00:00:00  * ircretaryquit (Remote host closed the connection)
00:00:08  * ircretaryjoined
00:02:25  * benoitcquit (Excess Flood)
00:04:27  * benoitcjoined
00:07:56  <MI6>joyent/node: isaacs master * ec74555 : benchmark: Set ephemeral ports properly on sunos (+5 more commits) - http://git.io/JoKNbw
00:11:09  * bradleymeckquit (Quit: bradleymeck)
00:36:51  * karupaneruraquit (Excess Flood)
00:38:00  <isaacs>sblom: any progress?
00:38:03  <isaacs>sblom: i'm still lost.
00:39:48  <MI6>joyent/node: isaacs v0.8 * 0c867ae : blog: Update blog post re streams2 - http://git.io/I1TwHQ
00:40:35  * Chip_Zeroquit (Ping timeout: 244 seconds)
00:40:42  * karupanerurajoined
00:43:13  * Chip_Zerojoined
00:44:53  * piscisaureus_quit (Quit: ~ Trillian Astra - www.trillian.im ~)
00:46:54  <MI6>joyent/node: isaacs v0.8 * 9d0103d : blog: More streams2 wordsmithing - http://git.io/A7GVkA
00:49:12  <sblom>isaacs: Starting to look at this right now.
00:49:32  <isaacs>alright, blog post posted.
00:50:00  <sblom>isaacs: Remind me where to look.
00:51:40  <isaacs>sblom: in src/node.js there's a call to net.Socket(fd)
00:51:48  <sblom>okay
00:52:01  <isaacs>sblom: this should be calling createPipe, because it's a pipe
00:52:07  <isaacs>er, no, wait, we call createPipe in the parent.
00:53:52  <isaacs>anyway, here's the test: require('child_process').spawn('node',['-e','console.log(1)']).stdout.pipe(process.stdout)
00:54:23  <isaacs>this works fine: require('child_process').spawn('node',['-e','console.log(1)'],{stdio:['pipe',1,2]})
00:54:30  <isaacs>(using the parent's shared fd)
00:56:12  <isaacs>sblom: ^
00:57:11  * kazuponjoined
00:57:50  <sblom>Okay.
00:58:00  <sblom>isaacs: ^
01:02:36  <isaacs>:)
01:06:51  <isaacs>sblom: also, it keeps node open... so it's waiting for something.
01:06:54  <isaacs>i'm not sure what
01:07:01  <isaacs>stdio bugs are the worst.
01:07:06  <isaacs>so tricky to track down
01:09:41  <isaacs>sblom: more info: it works fine when it's *not* a pipe, so it seems to me unlikely to be the code in lib/net.js that's to blame
01:10:01  <isaacs>sblom: of course, it starts when net.js changes... so. there's that.
01:18:28  * Chip_Zeroquit (Ping timeout: 245 seconds)
01:21:20  * Chip_Zerojoined
01:22:26  * kazuponquit (Remote host closed the connection)
01:22:52  <sblom>isaacs: Cool--I've got enough info to start poking around on this.
01:22:55  <sblom>Thanks.
01:26:44  * piscisaureus_joined
01:32:00  * piscisaureus_quit (Ping timeout: 264 seconds)
01:34:10  <isaacs>sblom: aha... i'm calling read(0) because i'm thinking that stdout/err are readable.
01:34:14  <isaacs>sblom: that could be related.
01:49:25  * c4miloquit (Remote host closed the connection)
01:55:47  * abraxasjoined
01:59:02  <isaacs>sblom: oh, i think i fixed it..
01:59:39  * pooyaquit (Quit: pooya)
01:59:54  <isaacs>hmm... at least partly
02:02:26  <sblom>isaacs: Oh, cool.
02:02:55  <sblom>isaacs: I think I'm still missing a repro step.
02:03:09  <sblom>Here's what I'm doing.
02:03:14  <sblom>1) building master
02:03:27  <sblom>2) looking in node.js for a call to net.Socket(fd)
02:03:30  <sblom>and I can't find it.
02:04:05  <isaacs>sblom: in src/node.js there's a function createWritableStdioStream(fd) {
02:04:13  <sblom>I'm guessing you mean net.Stream
02:04:13  <isaacs>sblom: line 406
02:04:15  <sblom>Yeah.
02:04:16  <sblom>Okay.
02:04:18  <isaacs>yeah, same thing
02:04:20  <isaacs>sorry :)
02:04:24  <sblom>Ahh, no problem.
02:04:27  <isaacs>net.Stream is the old timey classic name for it
02:04:31  <sblom>Okay.
02:04:46  <sblom>My secret decoder ring isn't very, ummm, well-populated yet. :p
02:05:01  <sblom>Okay.
02:09:24  <isaacs>sblom: Ok, so, this looks like it fixes at least some use cases: https://github.com/isaacs/node/commit/6377c814c65bc41dcf401bb97c57a4d15ba0e68d
02:12:00  <isaacs>sblom: https://gist.github.com/4350236
02:12:12  <isaacs>sblom: the first script runs fine now, with this patch applied.
02:12:17  <isaacs>sblom: but hte second still fails.
02:13:11  * dapquit (Quit: Leaving.)
02:13:52  <isaacs>sblom: node -e still fails, it looks like...
02:14:13  <isaacs>so... it could be that it's not running the script at all when stdout is a pipe? i dunno...
02:15:15  <isaacs>ok, it's definitely running the script.
02:15:42  <isaacs>and if it doesn't write to stdout, then it closes normally
02:17:53  * pooyajoined
02:18:42  <isaacs>i've gotta run.
02:19:22  <sblom>isaacs: Okay--I'll go run the second one and see what I can see. I'm still not fully caught up with where you are on this, and this is way over my node-head.
02:19:35  <isaacs>sblom: only way to learn to swim is to dive in :)
02:19:38  <isaacs>sblom: i believe in you :)
02:19:39  <sblom>Yeah.
02:19:41  <sblom>I'm diving.
02:19:53  <isaacs>ok, i might be back online tonight, maybe not until tomorrow.
02:19:57  <sblom>Okay.
02:20:10  <isaacs>if we can get 0.9.4 out tomorrow, that'd be great. might have to go out with -e busted. not the end of hte world.
02:20:12  <sblom>I'll be online all day tomorrow and the vast majority of my coworkers won't. :) So I know I'll be around.
02:20:13  <isaacs>(hah)
02:20:25  <isaacs>tomorrow afternoon i fly to SoCal for xmas
02:20:28  <sblom>oh
02:20:29  <sblom>Okay.
02:20:41  <sblom>I'll see how much time I can make for this tonight.
02:20:51  <isaacs>kewl, thanks a lot!
