[00:45] applications for GSoC starting to roll in. Man this is going to be so hard to pick 2 [01:50] lol did you see Hubert? [01:50] little late to the party [02:01] i almost feel bad [02:29] .func [02:29] Searches most recent copy of bookie repo for function definitions; Update repo with .pull; Syntax: .func [02:29] .help func [02:29] Searches most recent copy of bookie repo for function definitions; Update repo with .pull; Syntax: .func [02:29] .help list [02:29] no docs for command [02:29] .help [02:29] echo func help list pull relist sleep [02:29] .help asdf [02:29] no such command [02:37] jcastro: http://www.theregister.co.uk/2014/03/15/windows_desktop_and_laptop_market_share_dips_below_90_per_cent [02:37] About the only thing I got wrong in our bet was the timeframe. ;) [02:37] That and not counting iOS. :) [02:38] below 90% [02:38] i like how below 90% is considered a huge problem [02:38] Our bet was Mac marketshare at 10% [02:38] everyone else considers that a monopoly [09:59] morning [09:59] ugh [12:11] morning [12:12] and +1 ugh, but got to sleep in a bit so yay [12:14] Just a bit? [12:15] Yeh, acc. to the fitbit I got 5h30m sleep [12:35] morning [12:35] rick_h_: how was GR? [12:44] GR is awesome with every metric you choose [12:45] mrgoodcat: good stuff. Parking was meh but we had a good day [12:46] brousch: are you from gr? [12:47] West MI in general [12:47] i'll agree with you all day on that [12:48] rick_h_: I've never seen that parking ramp full except during ArtPrize and museum new exhibit openings. It was really odd for it to be so busy on the 3rd weekend [12:48] i love traverse city, kalamazoo, and grand rapids [13:06] woot! pocket casts now with chromecast support [13:07] <3 the world we live in [13:08] chromecast ias been really bad to me [13:08] why so? The only issue I find is that the chrome extention has to be disabled/enabled every so often as it loses the connection to the chromecast [13:08] mrgoodcat: Did it kill your good cat? [13:09] i had a really hard time with initial setup [13:09] and occasionally it will lose connection to my wifi [13:09] It has worked very well for me [13:09] the chromecast extension is total garbage [13:09] But i use it almost exclusively from Android [13:10] and the sound cuts out occasionally [13:10] i use it exclusively from android and chrome os [13:10] you would think it would work the best [13:10] it just feels like a beta product [13:10] Also I basically just use netflix and youtube [13:10] youtube is the worst for sound cut out [13:11] i use it for netflix, youtube, and google play movies [13:12] I've had no sound problems [13:12] mrgoodcat: How far away from your wifi is the Chromecast? [13:12] Played 2 different 1.5 hour long youtube videos yesterday [13:12] about 20 feet [13:12] accross the same room [13:13] line of sight [13:13] N? [13:13] yup [13:13] i even switched routers [13:13] i have a netgear r6300 right now [13:13] but i had a belkin before [13:14] does it go through the TV to get to the router? [13:15] sort of but not really. its on an angle [13:16] about 45 degrees [14:07] Trying duplicity again for local backup. [14:07] I think the biggest problem I have is trying to backup 10lbs of shit in a 5lb bag. [14:08] But did decide to turn off journaling on the backup drive as I think it was rather unnecessary. [14:21] duplicity is actually really good [14:22] rick_h_ has the patience of a saint [14:22] no way i could be a GSoC mentor [14:23] not that i believe anyone would actually want me to mentor them [14:30] meanwhile on the set of Ocean's fourteen http://bit.ly/1md2O9A [15:14] rick_h_ is a manager. You can call him Mister Manager. [15:14] huh? I'd not recommend it :P [15:19] my only problem with Duplicity is it takes forever to complete [15:20] cmaloney: my first sync to my NAS took 4 days [15:20] it's what it takes to start a backup [15:20] rdiff-backup is much quicker [15:21] but it also tends to lunch itself if the dest volume fills up [15:21] obnam! [15:21] which seems to happen to me if I use VBox. The image files change enough to make things not want to work right [15:21] greg-g: obnam is fine save for a few problems: [15:22] 1) I've never seen it finish. [15:22] 2) See above. [15:22] cmaloney: finishes for me [15:23] cmaloney: did you follow the guide and start with a subset? [15:23] obnam backup $HOME/somesubdir [15:23] greg-g: I had one version when I tried it that was completely borked. [15:23] No, I tried the whole hog [15:23] * greg-g shrugs [15:23] works great for me [15:23] It was before 1.0 released. [15:23] a incremental now takes about 10 minutes [15:24] speed isn't it's strong point, though [15:24] Right. I need something that's a little quicker than Obnam [15:24] also space efficient [15:24] obnam does dedup [15:25] also: could be conflating obnam with duplicity. [15:25] heh [15:25] I don't know the others too well anymore [15:25] other than "rsync" [15:25] ;) [15:25] Yeah, rsync is awesome [15:25] I swear if rsync stopped working I think I'd seriously have to consider not jumping off of a bridge. [15:25] if you don't want versioning, it Just Works [15:26] I used to use rsnapshot. Liked it a lot but rdiff-backup was more efficient [15:29] rick_h_: you made me chuckle out loud in the