[snarfed] [cleverdevil] API yes, but not scraping... which is the only usable part for most of us right now :(
#
[snarfed] do they have export?
#
[cleverdevil] I found this: https://github.com/althonos/instaLooter
#
Loqi [althonos] InstaLooter: Another API-less Instagram pictures and videos downloader.
#
[cleverdevil] So, I've already downloaded everything that way. The only thing it doesn't pull is the caption.
#
aaronpk the caption is kind of an important part haha
#
Loqi aaronpk: lol
#
aaronpk oh dear, does this mean i'm not going to be able to make ownyourgram pull all the history?
#
[snarfed] oh paging via scraping may well be possible. i just haven't implemented it
#
[snarfed] but in general for backfill, converting an official export will usually be more reliable and easier to implement
#
[cleverdevil] Well, I've downloaded everything using the above already, now I can loop through the outputs, and use granary to pull the caption.
#
[cleverdevil] Easy peasy.
#
[snarfed] cleverdevil++
#
Loqi cleverdevil has 18 karma in this channel (71 overall)
#
[cleverdevil] Sigh, Known doesn't support Micropubbing photos using JSON.
#
[cleverdevil] I suppose I can just do it the ole fashioned way 😉
[kevinmarks] joined the channel
#
[kevinmarks] Only one photo per post
#
[cleverdevil] One photo per post, and you can't do it via JSON.
#
[cleverdevil] Awesome. Now, all 400+ of my Instagram photos are being copied to my website.
#
[cleverdevil] (With Syndication links this time)
#
tantek wow
#
[cleverdevil] 200 down, 200 to go 🙂
#
[cleverdevil] Well... 243.
[eddie] joined the channel
#
[eddie] Seems like a lot of scratching going on over there 😉
#
[eddie] *itch, itch, itch*
#
[eddie] lol
#
[cleverdevil] Heh, pretty much.
#
[cleverdevil] I stalled on the Facebook photo thing.
#
[cleverdevil] So, I wanted to at least knock out another service
#
[eddie] Makes sense!
#
[cleverdevil] Wheeeeee, all done 🙂
chrisaldrich and [snarfed] joined the channel