02:21:18  <sblom>Yeah--no worries. I want to help with at least _one_ bug for 0.9.4 :)
02:32:54  * bradleymeckjoined
02:36:19  * sblomquit
02:50:10  * TooTallNatequit (Quit: ["Textual IRC Client: www.textualapp.com"])
02:51:23  * brsonquit (Quit: leaving)
02:57:34  * pooyaquit (Quit: pooya)
03:04:00  * bradleymeckquit (Quit: bradleymeck)
03:36:11  * kazuponjoined
03:37:04  * kazuponquit (Remote host closed the connection)
03:37:19  * kazuponjoined
03:40:45  * kazuponquit (Read error: Connection reset by peer)
03:41:09  * kazuponjoined
03:43:38  * kazuponquit (Read error: Connection reset by peer)
03:43:56  * kazuponjoined
03:49:53  * brsonjoined
03:50:52  * wolfeidauquit (Read error: Connection reset by peer)
03:51:30  * wolfeidaujoined
04:01:06  * sblomjoined
04:04:42  * wolfeidauquit (Read error: Connection reset by peer)
04:07:13  * wolfeidaujoined
04:08:28  * wolfeidauquit (Remote host closed the connection)
04:24:52  * bradleymeckjoined
04:24:59  * wavded_joined
04:27:41  * Raynos_joined
04:28:05  * dscape_joined
04:29:03  <sblom>isaacs: I'm running the two test cases that you pasted into the gist now.
04:29:14  <sblom>I don't see any difference in outcome between the two.
04:29:47  <sblom>I think maybe I'm missing a step where I make a change to src/node.js before running them?
04:30:01  <isaacs>sblom: ah, yeah, so they're only different if you apply that patch i mentioned
04:30:23  <isaacs>sblom: curl https://github.com/isaacs/node/commit/6377c814c65bc41dcf401bb97c57a4d15ba0e68d.patch | git apply
04:30:23  * indutny_joined
04:30:44  <sblom>isaacs: applying
04:30:54  <isaacs>sblom: or you can: git remote add isaacs git://github.com/isaacs/node; git fetch -a isaacs; git checkout 6377c814
04:31:30  <isaacs>then build, and the "normal" module test will work, but not the node -e test
04:31:49  <sblom>isaacs: building
04:35:59  <isaacs>sblom: i'm connected to IRC, but not actually in front of my real computer atm.
04:36:05  <isaacs>sblom: technically, i'm not here :);
04:36:15  <sblom>isaacs: Okay.
04:36:53  <sblom>Having applied the patch, I'm seeing the second one say https://gist.github.com/6112d3a802278b569a80
04:37:01  <sblom>Which looks about like I'd expect.
04:37:06  <sblom>(second one meaning b.js)
04:37:35  <sblom>(And if being not here makes things difficult, I can wait until you can make it back to a real computer.)
04:39:19  <isaacs>hmm... wait a second..
04:39:26  <isaacs>i was seeing b.js still fail with the patch
04:39:33  <isaacs>sblom: does a.js work with the patch applied?
04:39:47  <isaacs>sblom: if i just fixed the bug and stole your thunder, i'm sorry ;P
04:39:55  <isaacs>sblom: i really thought it was still busted in the -e case
04:40:47  <isaacs>sblom: because, yes, that is exactly the intended output
04:42:04  <sblom>isaacs: updated the gist https://gist.github.com/6112d3a802278b569a80
04:42:23  <sblom>isaacs: a.js seems to do work, too.
04:42:33  <isaacs>wow, so, they're both working
04:42:35  <sblom>isaacs: at least I get a PID and "child exit".
04:42:39  <sblom>Looks like it to me.
04:42:50  <isaacs>ok, i can probably chalk this up to "isaacs doing testing wrong on windows"
04:42:50  <sblom>Win Server 2012 64 bit, debug build of node.exe.
04:42:54  <sblom>Heh.
04:42:57  <isaacs>hm. ok.
04:43:10  <isaacs>i was using Windows 7 32bit build of release node.exe
04:43:22  <isaacs>the win version shoudln't matter, i'd think
04:43:29  <isaacs>but hte 32/64 test/release might
04:43:34  <sblom>Heh. Thinking if I have an easy way to double check that. Yeah--I can't believe they would've changed something so fundamental.
04:43:36  <isaacs>can you try with the relesae version?
04:43:42  <sblom>isaacs: Yeah. I'll try release.
04:43:47  <isaacs>nah, it would've fucked us already by now ;)
04:44:00  <isaacs>if they had changed how stdout worked between win7 and win server 2012
04:44:15  <isaacs>ohhh... you know what...
04:44:30  <isaacs>i had a bunch of other debugging type code in my windows working dir that i didn't add in the patch
04:44:37  <isaacs>i bet some of that was throwing it into some bad state.
04:44:46  <isaacs>since it might've been trying to console.error before stderr was set up or something.
04:45:50  * bradleymeckquit (Quit: bradleymeck)
04:45:53  <isaacs>i'll run a make test a bit later and see where we're at.
04:46:10  <sblom>isaacs: I'm hopeful that it's something like that. It just means we'll have to look elsewhere for my first 0.9.4 bug and by elsewhere, I mean 0.9.5. :)
04:46:20  <isaacs>hah
04:46:25  <sblom>I'll still test this release build that's in progress.
04:46:27  <isaacs>no worries, it's good to have you back all the same.
04:46:30  <sblom>:)
04:46:58  <isaacs>sblom: if you want to make it so that the etw/perfctr stuff is checked in, that'd be super handy.
04:47:15  <isaacs>since i'll have to build that tomorrow
04:47:38  <sblom>isaacs: okay--I can easily get that into a pull request tonight.
04:47:42  <isaacs>(or figure out how to install MSVS non-express version)
04:47:42  <isaacs>kewl
04:47:53  <isaacs>cc @piscisaureus in the notes
04:47:58  <sblom>Okay--will do.
04:48:28  <isaacs>thanks!
04:48:58  <sblom>isaacs: This version won't have the build script that complains if you _do_ have ctrpp.exe (i.e. if you're piscisaureus or sblom) and its output is different than what's checked in.
04:49:10  <sblom>But at least it'll build for you tomorrow.
04:49:52  <isaacs>ah, ok
04:49:56  <isaacs>that's probably fine
04:49:57  <sblom>I think I'll have to do that sanity check in a python script. Do we have any problems with new ones of those getting checked in?
04:50:18  <sblom>It seems like our build process so far is super simple.
04:50:19  <isaacs>as long as it runs in the same versions of python that our other stuff does, then whatever.
04:50:22  <sblom>Yeah.
04:50:23  <sblom>Okay.
