[07:45:35] <joakino>	 morning
[07:58:47] <phuedx>	 o/ joakino 
[08:02:50] <joakino>	 \o
[08:04:09] <phuedx>	 brb restart
[08:11:12] <phuedx>	 yup
[08:11:17] <phuedx>	 still restarting the other machine
[08:12:51] <phuedx>	 8 minutes remaining
[08:12:56] <phuedx>	 11 minutes remaining
[08:12:59] <phuedx>	 5 minutes remaining
[08:13:10] <phuedx>	 mac osx is the new windows
[08:35:06] <joakino>	 haha
[11:44:39] <phuedx>	 back from lunch
[12:15:17] <joakino>	 phuedx: http recording is awesome https://gerrit.wikimedia.org/r/#/c/284886/
[12:15:53] <joakino>	 makes all tests there run in 700ms
[12:19:13] <phuedx>	 joakino: neato!
[12:22:58] <phuedx>	 joakino: 40s -> 1s <3 <3 <3
[12:23:32] <joakino>	 in marketing&product world "Increased test running speed by 40x
[12:28:02] <joakino>	 it's beer o'clock
[12:31:49] <bmansurov>	 joakino: they usually say something like 4 million percent
[12:32:03] <phuedx>	 i'm watching cccen 32c3 talks and generating graphs
[12:35:41] <phuedx>	 joakino: did you see the core module graph?
[12:39:55] <joakino>	 phuedx: i think so, i found it pretty confusing
[12:40:05] <phuedx>	 because it's so big?
[12:40:14] <phuedx>	 (as is the mf one)
[12:41:58] <joakino>	 phuedx: more because it seemed a lot more horizontal and the entrypoints were utility libraries instead of *application modules* or *page entrypoints*
[12:43:59] <phuedx>	 i guess the clue is in the name: "core"
[12:43:59] <phuedx>	 ;)
[12:44:34] <joakino>	 not really, it is just the utility modules are added to entrypoints/pages in php :p (I guess?)
[12:44:51] <phuedx>	 not necessarily
[12:44:54] <phuedx>	 again, "core"
[12:45:00] <phuedx>	 but we can always check
[12:45:56] <joakino>	 grep the php sources for 'modulename'?
[16:49:36] <bd808>	 jdlrobson: I'm using the js from your chrome plugin via greasemonkey in Firefox and it seems to work pretty well -- https://github.com/bd808/userscripts/commit/80a721223bb50c18823d3f356a8bcd5e03202941
[16:49:41] <bd808>	 so thanks!
[17:02:43] <mdholloway>	 niedzielski: is ps2 of https://gerrit.wikimedia.org/r/#/c/284822/ just a rebase?
[17:04:52] <niedzielski>	 mdholloway:  yes, i just rebased that onto dbrant's menu patch
[17:11:54] <mdholloway>	 niedzielski: ok, cool.  i've rebased over https://gerrit.wikimedia.org/r/#/c/283671/ . 
[17:13:47] <mdholloway>	 one potential issue i've noticed is that when installing on a device i haven't used in a while, i got errors about not being able to find the database row readinglistpagedisk.filename -- had to clear my app data to do the upgrades and fix it
[17:17:27] <niedzielski>	 mdholloway: how long ago? remember we skipped the upgrade path for the filename column.
[17:28:10] <mdholloway>	 niedzielski: how long ago did i test with the device, or get this error?  the latter just a minute ago, the former i couldn't say.  but yeah, i remember about the filename column. 
[17:31:51] <niedzielski>	 mdholloway: i think it's probably expected if you haven't updated the app in a bit
[18:34:43] <maxbinder>	 dbrant: the cards in the feeds board are no longer prioritized, but as long as they are prioritized in the backlog it shouldn't matter
[18:35:12] <dbrant>	 maxbinder: cool, i'll make sure they are
[20:49:51] <mdholloway>	 all right!  the saved page service is hooked up and kind of working, and here's what i've found so far:
[20:50:41] <mdholloway>	 when creating a reading list for the first time, the page has a DiskStatus of OUTDATED and saves as intended
[20:51:12] <mdholloway>	 but unfortunately this happens whether the 'save for offline reading' slider is marked yes or no
[20:52:28] <mdholloway>	 all subsequent pages appear to enter the database with a status of UNSAVED, regardless of the 'save for offline' state of the reading list
[20:53:48] <mdholloway>	 and marking for for deletion seems to work correctly -- the service is finding rows with status DELETE as expected -- but based on limited testing right before retro it seems I'm getting null values from row.filename() for the deleted rows
[20:57:10] <dbrant>	 so then, action item: fix the DAO to give pages a status of ONLINE if the list isn't marked for offline use...
[20:57:30] <mdholloway>	 yes
[20:57:33] <mdholloway>	 also, a question:
[20:58:29] <mdholloway>	 when a reading list is edited, and the slider is moved to save for offline reading, that should kick off a job to save all pages in the list, correct?
[20:58:52] <mdholloway>	 and if the slider is moved back, should their saved files all be deleted?
[20:59:14] <mdholloway>	 or just left alone and not updated?
[21:00:23] <dbrant>	 when the slider is moved to the "on" position, the pages in that list will be marked as OUTDATED...
[21:00:37] <dbrant>	 and when the slider is "off", the pages will become UNSAVED.
[21:00:45] <dbrant>	 (and hsould be deleted from disk)
[21:01:39] <mdholloway>	 dbrant: ok, next thing i'll do then is fire up the service when that interaction happens and test whether it's behaving properly
[21:41:25] <mdholloway>	 next issue (i'm kind of just nothing them here as i find them): pages go from OUTDATED to ONLINE after saving rather than SAVED
[23:07:48] <mdholloway>	 https://gerrit.wikimedia.org/r/#/c/284984/ fixes all the problems i noted above except the deletion one so hopefully it's halfway sane in terms of the internal DAO logic
[23:15:29] <yurik>	 jdlrobson, around?
[23:34:02] <HaeB>	 https://twitter.com/designernewsbot/status/722792779961778177