2013-05-18 UTC
andreypopp, duckbillp, josephboyle, xtof, tilgovi, fmarier, b0bg0d and legastero joined the channel
duckbillp joined the channel
duckbillp, andreypopp and xtof joined the channel
duckbillp, b0bg0d, cweiske, tommorris and xtof joined the channel
duckbillp joined the channel
# 11:28 aaronpk ah cool! that's what I want to add next to pingback.me
# 11:28 aaronpk one thing I encountered on my site is I don't want to send pingbacks to *all* links on the page, just the ones inside the h-entry
# 11:49 cweiske do you have more content in your feed than the h-entry?
andreypopp joined the channel
xtof, barnabywalters and andreypopp joined the channel
brianloveswords, xtof and pdurbin_m joined the channel
# 15:49 pdurbin_m if ubuntu is invading privacy
# 15:56 pdurbin_m sending data about you into their cloud
xtof, seyz, spinnerin and peck_lx joined the channel
scor joined the channel
# 19:17 aaronpk cweiske: oh you meant my atom feed. I consider that deprecated at this point :)
# 19:32 aaronpk I want to encourage more use of microformats2, so anything new I develop is going to be based around my h-entry feed, not my atom feed
xtof joined the channel
# 19:36 cweiske so you really ditch an established, wide-supported feed technology for a json file?
# 19:37 aaronpk not a json file, microformats markup. I just happen to provide an already-parsed version of it at a URL
# 19:44 cweiske do you expect feed readers to support mf2 in near future?
# 19:44 Loqi cweiske meant to say: do you expect feed readers to support mf2 in future?
# 19:52 aaronpk i expect whatever replaces feed readers to support mf2. I think we're already seeing an extreme decline in the use of feed readers, especially now that google killed Reader
# 19:54 aaronpk facebook and twitter have proven that their version of a feed is liked more by people than traditional RSS readers
scor joined the channel
gRegorLove joined the channel
# 20:48 aaronpk thought about trying to do a HA setup with it, but it's too much work
# 20:50 tantek those of you showing original context on your reply posts - how are you storing original context in your storage?
# 20:50 tantek barnabywalters and aaronpk - you might be the only ones showing context on your replies
# 20:50 aaronpk tantek: I switched my storage to store both the raw HTML as well as the parsed JSON
# 20:51 barnabywalters tantek: I’m flattening the microformats and dumping it as JSON into storage. Soon my mentions will store both source HTML and flattened µf
# 20:51 tantek right that is what we discussed last time aaronpk
# 20:51 aaronpk yep. i'm not doing any parsing other than what php-mf2 does
# 20:52 tantek so entire source of the original? no filtering of scripts or other potential nasties?
# 20:52 tantek just treating it like a webcache snapshot - that makes sense
# 20:52 tantek so it would purely be for reparsing later as we discussed
# 20:53 aaronpk oh hey that's nice that he starts up automatically when this machine reboots
# 20:55 tantek first storing only the pieces of original context I need to create the original context display
barnabywalters joined the channel
# 20:57 aaronpk decide what to do with the original later, but keep it for now. it's not much extra code to do that since you already have to fetch the page to parse it
# 20:57 tantek are you storing HTTP headers returned as well?
# 20:59 tantek might be useful for retrying pingbacks/webmentions etc.
# 20:59 tantek and if we can/do store something warc-compatible, then we start archiving interesting (worth replying to) bits of the web in a very distributed manner
tantek joined the channel
# 21:02 tantek cweiske - when you say "do you expect feed readers to support mf2 in near future?" which specific feed readers are you thinking of? (since they seem to be dying off themselves)
# 21:25 tantek well without a specific example, it's perhaps not worrying about
# 21:27 aaronpk so how do we create more incentives for the large players like google/facebook/twitter to parse microformats-2 data?
# 21:27 tantek aaronpk, for larger players it's generally a simple formula of amount of deployment compared to cost of implementation
# 21:28 aaronpk do you think they are likely to take an existing open source parsing library into account in the cost of implementation? or would they rewrite a parser? or is that not significant?
# 21:29 tantek but rather than attempting to directly influence larger players (beyond casual suggestions to standards-friendly advocates inside their walls), it's likely to be more effective (short and long term) to influence up the long tail of sites, from smallest to largest.
# 21:30 tantek as we build a critical mass of interoperable sites and implementations, it naturally becomes more and more compelling to larger and larger players
# 21:30 aaronpk so then the challenge is showing the long-tail the benefits even when the major players don't support it
# 21:30 barnabywalters you guys seriously how can you not be watching eurovision right now I dont even
# 21:32 tommorris tantek: no, it's basically a grand unifying European-wide movement that brings together white-van-man racists, cynical Twitterers and very camp fashionistas to share their mutual distaste for the French.
# 21:32 barnabywalters tommorris: wow, you managed to sum up eurovision in a paragraph. it took me ages to explain it to brennan
sandeepshetty joined the channel
# 21:58 sandeepshetty barnabywalters: remember reading somewhere that you were moving to a file-system based store for Taproot?
# 22:00 barnabywalters so I store data in files with an ID, then store things about that file I want to query (e.g. tags) in the DB
tantek joined the channel
# 22:02 barnabywalters sandeepshetty: not really — I like being able to do complex queries over my data, so I decided to keep the DB in there
# 22:03 aaronpk I don't have any way to do complex queries right now. my site search goes to google
# 22:04 sandeepshetty aaronpk: so you have post organized by time and then a parallel hierarchy for tags?
# 22:05 aaronpk if you click a tag, it loads the tags.yml file and grabs all the filenames from there. I rebuild the tags file when I make a new post
# 22:06 tantek oh hey - welcome back sandeepshetty, nice to see you here
# 22:09 tantek didn't see anything in IRC for about 20 minutes after "how would we fit Disqus on the map?"
# 22:11 barnabywalters aaronpk: Loqi seems a little more talkative recently — have you been working on him?
# 22:11 aaronpk although his talkativity level does depend on the amount of traffic in other channels
# 22:12 tantek barnabywalters - can't reference it while the server is down
# 22:13 aaronpk uplodaed those from my phone...haven't seen that page since
# 22:13 tantek aaronpk - yeah - your uploads don't seem to scale well in the current mediawiki CSS
# 22:17 barnabywalters hm, an indieweb + silo islands map would be a really cool thing to generate automatically
# 22:27 aaronpk would be interesting to automate the indiemark score check
# 22:29 tantek aaronpk - I'd rather indiemark be meaningful than automated
# 22:29 tantek automation typically results in superficial gaming
# 22:30 aaronpk but there's a couple things that would be easy to check for
# 22:43 aaronpk fyi I'm migrating more servers, stuff will be down for about 30 minutes
# 23:00 tantek potentially sharing such warcs with ArchiveTeam or the Internet Archive if/when such sites go down
# 23:00 tantek is tempted to store a copy of any page I link to for that reason.
# 23:00 aaronpk that's not a bad idea. I haven't done anything with warc files before... any good getting started guides?
# 23:03 tantek ok well if anyone finds non-ISO documentation of how to construct a WARC file
gRegorLove and xtof joined the channel