04:50:36  <isaacs>you know what would be *awesome*?
04:50:40  <sblom>What's that?
04:50:43  <isaacs>if instead of vcbuild.bat we had a make.bat
04:50:54  <isaacs>that took exactly the same arguments as our Makefil
04:50:56  <isaacs>*Makefile
04:51:13  <isaacs>so i could run exactly the same command on windows or unix.
04:51:23  <isaacs>and have everything Just Work
04:51:43  <isaacs>`make release` would create a .deb and .rpm on linux, a .pkg on os x, and a .msi on windows
04:51:57  <isaacs>i fantasize about writing that make.bat file sometimes.
04:52:13  <isaacs>then i wake up, and remember how awful batch files are...
04:53:08  <sblom>Yeah. Just thinking out loud: since we already depend on python to build, is there any reason make.bat couldn't just invoke some python script?
04:53:49  <sblom>That at least means that Windows-free folks could more painlessly make changes to the Windows build script without having to learn the kooky .bat syntax.
04:54:24  <sblom>I don't know if there's much overlap between "tries hard to avoid Windows" and "has to make changes to Node-on-Windows build process", tho.
04:54:24  <isaacs>i guess, yeah
04:56:39  * avalanche123joined
04:56:42  * avalanche123quit (Excess Flood)
04:57:05  * isaacsis that overlap
04:57:08  <isaacs>i suspect
04:58:07  <sblom>heh
04:58:56  <sblom>isaacs: both tests seem to work fine with release also.
04:59:12  <isaacs>sblom: that is fantastic.
04:59:29  <isaacs>sblom: today i learned: On windows, DO NOT READ FROM STDOUT IT IS FOR WRITING DO NOT READ IT
04:59:42  <sblom>Heh.
04:59:44  <isaacs>unix seems to not care even a little
04:59:51  <isaacs>which is, admittedly, a bit odd.
05:00:16  <sblom>let me know if you can't reproduce my success.
05:00:32  <isaacs>will do
05:00:34  <sblom>I'll keep lurking a little more tonight, and I'll be back online around 9 in the morning.
05:00:48  <isaacs>ok. thanks again :)
05:01:13  <isaacs>sblom: btw, if you wanna get more exposed to node stuff, you can always chill in #node.js and try to answer questions and help noobs.
05:01:33  <isaacs>sblom: often it's just a matter of knowing how to look stuff up.
05:01:56  <isaacs>sblom: it is addictive and tiring, though. watch out.
05:02:55  <sblom>isaacs: I'll try to do that at least a little. I'm also pretty active on Stack Overflow--I should start watching the [node.js] tag
05:03:08  <isaacs>oh, great.
05:03:12  <isaacs>yeah, i'm not very activ there.
05:03:23  <isaacs>it'd be good if we had some core-team presence
05:03:40  <sblom>There are actually quite a few unanswered questions just quickly glancing through the list.
05:03:54  <sblom>I'll go look for low-hanging fruit there.
05:03:58  * avalanche123joined
05:03:58  * avalanche123quit (Excess Flood)
05:04:29  <isaacs>kewl
05:04:33  <isaacs>don't burn out, though :)
05:04:39  <isaacs>set a limit
05:12:41  * avalanche123joined
05:12:41  * avalanche123quit (Excess Flood)
05:13:57  * bradleymeckjoined
05:23:29  * kazuponquit (Ping timeout: 265 seconds)
05:27:48  * sblomquit (Quit: sblom)
05:27:59  * avalanche123joined
05:28:00  * avalanche123quit (Excess Flood)
05:30:29  * avalanche123joined
05:30:30  * kazuponjoined
05:30:32  * avalanche123quit (Excess Flood)
05:34:59  * avalanche123joined
05:35:00  * avalanche123quit (Excess Flood)
05:35:35  * avalanche123joined
05:35:37  * avalanche123quit (Excess Flood)
05:37:53  * mmaleckichanged nick to mmalecki[zzz]
05:41:39  * felixgejoined
05:41:39  * felixgequit (Changing host)
05:41:39  * felixgejoined
05:45:46  * mikealquit (Quit: Leaving.)
05:47:30  * kazupon_joined
05:48:36  * kazuponquit (Ping timeout: 248 seconds)
05:48:54  * pooyajoined
06:00:20  <indutny_>isaacs: you still there?
06:00:31  <indutny_>isaacs: wanna take a look at this https://github.com/joyent/node/pull/4449 ?
06:01:34  <indutny_>heh
06:01:34  <indutny_>ok
06:02:03  <isaacs>indutny_: hmm... how does it do with emit('data') without any args?
06:02:17  <isaacs>indutny_: i wonder if you're getting a gain here just because you'er guessing the right number of arguments.
06:02:18  <indutny_>I suppose it doesn't change anything at all
06:02:29  <isaacs>indutny_: because if that's the case, we should guess 1 arg
06:02:31  <indutny_>I'll amend benchmark
06:03:58  <isaacs>i'm done for tonight
06:04:07  * bradleymeckquit (Quit: bradleymeck)
06:04:08  <isaacs>i'll take a look tomorrow.
06:04:25  <indutny_>:)
06:04:26  <indutny_>ok
06:04:32  <isaacs>indutny_: if it makes http_simple go faster, then that's usually a better test. it calls emit() a zillion times, in a lot of different ways
06:04:41  <isaacs>so it's a bit less likely to get eaten by the sphinx ;)
06:04:48  <indutny_>hahaha
06:04:52  <indutny_>well, I
06:05:01  <indutny_>I'd like to test it synthetically too
06:05:05  <isaacs>sur sure
06:05:18  <isaacs>but it ought to be an improvement in both places.