office: https://twitter.com/shmcmahon/status/445329656602820608 [15:29] such great stuff http://docs.python.org/3.4/whatsnew/3.4.html [15:29] greg-g: <3 [15:31] time to move to python3 [15:31] rsync isn't backups. [15:31] if you delete a file, and rsync runs and deletes the synced files, you cannot recover. [15:32] or rather... rsync MIGHT be a backup - if you have a stupid recovery specification [15:32] jrwren: that [15:32] well, there's different kinds of backups. Full versioning isn't always necessary and if you don't do -e you can rsync and keep old files around [15:32] but they're the latest version [15:32] and you are backing up multiple locations right? [15:32] I've been bitten by the "new version wasn't right" bug [15:32] i like to say - there is no such thing as backups. There is only restores. [15:33] greg-g: yea, but that's another level of 'data lost' [15:33] jrwren: Deep maaaaaaaan [15:33] luckily I could go through git-annex's previous versions [15:33] tar -cf - / >/dev/null # is a backup [15:33] OK, I do not like git annex. [15:33] I tried it for my book collection and hated it [15:33] i neither like nor dislike git annex. [15:33] cmaloney: which part? [15:33] greg-g: The "files are not real files" part. [15:34] the "syncing takes longer than rsync" part [15:34] ah, so, "git-annex direct" will make you happy [15:34] the "holy shit I think I just did something stupid and now I have to restore from backup" part. [15:34] that part. [15:34] cmaloney: there's this pattern I see in you: you want speed at the expense of well written software that guards against data loss ;) [15:35] greg-g: Point [15:35] but seriously, git-annex direct makes the files real files, but you loss a ton of git-annex's safe guarding [15:35] I think the main part was I didn't understand what it made it tick [15:35] and that ultimately wasn't what I had in mind. [15:35] http://git-annex.branchable.com/direct_mode/ [15:36] Yeah, familiar with it [15:36] but what I really wanted (in this case) was a way to sync my books to my laptop [15:36] yeah, I've gotten to the point where I know how it works (mostly, not all) and thus I grok it, and when grok something, you love something ;) [15:36] and rsync does that nicely [15:37] greg-g: http://www.youtube.com/watch?v=SS516maqfAg woot looks like it's up [15:38] w00t! [15:39] it's a talk I should try to give again. I could use a few iterations on it. [15:42] * greg-g is listening [15:44] lol i love the description [15:44] "he's been a team lead for about 10 minutes" [15:44] ;) [15:45] talk the talk with a giant salt lick [15:51] and a shot of tequila [15:56] rick_h_: when are your daily standups? time of day? where does your team live? [15:56] greg-g: so ours is 11am est. We have folks in italy, EDT, and mountain time in CO and Canada [15:57] greg-g: so 9am for the western folks, after lunch for guy in italy [15:57] * greg-g nods [15:57] we have a senior dev in Australia [15:57] kinda fucks everything over [15:58] greg-g: yea, we've got a design guy in AU. So I'll meet with him at night as needed and we tend to have more of a weekly catchup with him [15:58] greg-g: but we track his work on the board [15:58] * greg-g nods [16:03] rick_h_: slides with links to your boards? (sorry for interrupting, but you're a manager now, you HAVE to multitask) [16:03] :P [16:03] greg-g: :P [16:03] greg-g: the big work board isn't publicly avail [16:03] greg-g: I'd be happy to do a hangout/screenshare if you want a walk through/questions [16:03] but it's behind a login [16:04] * greg-g nods [16:04] understandable [16:04] https://trello.com/b/jXSwmBMC/bookie [16:04] and shared the slides doc with you [16:05] ty sir! [16:05] interesting! [16:05] the requirements on code review based on diff size [16:05] we don't have such crazy rules like that [16:06] it helps break code down and prevent massive diffs that talk days to review [16:06] and I think is a really really good thing [16:06] yeah, I like it [16:06] helps break that 3 day card to two 1 day tasks [16:06] we don't have too much of a problem with that, but it does happen [16:06] we have a huge rewrite of one of our extensions just sitting there because, well, it's a rewrite, all at once [16:07] yea, sometimes we have to do more work to do things behind feature flags and such but it makes things not get hung up for so long and impossible to merge [16:07] trade wasted time in making things landable as you go vs removing that cruft once it all does land [16:08] totally [16:08] people need to get more comfortable with feature flags [16:08] <3 [16:08] rick_h_: does whatever ya'll use for code reviews enforce that, or is it team enforced? [16:08] we've got a few good lessons on using those. Working in the negative case vs the positive for 'is the flag on' to aid cleaning up a closed flag [16:09] greg-g: team enforced [16:09] * greg-g nods [16:09] once in a while we let one go because of the nature of the work [16:09] yeah [16:09] or things like "mechanical change, moved x to y" [16:09] a tool enforced thing would need an override feature [16:09] yeah [16:09] but nice because you need break the work from mechanical vs actual changes in a big branch [16:09] huh, yeah [16:13] whoa! pep442 finally makes python GC not crazy! :) [16:56] I