[00:17] brousch: Yeah, we've all moved to pycharm [00:18] autocomplete FTW [12:18] G'morning [12:20] morn [12:56] Happy Friday [12:57] don't forget Penguicon "bring a friend" party tonight. With free-as-in-beer beer (I think) [12:57] eh? [12:58] there's a party going on somewhere? [12:59] Dearborn [12:59] oh heh to far [12:59] Penguicon is hosting a party at the hotel where the convention will be this year. [13:00] It's specifically aimed at getting people who haven't been to Penguicon to come out and get an idea what it's like. [13:00] So brousch, get out there and check it out :-) [13:00] too scary [13:01] http://www.penguicon.org/CMS/?page_id=216 [13:01] About: 404 [13:04] the about page should just redirect to the rules page [13:05] hmm this party mostly programers? [13:06] i get the impression it's mostly wackos [13:07] hmmm drag show lol [13:07] yep i c [13:17] I resent being called a wacko [13:19] You will find virtually any (and every) flavor of geek at P-con [13:19] some of them are even too strange for me... [13:19] Wolfger: mostly wackos, you might fall into the non-wacko segment. might [13:19] doubt it :-) [13:19] not that there's anything wrong with that [13:20] I fly my freak flag(s) proudly [13:20] i heartily support your right to be a wacko [13:21] what about a Yakko or a Dot? [13:21] or Pinky and the Brain? [13:21] oh wow, the BBB sent me an urgent notice using my own email address. i'd better jump right on this [13:22] lol [13:23] omg. someone is complaining about their dealership with me [13:25] Something about the phrase "multiple life sentences" just screams "your legal system is completely FUBAR" [13:29] seems like one life sentence should be sufficient [13:29] you would think [13:30] is it just me, or do the "oldies" on this list seem overpriced? http://shop.oreilly.com/category/deals/oldies-but-goodies.do?imm_mid=07ed2a&cmp=em-orm-books-videos-oldies-goodies-short-elist [13:31] DNS and Bind from 6 years ago for $20? [13:31] for an ebook? [13:32] also 2 ruby books, no python? [13:32] brousch: How many changes are you aware of in DNS? [13:32] Outside of DNSSEC, not much has changed. :) [13:32] and I'm not sure that's all that new. [13:34] hm, maybe you're right. it is $21 used on amazon [13:42] brousch: I'd expect Wolfger to be fielding dealership complaints sooner than I'd expect you to be fielding them. :) [14:42] just looked clojure syntax, all I can say is yikes [14:43] I can read it but I think using it for anything complex is just going to hurt [14:46] nullspace: ever played with Lisp? [14:46] It's very similar. [14:49] brousch: yes, everything on that list is overpriced [14:50] heh, the fact that they half off new books for DoTD, give any UG 35% off the top means that books are just overpriced [14:53] clojure is a lisp isn't it? [14:53] its scheme for jvm, right? [14:53] yup [14:54] "Clojure is a dialect of Lisp," from clojure.org [14:54] I don't think there's a "one true lisp" [14:54] right, that is what I'm saying. [14:54] It's like English [14:54] its a lisp [14:54] like scheme is a lisp [14:54] like elisp is a lisp [14:54] like common lisp is a lisp [14:55] Like MIT Scheme is not like GNU Scheme is not like ... [14:55] well you could argue that John McCarthy's 1960 paper defines the one true lisp [14:56] exactly! MIT scheme would be > GNU Scheme :) [14:56] Hah [14:57] Just seems whenever someone creates a LISP interpreter, they also create a new dialect to go with it [14:57] And we're not talking separate accents, but whole categories of slang. [14:58] right [14:58] and there are thousands, becuase its a project that 1/2 of all computer science students do at one point :p [14:58] jrwren: truth [15:09] snap-l: have yet to play with Lisp, though I am intrigued [15:11] when I see a new language that I understand but I don't see the reason for why it's doing things in a certain way I think that there has to be a good reasona nd maybe a godo use for it [15:11] otherwise why the hell was it created [15:12] nullspace: the enduring languages have purposes [15:12] except brainfuck, that's just there to screw with your head [15:12] sometimes languages are just created as a programming exercise. [15:14] I think I'll give clojure a spin through the first ten problems of project euler this weekend [15:24] <_stink_> and even within common lisp implementations, there are enough significant differences to be totally