06:05:20  <isaacs>ok, i'm ot
06:05:22  <isaacs>out
06:05:23  <isaacs>g'nite
06:05:31  <indutny_>g'nite
06:06:06  * kazupon_quit (Ping timeout: 256 seconds)
06:08:08  <indutny_>isaacs: you're right, it's slower with no args
06:09:51  * indexzerojoined
06:10:48  * kazuponjoined
06:11:38  * kazuponquit (Read error: Connection reset by peer)
06:11:44  * mikealjoined
06:12:04  * kazuponjoined
06:26:23  * deoxxachanged nick to science
06:26:39  * sciencechanged nick to deoxxa
06:31:39  * avalanche123joined
06:31:39  * avalanche123quit (Excess Flood)
06:37:07  * felixge_joined
06:37:12  * felixgequit (Read error: Connection reset by peer)
06:37:20  * felixgejoined
06:41:47  * felixge_quit (Ping timeout: 265 seconds)
06:52:02  * avalanche123joined
06:52:03  * avalanche123quit (Excess Flood)
06:57:35  * avalanche123joined
06:57:36  * avalanche123quit (Excess Flood)
07:08:40  * brsonquit (Ping timeout: 265 seconds)
07:13:11  * avalanche123joined
07:13:12  * avalanche123quit (Excess Flood)
07:16:20  * felixgequit (Quit: felixge)
07:17:16  * indexzeroquit (Quit: indexzero)
07:19:49  * felixgejoined
07:19:49  * felixgequit (Changing host)
07:19:49  * felixgejoined
07:29:16  * avalanche123joined
07:29:22  * avalanche123quit (Excess Flood)
07:30:43  * rendarjoined
07:41:18  * `3rdEdenjoined
08:00:04  * AvianFluquit (Remote host closed the connection)
08:04:31  * stagasjoined
08:18:06  * abraxasquit (Remote host closed the connection)
08:19:08  * abraxasjoined
08:19:13  * abraxasquit (Remote host closed the connection)
08:44:06  * dscape_quit (*.net *.split)
08:44:06  * skebcio_quit (*.net *.split)
08:44:06  * pquernaquit (*.net *.split)
08:44:06  * tristramquit (*.net *.split)
08:44:06  * davispquit (*.net *.split)
08:46:15  * dscape_joined
08:46:15  * tristramjoined
08:46:15  * skebcio_joined
08:46:15  * pquernajoined
08:46:15  * davispjoined
08:50:54  * bentkusjoined
08:51:06  <bentkus>where is pisci, where is bnoord
08:51:24  <bentkus>I guess no libuv presents for me this year :(
08:53:28  <indutny_>bentkus: what do you want anyway?
09:00:00  <bentkus>A comment
09:00:03  <bentkus>thats it
09:00:12  <bentkus>"do number 2"
09:01:02  <bentkus>like solution number 2, not the number 2 of the pseudo english lang
09:01:37  <bentkus>you work there too
09:01:45  <bentkus>indutny_: how cold is it now at your place?
09:01:52  <indutny_>pretty cold
09:02:03  <indutny_>-20
09:02:10  <indutny_>http://www.bbc.co.uk/weather/524901
09:02:35  <indutny_>well, bbc lies a little bit
09:02:39  <indutny_>but that's about it
09:02:42  <bentkus>moskva
09:03:07  <bentkus>-14 here, in 2 days it will be -20
09:03:17  <bentkus>so good
09:03:46  <bentkus>O you are ahead one hour in the 21th apocalypse
09:03:53  <bentkus>how is it going? no meteors or shit?
09:04:17  <indutny_>so far looks good :)
09:04:57  <bentkus>You should drink some vodka, it might be the last time
09:06:12  <bentkus>I drank yesterday for today and tomorow
09:06:14  * loladiroquit (Quit: loladiro)
09:06:23  <bentkus>vodka biz pivo, dzengi na veter
09:07:28  <indutny_>indeed
09:08:25  <bentkus>2 days ago when traveling to the airport i memorized the cyrilica
09:10:32  <bentkus>But these russian keyboards
09:10:40  <bentkus>I need one
09:11:34  * Chip_Zeroquit (Ping timeout: 244 seconds)
09:12:27  * Chip_Zerojoined
09:13:16  <bentkus> пыэда
09:13:32  <bentkus>do I need to use the short I for that?
09:13:47  <deoxxa>ya machinu!
09:14:03  <indutny_>bentkus: erm
09:14:08  <indutny_>bentkus: do you want to say cunt?
09:14:19  <bentkus>Yes
09:14:22  <indutny_>you should use и instead of ы
09:14:25  <indutny_>and
09:14:31  <indutny_>з instead of э
09:14:38  <bentkus>Thats the same symbolo
09:14:40  <indutny_>it's different character
09:14:44  <indutny_>no
09:14:52  <bentkus>fucking online russian keyboard
09:14:54  <indutny_>з is more like 3
09:15:01  <bentkus>O yeah
09:15:04  <bentkus>I see it now
09:16:25  * loladirojoined
09:16:45  <bentkus>пизда
09:17:32  <bentkus>отлична
09:19:32  <indutny_>ok
09:19:35  <indutny_>now another thing
09:19:41  <indutny_>this is a curse word :)
09:19:57  <indutny_>one of the shittiest
09:20:16  <indutny_>and отлична should be отлично
09:20:34  <indutny_>bentkus: are you from poland?
09:20:42  <indutny_>or from bulgari?
09:21:11  <bentkus>from the country, which left first the USSR in 1990
09:21:25  <bentkus>Lithuania
09:21:29  <indutny_>aah
09:21:33  <indutny_>ok
09:21:54  <bentkus>no border with russian
09:22:10  <bentkus>o yes we have
09:22:16  <bentkus>but like that part of russia which belonged to germany
09:22:48  <indutny_>you mean kenigsberg?
09:23:01  <indutny_>or kaliningrad
09:23:07  <indutny_>which is the same thing
09:25:50  <bentkus>königsberg
09:26:14  <bentkus>it used to be deutsches Reich
09:26:55  <indutny_>yeah
09:28:14  <bentkus>индути федор
09:28:31  <indutny_>индутный федор
09:28:48  <indutny_>indutny <- is not exact transliteration
09:28:56  <indutny_>it just looks more adequate
09:30:20  <bentkus>that nick is nice
09:30:31  <indutny_>that's not a nick
09:30:35  <indutny_>it's my family name
09:30:48  <indutny_>and it's in my foreign passport
09:30:52  <indutny_>but with two Y
09:30:54  <indutny_>indutnyy
09:32:09  <bentkus>залез па дзйрево
09:32:19  <bentkus>this is my atempt on swearing in russian
09:33:16  <bentkus>i need to learn the difference between й и ы е
09:33:26  <bentkus>at which class do you go through this stuff?
09:33:59  <indutny_>em... I suppose 2nd
09:34:03  <indutny_>not really sure
09:34:10  <indutny_>probably later
09:34:20  <indutny_>well, we learn how to write words
09:34:25  <indutny_>at 1st class
09:34:28  <indutny_>and how to read them
09:34:36  <bentkus>and later you learn the rules
09:34:41  <indutny_>yes
09:36:15  * kazupon_joined
09:36:40  <bentkus>it will be funny to go to a bookstore and buy like a 4rd graders excercise book
09:36:48  * avalanche123joined
09:37:01  <bentkus>But it's awesome, at last I able to decypher the russian torrent sites
09:37:35  * kazuponquit (Ping timeout: 255 seconds)
09:37:41  * bnoordhuisjoined
09:37:46  <bentkus>I am
09:37:53  <bentkus>I already speak like a russian
09:38:06  <bentkus>but russians don't have articles
09:38:18  <bentkus>That freaking german invention
09:40:51  * avalanche123quit (Remote host closed the connection)
09:40:52  <indutny_>haha
09:40:53  <indutny_>yeah
09:41:00  <indutny_>life is much simplier without all this crap
09:43:22  <bentkus>but why did russians borrow so many words from germans
09:43:25  <bentkus>like tanzivat
09:44:15  <bentkus>танзиват
09:44:20  * bnoordhuisquit (Ping timeout: 248 seconds)
09:44:26  <bentkus>o there was bnoord
09:44:29  <bentkus>Totally missed him
09:44:37  <bentkus>cyka quit
09:44:59  * pooyaquit (Quit: pooya)
09:45:54  <bentkus>what is this txdv faggot doing in here
09:45:58  * loladiroquit (Quit: loladiro)
09:47:29  <indutny_>bentkus: tranzivat?