heard one anecdote from a developer that when he worked at $COMPANY they had an onerous process for landing program changes into production [16:57] so the developers learned to make all of their code work via the database instead. [16:57] because that was less controlled than the codebase [16:57] brings a whole new meaning to "data is code, code is data" [16:57] yea db needs the same process [16:57] Their code essentially became a boot-loader for code in the database [16:57] lol [16:58] I'd rather the lesson be that you can get to the point where your process is more important than actually getthing things done. [16:59] (but yeah, your database should also have some rigor to it) [16:59] oh good: http://www.google.com/appsstatus#hl=en&v=status [17:00] well the trouble is that is takes a lot of time to get the process right [17:00] also, yay depending on a service without an SLA [17:00] until you've got the smooth running process changes tend to be big and infrequent which just makes them harder [17:00] but until you get it smooth you won't get small frequent changes that are more ideal [17:00] greg-g: yea :/ [17:00] we'll see if I'll be doing my hangout interviews this afternoon [17:00] guess who has a Ops-wide team meeting over hangouts soon? [17:01] rick_h_: Agreed, but developers tend to see processes that are too strict as damage and route around it. [17:01] cmaloney: yea, understand [17:01] ++ for creative solutions, but their code suffered as a result. [17:02] We saw that with FieldConnect and their backward waterfall process. [17:04] It was a challenge to get anything promoted to production so it felt like a major accomplishment to land any change. [17:05] right, the goal of any process should be making it faster and easier to land things safely [17:05] much like figuring out the controls to an arcade game. [17:09] It should be this simple: http://www.tokensonly.com/images/events/morgan-beckman/images/morgan-b-gameroom-40.JPG [17:10] (That's Badlands, a laser disc game that had one single button). [17:10] (Note: The game was terrible). [17:11] http://www.emuparadise.me/MAME/cabinets/thayers.png <- This was our launch process. [17:15] Also: Thayer's Quest was interesting, but not terribly friendly to newcomers. [17:57] "Since GateKeeper is essentially a runtime business rules engine, it was heavily abused to effectively execute code living in a database. Avoid this through simpler design or a policy of sanity." [17:57] heh [17:57] from https://secure.phabricator.com/book/phabflavor/article/recommendations_on_branching/ [18:02] :) [19:25] greg-g: re: Techcrunch article on harassment: I take most things from TechCrunch with a grain of skepticism. [19:26] But even if 70% of what is in that article pans out, that's reprehensible. [19:26] (was re: http://techcrunch.com/2014/03/15/julie-ann-horvath-describes-sexism-and-intimidation-behind-her-github-exit/) [19:45] maybe i should ad a rickstatus command to bookiebot lol [19:45] mrgoodcat: :P [19:45] seems that way "eyeballs averted don't type out 10 things" [19:50] i was considering making it so if a new person joined that the bot had never seen before it would greet him with some sort of message linking to the docs and such [19:51] maybe after i finish the switch to sqlalchemy [20:02] nah, just have it display that every 5 minutes. ;) [20:03] cmaloney: reprehensible! [20:03] lol [20:03] cron job [20:04] i'm rewriting my go bot as well [20:06] mrgoodcat: https://bmark.us/craig/redirect/7f0f9f24bb793f [20:06] (that's what I think of whenever I hear 'go bot' ;) [20:10] interesting idea... maybe i should write a bot in go to play go [20:10] i've made a few sudoku solvers [20:20] jrwren: w...t...f? 3252 queries? [20:20] queries on what? [20:21] bmark? [20:21] twitter [20:21] which only takes 7s, which i'm totally OK with. [20:21] no, we don't do that anywhere [20:21] cmaloney: yeah, github has a culture problem. always had. They just now (Jan '14) hired a head of HR. That shows a real problem. [20:21] i'm confused [20:21] jrwren: ugh, any web thing taking 7s needs to be shot unless it's a one off admin only view [20:21] assume its not web. [20:21] still, that'll drop your job [20:21] jaw [20:21] bah [20:21] I'm done for today, wtf [20:21] rick_h_: don't load the Barack Obama article on WP (slow because of how well cited it is) [20:21] it's 4:20?! [20:21] still, the 70s total is what I find unacceptable. [20:22] holy crap, /me missed the last 2 hours [20:22] rick_h_: spark it up! [20:24] OK, did anyone follow anything that rick_h_ and jrwren said since 4:20? :) [20:24] nope [20:25] context https://twitter.com/JayRWren/status/445654479648460800 [20:25] I'm a little bit less crazy than I seem...usually [20:25] What the fuck? [20:26] lol [20:26] jrwren: why are you making so many queries on one webview? [20:26] What the hell are you loading? [20:28] whelp, that's the end of my workday [20:29] Yeah, same here. [20:29] ttyl! [20:29] new OMC tonight with any luck. [20:33] this thread makes me sad: https://twitter.com/faidonl/status/424108562579591168 [20:33] (Faidon is an Ops engineer at WMF, also a DD (paravoid)) [20:50] greg-g: :( [23:49] Evening [23:50] So basically what Varnish just did was give everyone a reason to not use Varnish.