annoying. [15:24] <_stink_> especially with system interfacing stuff. [15:24] <_stink_> i like the competitive ecosystem... until i try to accomodate it. then i hate it. [15:28] _stink_: ++ ++ ++ ++ ++ [15:45] rick_h: did you see the nodejs video I retweeted? [15:45] i LOLed a lot [15:47] jrwren: will take a peek [15:57] i just created a custom django templatetag and corrected a django snippet. i think i am becoming assimilated [16:00] also, heroku is fun. `git push heroku master` and my updates are live [16:11] rick_h: it was orig a jwz tweet if you can't find it [16:27] jrwren: thanks, saw the tweet go by but didn't "understand" what it was so skipped by [16:28] What the fuck [16:28] Just discovered UNity has some keyboard shortcuts for tiling [16:28] Accidentally hit CTRL-ALT-NUmpad 0 and my window went fullscreen [16:29] CTRL-ALT-2,4,6,8 also move the window to that side [16:29] Hell, the whole thing is boobytrapped. :) [16:29] DBO: This is your doing!!!! [16:36] they're trying to make rick_h happy [16:36] heh, not quite. anything + a number is a horrible shortcut [16:37] Well, and it's only for a numberpad [16:37] ?! [16:37] so rick_h gets no love there either [16:37] so doesn't work if you dont' have a numpad on your laptop? [16:37] bah, wasted code imo [16:38] Heh [16:40] why do we have to have the most retarded setups for things? ugh! [16:42] ? [16:44] just getting annoyed at the way every little bit of code in the world we put together is strange and half working as #@$@#$@# [16:44] having a ranty friday [16:45] o.o [16:48] packages with buildout.cfg, but they don't work, then the README says to do this other thing, and rather than any sane test thing, find out we've built some magic https://launchpad.net/testrepository thing that this package uses [16:49] heh, and I see this is written because we've got giant ass test suites that day a workday to run: http://rbtcollins.wordpress.com/2010/11/30/testrepository-iteration-for-python-projects/ [16:49] a whole day? [16:50] that's insane [16:50] 4-6hrs [16:50] it's why we run the damn things in ec2, it's fire/forget [16:56] this is crazy traffic to my work website. i led a grwebdev meeting on the 23rd but never mentioned my work at all. http://dl.dropbox.com/u/101667/traffic_from_grwebdev_wtf.png [17:18] rick_h: no numpad on your keyboard? How do you live? [17:20] all day tests, fire and forget? No way. "Quit playing games and get back to work." "I am working. I'm running tests on the code." :-) [17:20] single-threading ftw [17:21] Wolfger: this is what pqm is for [17:22] Wolfger: and yes, I don't have a keypad on many of my keyboards <3 [17:23] I always wanted the numpad on the left, not the right [17:23] those keyboards are hard to find and/or obscenely expensive [17:24] numpad is bad. [17:24] not ergo. [17:24] mouse should be there. [17:24] if mouse is past the numpad it is too far and you have to stretch to far to reach it. [17:24] bad ergo [17:28] Why are you using a mouse? Just keep your hands on the keyboard. [17:28] oh come on Wolfger, I can't believe a kde user isn't moues heavy [17:28] if you truly didn't usea mouse much you'd be on a different DE [17:28] but yeah, that's why I want the numpad on the left... No reaching for mouse [17:29] exactly, which is why I like my 10less keyborads [17:29] keyboads [17:29] bah! [17:29] I give up...going home and calling in sick [17:29] ROFL [17:30] "Where's Rick?" "He was here earlier, then I got an e-mail that he wasn't feeling well and wouldn't be in today..." [17:39] snap-l: gah! your tweet and blog post... i still don't have anyone to open the door for you. [17:40] jrwren: No worries. Just LMK if it's not going to happen. [17:45] snap-l: crap, wife's working on that day :( [17:48] urgh [17:48] WEll that's no good. [17:49] I might be able to see if grandpa can watch him for an hour/two between me leaving and mom getting home since he's in town [17:49] but not sure [17:49] OK [17:49] I'll shoot for it, but might not be able to make it out. No offense, but not going to run out at 2pm when wife gets home [17:50] too much driving for too little time [17:50] Yeah, understandable [18:42] wow. got my second recruiter call ever at work. they want me to do iOS development [18:42] damn fools [18:45] https://plus.google.com/109919666334513536939/posts/a5DdsUPEvhH [18:46] heh [18:52] Oh geez [18:55] wat [18:55] come on, that's one cool looking michigan logo there :) [18:56] <3 [18:56] Waterloo's amuses me [18:59] ok, who wants to take this one? http://www.quora.com/What-can-vim-do-that-nano-cant?