09:47:56  <indutny_>bentkus: actually, we've pulled a lot of words from another languages
09:47:57  * avalanche123joined
09:48:02  <indutny_>most from french
09:48:10  <indutny_>some from english
09:48:14  <indutny_>and much less from german
09:48:24  <indutny_>I mean, there're some words that english and french pulled from german
09:48:29  <indutny_>well, there're many
09:48:36  <indutny_>but we pulled them from french and english
09:48:39  <indutny_>not german itself
09:51:09  <bentkus>Hm
09:51:10  * avalanche123quit (Remote host closed the connection)
09:51:17  <bentkus>I thought Pioter the great was influenced by german culture
09:51:21  <indutny_>yes
09:51:39  <indutny_>but you should also know that some time ago french was default foreign language to learn
09:51:51  <indutny_>for one age
09:52:11  <bentkus>O yeah
09:52:27  <bentkus>We are 3 guys living in a flat in germany (students)
09:52:37  * Raynosquit (Remote host closed the connection)
09:52:37  * indutnyquit (Remote host closed the connection)
09:52:37  <bentkus>one is russian, his first foreign language was french
09:52:37  * wavdedquit (Write error: Broken pipe)
09:52:38  * dscapequit (Write error: Broken pipe)
09:54:28  <indutny_>oh, ok
09:54:36  <indutny_>my wife is graduated filologist
09:54:51  * Raynos_quit (Remote host closed the connection)
09:54:51  * dscape_quit (Remote host closed the connection)
09:54:51  * indutny_quit (Remote host closed the connection)
09:54:51  * wavded_quit (Remote host closed the connection)
09:54:55  <bentkus>What language?
09:54:59  <bentkus>Like russian filologist?
09:55:04  <bentkus>French likely
10:07:01  * kazupon_quit (Remote host closed the connection)
10:08:25  <bentkus>O no he left
10:08:26  <bentkus>wtf
10:08:34  <bentkus>japan and russia gut cut off the internet
10:08:43  <bentkus>21th, it's comming
10:12:26  * indutny_joined
10:14:34  * dscape_joined
10:16:43  * Raynos_joined
10:21:31  * wavded_joined
10:25:29  * piscisaureus_joined
10:25:39  <piscisaureus_>notes
10:25:49  <piscisaureus_>ircretary: notes
10:27:47  <piscisaureus_>guys I cannot work today or this weekend
10:27:48  <piscisaureus_>:-(
10:27:52  <piscisaureus_>or :-)
10:28:04  <piscisaureus_>anyway, I have no laptop until monday (earliest)
10:40:43  <bentkus>piscisaureus_: :(
10:40:55  <bentkus>but wait, developing on libuv is not work, it is fun!
10:40:58  <bentkus>so have a fun time!
10:42:09  <bentkus>btw the best christmas present you can do is to comment on issues
10:43:09  <bentkus>like, you know, the one i did
10:43:23  * bentkusquit (Quit: Lost terminal)
10:48:06  * stagasquit (Ping timeout: 244 seconds)
10:57:24  * piscisaureus_quit (Ping timeout: 248 seconds)
11:02:45  * abraxasjoined
11:03:42  * abraxasquit (Remote host closed the connection)
11:05:14  * stephankquit (Ping timeout: 240 seconds)
11:12:22  * bentkusjoined
11:12:32  <bentkus>indutny_: no apocalpyse in russia?\
11:28:28  * stephankjoined
11:50:48  <bentkus>stephank: where are you from?
12:09:48  <indutny_>no apocalypse so far
12:10:01  * indutny_changed nick to indutny
12:41:01  * sgallaghjoined
12:47:04  * felixgequit (Quit: felixge)
12:53:34  * stephankquit (Ping timeout: 240 seconds)
13:05:41  * mmalecki[zzz]changed nick to mmalecki
13:09:11  <mmalecki>I expected apocalypse to actually start in Russia
13:42:56  * stephankjoined
13:44:25  * c4milojoined
13:53:06  * indutnyin flames
13:53:16  * yawnttakes a vodka fire estinguisher
13:53:29  <yawnt>*ex
14:09:44  * avalanche123joined
14:09:50  <indutny>where is ben?
14:10:00  <indutny>looks like doomsday is only in AMS
14:14:36  * avalanche123quit (Ping timeout: 264 seconds)
14:48:28  * bentkusquit (Ping timeout: 252 seconds)
14:51:18  * stagasjoined
14:57:57  * TheJHjoined
14:58:44  * AvianFlujoined
14:59:54  * bradleymeckjoined
15:14:11  * stagasquit (Ping timeout: 255 seconds)
16:06:41  * `3rdEdenquit (Remote host closed the connection)
16:37:01  <isaacs>indutny: for that EE patch, if you want to do that, we need to know what number of args is most common.
16:37:37  <isaacs>indutny: you could put some logging in there, and then run an http benchmark or something, and tally the results.
16:39:47  <MI6>joyent/node: isaacs reviewme * 6377c81 : stdio: Fix windows stdio pipes in streams2 land Hopefully. - http://git.io/oES4oQ
16:39:48  <isaacs>review please? ^
16:42:55  <MI6>joyent/node: isaacs v0.8 * d30e76e : npm: upgrade to 1.1.70 - http://git.io/zKmEdw
16:48:57  <MI6>joyent/node: isaacs master * c048c81 : http: Trivial fix for comments and 'this.read' - http://git.io/KsxvfA
16:50:24  * bradleymeckquit (Quit: bradleymeck)
16:51:59  <MI6>joyent/node: isaacs master * fb915ed : lint - http://git.io/cr-mag
16:56:48  <mmalecki>isaacs: you prefer configure or env variable during build?
