#dev 2016-08-30

2016-08-30 UTC
miklb, thebaer, KevinMarks, ben_thatmust, KevinMarks_, mblaney, cweiske_, doesntgolf, tantek and cweiske joined the channel
#
cweiske
initial crawling for chat.indieweb.org will take years :/
#
aaronpk
whoa what! how long is each request taking?
#
cweiske
1 second it seems
#
cweiske
and it's just crawling, not indexing
#
cweiske
it crawls every single chat log line
#
cweiske
which are many
#
cweiske
and then, when it's done with that, it'll index them
#
aaronpk
hm that doesn't seem right. it loads in less than 1s in the browser
#
cweiske
i'm running 10 parallel instances now
#
cweiske
aaronpk, maybe it's internal processing that takes time
#
aaronpk
could be. also http request overhead.
#
aaronpk
it supports http2 so if your client supports that you'll save ssl overhead
#
cweiske
gearadmin --status
#
cweiske
phubb_notifysubscriber 0 0 3
#
cweiske
phubb_verify 0 0 3
#
cweiske
phinde_crawl 784 10 10
#
cweiske
phubb_publish 0 0 3
#
cweiske
phubb_cleanup_pingrequest 0 0 3
#
cweiske
phinde_index 1859 0 10
#
cweiske
I guess I should prioritize indexing before crawling
#
cweiske
and I need a way to disable crawling of log line urls
#
cweiske
they won't give new links that I need
#
cweiske
ok, I'll stop that for now until I fix those issues.
#
miklb
  
cmal, kline, thebaer, tantek, KevinMarks and cweiske joined the channel
#
cweiske
aaronpk, we need to fix the log line page titles
#
cweiske
.. or I need phinde to always display some content extract, not only when it matches the search terms
#
cweiske
it's now slowly crawling everything
#
cweiske
hm. some config error somewhere. only crawled today
#
cweiske
more tomorrow.
#
tantek.com
edited /Instagram (+27) "/* Single Photo Bookmarklet */ update to work for multiple ways of navigating to a permalink, from recent photos homepage or directly"
(view diff)
#
aaronpk
Fix the title?
#
aaronpk
Not sure I want the title tag to be the chat like content
#
aaronpk
seems like it would be hard to get that to look right in all cases
#
aaronpk
showing a summary of the page content even if that's not a match of the search seems reasonable. Pretty sure Google search results always show something in the content summary
tantek, ChrisAldrich, gRegorLove and rMdes joined the channel
#
rascul.xyz
created /geocoding (+687) "create page and list some services"
(view diff)
#
rascul
what is geocoding?
#
Loqi
Geocoding is the process of turning a text form location (such as an address or name of a place) into a set of geographical coordinates https://indieweb.org/geocoding
#
rascul.xyz
edited /Nominatim (+4) "link to the geocoding page"
(view diff)
#
rascul
doesn't esri have a geocoding service somewhere?
#
aaronpk
developers.arcgis.com
#
rascul
ahh ok i'll add it
#
rascul
eh you're probably better qualified to list it aaronpk
ChrisAldrich joined the channel
#
rascul.xyz
edited /geocoding (+59) "/* Geocoding Services */ list esri"
(view diff)
#
gregorlove.com
edited /geocoding (+162) "/* Geocoding Services */ geonames.org"
(view diff)
#
gRegorLove
What is geocode?
#
Loqi
It looks like we don't have a page for "geocode" yet. Would you like to create it?
#
loqi.me
created /geocode (+22) "prompted by gRegorLove and dfn added by gRegorLove"
(view diff)
#
rascul
what is geocode
#
Loqi
Geocoding is the process of turning a text form location (such as an address or name of a place) into a set of geographical coordinates https://indieweb.org/geocode
#
rascul
nifty
#
rascul
is learning how to be a wiki gardener
cmal joined the channel
#
rascul
pretty soon i'll have drank enough beer to work on my site
#
rascul
once i implement checkins then i'll be ready to have a real site instead of the placeholder i have
#
aaronpk
Everyone is beating me to checkins :-P
#
rascul
the only reason i care about geocoding is so ensure i have a set of coordinates to display a map
#
aaronpk
That's a good reason to care about geocoding
#
rascul
indeed
#
rascul
openstreetmap can never seem to parse the addresses i give it
#
rascul
the us census geocoder parses my addresses just fine and doesn't require an api key so i'll pprobably use that
#
gRegorLove
It's not normal/correct to nest XML inside CDATA, right?
#
gRegorLove
This RSS feed I'm parsing has:
#
gRegorLove
<description><![CDATA[Publication date: <pubDate>October 2016</pubDate>...
#
Zegnat
XML inside CDATA is no longer XML. So. No. Not correct.
#
Zegnat
Though it is pretty normal to put HTML inside CDATA inside feeds. So it is easy to get confused.
#
aaronpk
Maybe they wanted the literal < to show up as the description
#
gRegorLove
Yeah, I figured they just messed up.
#
gRegorLove
<pubDate> is surely meant to be an XML element in the feed.
#
Loqi
rss has -1 karma
#
rascul
it's certainly not an html tag heh
#
gRegorLove
This is an academic journal, though, so good luck getting them to fix it or adopt mf2 :D
#
gRegorLove
Also weird, each <item> has <feedDate>, and the date is the same for each item.
chrisaldrich1 joined the channel
#
GWG
Afternoon
tantek, miklb and kline joined the channel
#
tantek.com
edited /Instagram (+33) "/* Single Photo Bookmarklet */ update prompt text instead of "Photo" placeholder"
(view diff)
#
bear
techinically you can put *anything* inside of CDATA as it is defined to be where you put data that is not to be parsed as XML
#
gRegorLove
Ok. That's what I thought and was meaning to confirm.
tantek, chrisaldrich1 and AngeloGladding joined the channel
#
aaronpk
ben_thatmustbeme: are you using Cordova still?
#
aaronpk
Google released an SDK for OAuth for Cordova and iOS and android and I think it could be used for IndieAuth login for Micropub. Need to try that out.
KevinMarks joined the channel
#
www.boffosocko.com
edited /publics (+414) "comparison with author identies"
(view diff)
KevinMarks_ joined the channel