__snids__=33969307 [18:59] you win one internets [19:01] Oh jesus [19:01] That's like asking what's the difference between a woman and a Barbie doll. [19:02] Or the difference between a Navy Seal and a GI Joe action figure. [19:03] so ... you're saying vim can have real sex instead of just pretend? [19:03] "What's the difference between theory and practice? In theory, they're the same" [19:03] * brousch checks his vim book's ToC more closely [19:03] brousch: There's a plugin for that [19:03] gspot.vba [19:04] also, do not google gspot vim [19:05] based on those people i know who use vim, that is not an image search i want to encounter [19:12] snap-l: whoa http://www.microcenter.com/single_product_results.phtml?product_id=0242478 [19:13] Yeah, they're cheeeeeeep [19:13] Even 4GB is $5 each [19:14] i actually can't find any 1GB for under $2 each [19:15] i bought a 5 pack 2 years ago for $6 [19:15] That's still pretty cheap. [19:15] Yeah, I think they bottom out after a while [19:15] some how I missed this whole non blocking arguement the node.js fan boys are thrwoing around [19:16] nullspace: How so? [19:16] nullspace: huh? [19:16] that's the whole things with eventlet/node/etc [19:16] whole point that is [19:17] don't solve the concurrency problem, just work around it by making things not concurrent :) [19:17] ouch [19:17] Well, and use a language that is event based [19:18] or very good at events along with a seriously optimized engine [19:19] well the events just facilitate jumping in/out of the reactor/main thread [19:19] nullspace: so anyway, what's the issue or whatever with the realization? [19:20] I don't understand there arguement. did this make sense any time node.js came into being [19:20] yes, the idea is that all the libraries/etc you use are non-blocking so that whenever you make a call to mysql/etc [19:21] it happens outside the main thread allowing for another bit of code to be processed [19:21] in this one, one thread handles more work faster by getting anything that hangs cpu cyles out of the way [19:21] /this one/this way [19:22] for instance, in python, when I run a db query, it sits and waits for the db to respond, that python thread doesn't do any other work while waiting [19:22] but it could be serving someone else's cached result in that time [19:23] isn't there ways to write multi thread python? [19:23] yes, but there's a GIL you have to deal with [19:23] and it's really ineffecient for many many things [19:23] GIL being the Global Lock [19:24] and any time you get multi-threaded you have to think harder [19:24] (I being ... :) ) [19:24] see java multi threading/locks/etc [19:24] Global Interpreter Lock [19:24] ah [19:26] yeah that's not really a problem I can say I run into [19:26] nullspace: Well, you might run into it in small forms [19:27] but much like benchmarking different CPUs nowadays, you'll care a lot if you care a lot. ;) [19:27] otherwise, you won't care that much until it starts breaking things. [19:28] I think there are way bigger fish to fry [19:28] Yeah, like setting up Eclipse. ;) [19:30] hey, eclipse works just fine for me [19:30] anyone else off monday? [19:31] rick_h: Not I [19:32] Unless you work for the government, I don't think you get Monday off. [19:46] hmm well node could be really handy on embedded devices with a web frontend [19:48] nullspace: node is very handle for very fast little services and some people have built really large apps with it [19:48] handy that is [19:48] I did a contact lookup thing that could handle many more req/s on the one cpu than python could [19:50] the other handy space is the JS from back end to front end [19:51] lol, hilarious. [19:52] jrwren: which part :) [19:56] nano [19:56] ah, yea [19:59] nullspace: the nonblcoking arguement: see Twisted Python [20:00] or even stackless python [20:00] and maybe pypy greenlets [20:00] gevent, etc [20:02] rick_h: do you have an opinion on gunicorn/geven vs uwsgi? I was trying to wrap my head around that last night. [20:02] gunicorn is theoretically production-class, right? not just some little dev server? [20:04] Blazeix: supposedly [20:04] i'm using it to serve tens of requests per day on heroku [20:04] Blazeix: yea, but it's pretty bare bones [20:04] Blazeix: I'd