16:57:30  <tjfontaine>I would say configure makes the most sense
16:57:51  <mmalecki>ya, agreed
16:58:09  <tjfontaine>that fits most existing styles, would plug into .deb and .rpm lands nicely
16:59:02  <mmalecki>I have no idea about building .deb and .rpm tbh
16:59:16  <tjfontaine>heh, well take my word for it then :)
16:59:36  <MI6>joyent/node: isaacs master * 0edd93d : test: Fix simple/test-http-localaddress - http://git.io/kxsl5Q
16:59:54  <isaacs>mmalecki: first one, then the other
17:00:08  <isaacs>mmalecki: like we do with --prefix=/foo PREFIX=bar
17:02:18  <mmalecki>so ./configure --tag=my-build and MAKE TAG=my-build?
17:02:30  <isaacs>mmalecki: yeah
17:02:41  <isaacs>priority is: env, config flag, default
17:02:44  <mmalecki>word, on it :)
17:03:11  <isaacs>though... actually, since it's something that goes in the actual code, i guess we don't ever use an env for that...
17:03:26  <isaacs>but it'd be more convenient
17:03:52  <mmalecki>we can tho, -DNODE_TAG=blah
17:03:55  <isaacs>git checkout blerg; TAG=blerg make
17:03:58  <isaacs>right
17:04:14  <isaacs>it always ends with a -D to gcc
17:04:26  <isaacs>or whatever the equivalent is on windows
17:04:48  <mmalecki>as I recall from the dark ages, it's the same
17:05:57  * mmaleckicalls his Windows times "dark ages"
17:11:50  <isaacs>With that patch on reviewme, windows looks like this: [04:11|% 100|+ 489|- 13]: Done
17:11:54  <isaacs>huge improvement
17:18:54  <mmalecki>am I the only one who hates the complexity involved into build systems?
17:19:09  <isaacs>mmalecki: no, we all hate it
17:19:59  <tjfontaine>and they all suck
17:20:48  * sblomjoined
17:22:44  <isaacs>sblom: https://gist.github.com/4354230
17:22:58  <isaacs>sblom: still getting that with the patch i made last night
17:23:12  <isaacs>sblom: if you're *not* getting that, then i'm confused
17:23:39  * pooyajoined
17:28:44  <mmalecki>I'll write my own one, with blackjack and hookers
17:36:04  <mmalecki>isaacs: done with configure, but not sure how to do that with make yet
17:37:43  * stagasjoined
17:39:40  <isaacs>mmalecki: well, reading the env in make is easy. `NODE_TAG ?=` and then later do `-DTAG=$(NODE_TAG)`
17:39:51  <isaacs>although, i guess it goes through to gyp or whatever.
17:39:58  <isaacs>this is TooTallNate and bnoordhuis magic.
17:40:42  <mmalecki>I got the gyp part figured out already
17:43:42  * avalanche123joined
17:43:43  * avalanche123quit (Excess Flood)
17:45:04  * avalanche123joined
17:45:05  * avalanche123quit (Excess Flood)
17:46:17  * avalanche123joined
17:46:17  * avalanche123quit (Excess Flood)
17:50:10  <sblom>isaacs: I'll try again. But I'm absolutely sure it was working.
17:50:32  * mralephjoined
17:50:44  <sblom>isaacs: Here's one sanity check. Before I applied your patch, I saw "1" printed by b.js, and then it just sat there stuck.
17:50:55  <sblom>Is that what I should have expected _before_ applying the patch?
17:52:12  * avalanche123joined
17:53:55  <isaacs>sblom: weeeiiird.
17:54:02  <isaacs>sblom: you have checked my sanity right into the walls
17:54:05  <isaacs>sblom: :)
17:54:11  <isaacs>sblom: i'm seeing nothing printed, then sitting there stuck
17:56:32  * avalanche123quit (Ping timeout: 252 seconds)
17:57:45  * pooyaquit (Quit: pooya)
18:01:28  <isaacs>oohhhhh!!!!!
18:01:30  <isaacs>sblom: i'm dumb
18:01:39  <isaacs>sblom: i'm execing "node", not process.execPath
18:02:07  <isaacs>sblom: haha, you're absolutely correct, it does fix things.
18:02:18  <isaacs>sblom: i'm going to add a test and then land it in master, then
18:02:38  <isaacs>sblom: if you can do that perfctr/etw generated thing checkin, we can get 0.9.4 underway
18:03:34  <tjfontaine>he has a pullreq in for it, but bert is without a laptop until monday it seems
18:05:09  * loladirojoined
18:06:35  <isaacs>oh, ok
18:07:34  <isaacs>sblom: i'm gonna try building with this thing. if it works, i'll land it, and it can be our xmas present to piscisaureus
18:09:13  * c4miloquit (Remote host closed the connection)
18:18:22  * c4milojoined
18:30:02  * TooTallNatejoined
18:35:17  * mjr_joined
18:41:56  <isaacs>build works.
18:42:20  <isaacs>adding test for stdout reading, and then on to 0.9.4!
18:42:30  <isaacs>sblom: you got a bug fix in this one. you fixed the build :)
18:51:58  * `3rdEdenjoined
18:53:21  * TheJHquit (Ping timeout: 265 seconds)
18:54:16  * loladiroquit (Quit: loladiro)
18:55:32  * Raynos_changed nick to Raynos
18:59:43  * joshthecoderjoined
19:01:13  * toothrchanged nick to toothort
19:01:17  * toothortchanged nick to toothrot
19:02:24  * avalanche123joined
19:02:24  * avalanche123quit (Excess Flood)
19:04:22  * sblomquit (Ping timeout: 252 seconds)
19:05:11  * loladirojoined
19:05:20  * pooyajoined
19:06:52  <MI6>joyent/node: Scott Blomquist master * 841b7f5 : Ease building with VS Express by checking in generated files. (+1 more commits) - http://git.io/8SlAyg
19:12:58  <TooTallNate>isaacs: i wanna land https://github.com/joyent/node/issues/3226 real quick for v0.9.4
19:13:07  <TooTallNate>isaacs: need to update the docs in the patch first though
19:18:07  * brsonjoined
19:18:17  * brsonquit (Client Quit)
19:18:25  * brsonjoined
19:19:13  * sblomjoined
19:19:28  <sblom>isaacs: glad to hear everything is working well this morning.
19:21:27  <isaacs>TooTallNate: let's do it for 0.9.5
19:21:31  <isaacs>TooTallNate: the changelog is already insane
19:21:37  <TooTallNate>isaacs: hahah, ok fair enough
19:21:38  <isaacs>it's been like 2 months :)
19:21:43  <TooTallNate>isaacs: do you have a branch for it already?
19:21:46  <TooTallNate>so that i can push to master...
19:24:55  <mmalecki>TooTallNate: herro!
