#ZegnatHmm, not sure if I like how IndieNews lists articles based on when they were submitted rather than when they were originally posted. I could totally see how more recent blog posts could be drowned out by someone adding a backlog of posts like pfefferle just did.
#voxpelliOn the other hand it's hard to find newly submitted posts if they aren't added on the top :/ Have that problem with FeedBin – lots of "new" blog posts are added among old ones I've already receives – making them easy to miss
mdik_ joined the channel
#voxpelliTricky that – I have on my todo to enable webmentions from my endpoints to be listed by receive date in addition to the claimed publish date it has now
#ZegnatYeah, I guess that could be problematic. But for some reason I would expect to see the latest news on news.indiewebcamp.com, not "old" news
#ZegnatHow do the old-school "planet" sites do this?
snarfed joined the channel
#ZegnatNote to self: check planets after dinner (planet mozilla, planer firefox, planet performance, come to mind)
mdik, eschnou, misuba, wrk3, shiflett and nitot joined the channel
#ZegnatHmm, it seems to me that most planets work based on RSS, and then sort by publishing date, but not sure :(
#voxpelliWhen they add a new feed they probably don't import retroactively – which results in only getting new posts
#[aaronpk]Zegnat: re above, it's because of the clustering, it's not practical to include more than one post name all the time. Also some post names are really long, so that's awkward.
#rhiaroThis is driving me mad. Is everything an edge case? There's no consistent way to know where to look for stuff? (aside from representative h-card thing)
#rhiaroI didn't realise expanding my reposts was going to be such a PITA
#rhiaroThis was supposed to be a quick thing I could do so I felt like I accomplished something
#rhiaroWill just assemble a list of people whose markup I definitely don't want to deal with and never repost them :)
#ZegnatYou should document as many edge-cases as possible on the wiki, rhiaro. I think people don’t mind changing their markup a bit if some consensus can be reached
#rhiaroFinding authors has been reasonably well thought out, so it's not so bad, but I still have to write a bunch of really boring author-finding code
#gRegorLoveWhich language? I can share my PHP authorship code if it helps
#rhiaroBut there are other things too. People nesting things in feeds differently. No specific example from what I'm doing right now, but ran into this before a while ago. I think I was trying to find the most recent post on any homepage that had a location. Had to jump through so many hoops and look for so many cases
#kylewmZegnat: you joke, but technically compliant markup that exercises the edge cases is pretty much tantek's M.O.
snarfed1 joined the channel
#Zegnatkylewm: so what you are saying is I should aim to give tantek a headache?
snarfed and [kevinmarks] joined the channel
#[kevinmarks]yes, Tantek made my hovercard code blow up a lot too (he has a list of orgs, one of which is an h-card)
#[kevinmarks]but maybe as well as consistent parsers across language implementations we need consistent utility functions for 'get representative h-card, get author , is this a note? etc
nitot joined the channel
#Zegnat[kevinmarks]: interesting idea. So, formalise the algorithms for finding information?
wolftune, KevinMarks, keroberos and [kevinmarks] joined the channel
#[kevinmarks]we kind of do that on the wiki already; so doing it with examples and testcases seems a logical extension
moredhel and wolftune joined the channel
#M-Kodoformalising the algorithm would be very useful
#M-Kodothe whole here is the html here is the json is like, here are two circles, now draw the owl