still put it behind nginx/company [20:05] Blazeix: uwsgi has some extra added nice features [20:05] but I think gunicorn is winning the battle tbh [20:05] but one thing at morpace I wold them on, was that it could "shut down" an app, and reload it on first request [20:05] it doesn't really do static files, so you still need something else in front [20:05] so let's say you were deploying 50 wsgi apps, rather than 50 apps + 10 workers per app, sitting idle [20:05] it could actually shut down the apps to the 5 current accepting requests [20:05] and if #6 got a request, it'd take the time to load it [20:06] brousch: right, static files, ssl, caching, etc [20:06] Blazeix: and uwsgi is pure python, so no C libraries/compiling while gevent needs some C libs [20:07] the pure python is the most awesome part. pip install gunicorn [20:08] Blazeix: nvm...it's not pure python, I lied [20:08] but yea, it's pip installable, gunicorn is as well [20:09] it's not? [20:09] no, gunicorn needs some c compiled bits for the gevent support [20:10] damn, maybe I'm having a bad friday...don't see it [20:11] ok, cool. it kind of seems like gunicorn exploded on to the scene. [20:11] but maybe i'm just too far from the python community right now. [20:11] ah, ok so gunicorn is pure python ootb, but yuou can change the worker model to eventlet of gevent for performance [20:11] http://gunicorn.org/faq.html#worker-processes [20:11] and those are C-based, I thoght it used gevent ootb [20:12] Blazeix: yea, gunicorn is winning I think because it started out with built in django support [20:12] and has since really gotten helpers to make it a lot easier/performs well [20:12] and docs that don't suck as hard as uwsgi [20:12] I love uwsgi --help and seeing [20:12] -X "not documented yet" [20:12] for options in there lol [20:16] well `python manage.py run_gunicorn` is a lot easier than farting around with apache+mod_wsgi [20:17] yea, definitely [20:17] I see django 1.4 will finally have a .wsgi ootb now [20:17] i should say much easier than python+mod_wsgi+ virtualenv, because that's what really makes it a bitch [20:17] * rick_h rants on django some more [20:17] rick_h: 'bout fucking time [20:17] http://us4.campaign-archive2.com/?u=9735795484d2e4c204da82a29&id=e9564edf16&e=6bda5f785a [20:18] we don't need no stinking .wsgi [20:18] * rick_h smacks brousch upside the head with a cluebat (no baby sticks here) [20:19] but yea, that's why gunicorn got a boost, it knew django users didn't wsgi ootb so it made it easy up front [20:19] while everyone else just supports wsgi dammit [20:19] There's no reason not to support wsgi [20:20] None. Zero. Zip. [20:51] man precise downloads so painfully slow :/ [20:59] bittorrent? [21:07] rsync daily from testdrive :[ [21:11] anyone have an opinion on the "Agile and Beyond" conference that's coming up? [21:12] my work wants me to go it, but for some reason i have a bad taste in my mouth about it. [21:12] i guess i'm worried it's going to be bunch of Big-A Agile people. [21:12] What's the worst that could happen? [21:13] A little Kool-Aid, some sneakers, and a loooong nap? [21:13] my brain could leak from my ears. [21:14] Blazeix: yea, I couldn't sit through it [21:14] Agholes? [21:14] Blazeix: Searls gave the talk at detroit dev days on jasmine [21:14] but I mean side v side kanban talks? rly? [21:15] heh, the improv talk from detroit dev days is going on there again [21:15] i've only ever used Kanban as a punchline, which I fear wouldn't go over well there. [21:15] hah [21:16] Blazeix: yea, it sure seems like a mgr thing than a code in an editor thing :/ [21:16] exactly [21:35] as i work on this project i find i'm collecting a bunch of little bash scripts for repetative tasks, like refreshing static files locally and on the server. is this where i should be using fabric? [21:39] If you're repeating a script, yes. [21:39] If you have to keep a list of what script to run next, yes. [21:52] ok [22:01] brousch: fabric or Makefile [22:01] I'm coming around to the Makefile way of life [22:01] thuogh still <3 fabric [22:17] uh oh. :) [23:09] Loaded up the good reads app on my nook [23:09] I'm a little disappointed with it [23:10] would be nice to have it automatically grab titles from the device [23:10] instead, it's essentially the website in app form [23:41] snap-l rick_h sort of good news http://paste.mitechie.com/show/542/ [23:49] brousch: ah, nice