19:25:04  <TooTallNate>mmalecki: yo yo
19:25:28  <mmalecki>TooTallNate: so I made https://github.com/joyent/node/pull/4448
19:25:42  <mmalecki>works with configure, but been wondering how to make it work with `make`
19:25:42  * TheJHjoined
19:25:54  <mmalecki>so, `make TAG=custom-build`
19:27:39  <isaacs>TooTallNate: yeah, go for it
19:27:52  <isaacs>TooTallNate: i'll merge it in after the release as usual
19:29:03  <TooTallNate>mmalecki: well hmm… the way you have it in the patch is nice since you can also check `process.config.variables.tag`
19:29:14  <TooTallNate>mmalecki: but with the make TAG=blah way, that wouldn't work
19:29:43  * loladiroquit (Quit: loladiro)
19:32:02  * `3rdEdenquit (Quit: zomgbrb)
19:34:40  <isaacs>mmalecki: yeah, i think maybe just skip the Makefil way, and make it a config var only
19:34:49  <isaacs>mmalecki: that's the more "correct" place for it anyway
19:35:14  * loladirojoined
19:35:15  <isaacs>here we go..
19:35:24  <TooTallNate>^ Mario
19:35:30  <MI6>joyent/node: isaacs created branch v0.9.4-release - http://git.io/mwcZ_g
19:35:54  <TooTallNate>big diff is big
19:36:11  <mmalecki>TooTallNate: word
19:36:15  <mmalecki>so, good for merge?
19:36:31  <TooTallNate>mmalecki: can you check on my comment in the thread first?
19:36:54  <TooTallNate>mmalecki: a "tagged" version of node-pre should still throw the "use the --nodedir flag" error
19:36:55  * loladiroquit (Client Quit)
19:37:33  <mmalecki>ah, sure
19:38:21  <isaacs>http://nodejs.org/docs/v0.9.4/api/stream.html
19:38:23  * sgallaghquit (Remote host closed the connection)
19:39:07  <mmalecki>TooTallNate: it's affected, yeah
19:39:19  <mmalecki>cause version[5] is '<tag>-pre'
19:39:33  <TooTallNate>mmalecki: ahh, so putting it after wouldn't work either
19:39:39  <mmalecki>maybe you could check if it ends with -pre instead? I feel that it might be better semantically
19:39:40  <TooTallNate>mmalecki: damn, need to update node-gyp then
19:39:50  <TooTallNate>mmalecki: ya for sure
19:39:52  <mmalecki>I can do a PR too :)
19:39:59  <TooTallNate>mmalecki: haha, sure thing ;)
19:40:13  * pooyaquit (Quit: pooya)
19:40:16  <mmalecki>but not now, fighting a fire
19:42:26  <isaacs>changelog notes? https://gist.github.com/4355243
19:45:15  <isaacs>oh, i'm dumb, i'm building the mac binary on my laptop, and i have a new mac mini sitting here with 8 cpus doing nothing.
19:45:29  <MI6>joyent/node: Nathan Rajlich master * 837df70 : repl: don't touch `require.cache` Fixes #3226. Consider a production se - http://git.io/Rr8AtA
19:45:37  <TooTallNate>isaacs: is the mini faster?
19:45:42  <isaacs>TooTallNate: much
19:45:43  <TooTallNate>isaacs: oh, you're on an air right?
19:45:55  <isaacs>TooTallNate: no, i'm on a mbp, but it's a year or two old
19:46:04  <TooTallNate>ahh
19:46:09  <isaacs>meh. id' have to copy over all my keys and whatnot for signing
19:46:16  <isaacs>that's kind of a project. next time
19:55:06  * loladirojoined
19:55:58  * sblomquit
19:57:16  * EhevuTovjoined
20:03:37  * bradleymeckjoined
20:05:15  * loladiroquit (Quit: loladiro)
20:06:04  <isaacs>TooTallNate: Can you review the changelog? anything you think is pointless there? We need to cut like 1/3 of that
20:06:07  <isaacs>at least.
20:06:14  <TooTallNate>ok
20:07:30  * `3rdEdenjoined
20:08:51  <mmalecki>TooTallNate: https://github.com/TooTallNate/node-gyp/pull/176
20:09:00  <TooTallNate>maybe consolidate the "* build: allow linking against system libuv (Stephen Gallagher)" ones into one
20:09:04  <mmalecki>that was easy :)
20:09:26  <mmalecki>actually, wait, have to check against undefined
20:09:41  <TooTallNate>mmalecki: you could just check the entire versions tsring
20:09:42  <TooTallNate>string
20:09:45  <TooTallNate>rather than [5]
20:11:22  <mmalecki>there, fixed
20:11:28  <mmalecki>well, both work
20:11:34  <TooTallNate>remove "* zlib: reduce memory consumption, release early (Ben Noordhuis)"
20:11:34  <isaacs>TooTallNate: yeah, they're all the same dude
20:11:48  <TooTallNate>maybe remove: "* benchmark: Add http-flamegraph (isaacs"
20:11:57  <TooTallNate>maybe remove: "* crypto: fix ssl error handling (Sergey Kholodilov)"
20:12:12  <TooTallNate>isaacs: a lot of them seem mentionable though
20:13:08  <isaacs>TooTallNate: update: https://gist.github.com/4355452
20:14:08  <isaacs>removed the tap output one, too. it's not enabled by default, anyway.
20:14:23  <isaacs>tjfontaine: sorry :( At least you can know that I personally appreciate the hell out of that :)
20:14:25  <TooTallNate>isaacs: lgtm i think, not too long now
20:14:28  <isaacs>kewl
20:14:29  <isaacs>thanks
20:14:32  <tjfontaine>isaacs: haha ya I'm not too fussed :P
20:14:49  <tjfontaine>before 0.9.5 we *really* need to fix setImmediate
20:14:56  <TooTallNate>"* readline: don't emit "line" events with a trailing 'n' char (Nathan Rajlich)"
20:15:01  <TooTallNate>^ didn't that one go into v0.8?
20:15:06  * TooTallNatecan't really remember...
20:16:36  <mmalecki>TooTallNate: wouldn't that be an API change?
20:17:18  <TooTallNate>mmalecki: ya… but it was a normalization, in "raw mode" (terminal:true) you don't get the \n
20:17:22  <mmalecki>also, can we merge my commit into 0.9 too?
20:17:25  <TooTallNate>so it kinda could have gone either way
20:17:29  <mmalecki>I can do the forward port
20:17:29  <mmalecki>I see
20:17:38  <TooTallNate>mmalecki: 0.9.5 probably
20:17:42  <TooTallNate>mmalecki: but i'll merge in a sec
20:21:09  <TooTallNate>isaacs: are you ok with mmalecki's patch on v0.8 or should it go to master?
20:21:18  <TooTallNate>i was thinking master initially
20:21:24  <TooTallNate>but it's targeted for v0.8
20:25:32  <isaacs>which one?
20:25:34  <isaacs>the TAG one?
20:25:39  <TooTallNate>isaacs: ya
20:25:41  <isaacs>that's a feature addition. master.
20:25:45  <TooTallNate>k
20:25:47  <isaacs>sorry, mmalecki.
20:25:53  <TooTallNate>mmalecki: needs a rebase against master then
20:26:01  <TooTallNate>it doesn't apply cleanly as-is
20:26:22  <isaacs>TooTallNate: if you apply to your local v0.8 branch, you can probably cherry-pick it onto master
20:26:37  <isaacs>unless the underlying stuff changed a lot
20:26:44  <TooTallNate>good point
20:27:43  <isaacs>TooTallNate: or you can make mmalecki do it :)
20:27:57  <isaacs>TooTallNate: i usually find it's easier to just do the git-fu myself rather than explain what i want done
20:28:04  <TooTallNate>indeed
20:28:07  <isaacs>ymmv
20:28:07  <TooTallNate>git-fuuuuu!!!!
20:28:18  <isaacs>omg, this ppkg build.
20:28:18  <mmalecki>it doesn't apply cleanly, I checked already
20:28:19  <isaacs>kills me
20:28:24  <isaacs>ah, ok
20:31:27  <mmalecki>TooTallNate: https://github.com/joyent/node/pull/4452
20:32:36  <TooTallNate>mmalecki: kewl, one sec. eating pizza
20:34:03  <MI6>joyent/node: isaacs created tag v0.9.4 - http://git.io/QNVxzQ
20:34:16  <MI6>joyent/node: isaacs v0.9.4-release * d86d83c : 2012.12.21, Version 0.9.4 (Unstable) * streams: Update all streaming int - http://git.io/goe8MQ
20:34:46  <isaacs>TooTallNate: are you about to land something in master? i'm gonna merge 0.9.4 in
20:35:12  <TooTallNate>isaacs: i landed the repl patch already
20:35:18  <isaacs>oh, you did that, ok
20:35:18  <isaacs>kew
20:36:46  <isaacs>merged in!
20:36:57  <MI6>joyent/node: isaacs master * 70eb227 : Now working on 0.9.5 (+7 more commits) - http://git.io/PJLYmQ
20:40:10  * wolfeidaujoined
20:40:41  * loladirojoined
20:40:50  * wolfeidauquit (Remote host closed the connection)
20:42:03  * wolfeidaujoined
20:44:43  <MI6>joyent/node: isaacs v0.8 * 6738e68 : blog: Post about v0.9.4 - http://git.io/kus_CA
20:47:19  <TooTallNate>mmalecki: this is a cool feature btw, thanks :)
20:47:48  <isaacs>yeah, really nice for comparative benchmarking
20:48:02  <MI6>joyent/node: Maciej Małecki master * f84bf5b : build: allow to specify custom tags When building custom `node` versions - http://git.io/sMN9Sw
20:48:04  <isaacs>i frequently get confused what i'm actualy testing when switching back and forth
20:48:21  <TooTallNate>isaacs: yup, i wanna use it for custom compiler flags
20:48:33  <TooTallNate>like my node binaries tailored for Raspberry Pi
20:48:57  <isaacs>ok, so, 0.9.4 is out. i'm gonna get some lunch, and hten i'll be packing for the rest of the afternon, and flying to LA until next week.
20:49:20  <TooTallNate>mmalecki: merged!
20:49:46  <isaacs>if anything is drastically missing or broken wiht hte release, leave a message and i'll check in later, but otherwise, i'll probably be offline until after christmas.
20:49:58  <wolfeidau>Question where do the node binaries get built and how?
20:50:16  <TooTallNate>wolfeidau: isaacs' laptop so far, and via the `make binary` command
20:50:50  <wolfeidau>TooTallNate: O so it does all the different os's in that "routine"?
20:51:40  <wolfeidau>I am asking as i noted that ./configure && make seems to produce slightly different binaries than those posted
20:51:54  <tjfontaine>different how?
20:51:54  <TooTallNate>wolfeidau: you run `make binary` and it'll generate a tarball for the current platform. so you have to run it on each platform
20:52:14  <tjfontaine>on osx for instance it will generate a lipo'd 32bit and 64bit binary
20:52:43  <wolfeidau>tjfontaine: The binary version includes at least one extra GCC flag for related to buffer overflows
20:53:03  <wolfeidau>TooTallNate: Thanks for the information I will take a look at the Makefile
20:53:25  <TooTallNate>wolfeidau: how do you know which gcc flags were used in the binary?
20:53:26  <tjfontaine>wolfeidau: it does ./configure --without-snapshot
20:54:22  <wolfeidau>TooTallNate: Well i ran into an interesting issue with sprintf vs snprintf in a binary module I was building, it only blew up on machines where the node binary release was installed
20:54:59  <wolfeidau>TooTallNate: I replicated the issue by adding -D_FORTIFY_SOURCE=2 to the GCC flags
20:55:20  <wolfeidau>without adding that the "generic" build method doesn't have an issue
20:55:38  <wolfeidau>by generic i mean ./configure && make
20:55:59  <TooTallNate>wolfeidau: well i think you should just use snprintf(), no?
20:56:03  <TooTallNate>and not buffer overflow :p
20:56:35  <wolfeidau>TooTallNate: Yes, as someone that hadn't done c for 5 years that is now apparent :)
20:56:53  <tjfontaine>you should have been doing snprintf 5 years ago too :P
20:57:13  <wolfeidau>Yes I may have been, but the guys code i was copying didn't :)
20:58:14  <wolfeidau>But I, aside from my mistakes on that side I was supprised node didn't include the flags it was compiled with in the --version
20:58:31  <wolfeidau>Or even a text file with the distrubution
20:59:28  <wolfeidau>I will look at hacking up a change to the "dist" routine to see if i can add it
20:59:47  <TooTallNate>wolfeidau: well you can check the `process.config` object but that only gives you a partial picture
21:00:04  <wolfeidau>Mite help others in the future to understand exactly what flags node distributions are built with
21:00:27  <wolfeidau>TooTallNate: GCC also has a flag to embed the "comple flags" into the binary
21:00:51  <wolfeidau>But i understand it isn't the only compiler
21:01:11  <isaacs>wolfeidau, TooTallNate: No, make binary will only make for the default arch
21:01:19  <isaacs>you have to do DESTCPU=blah for the other one
21:01:27  * EhevuTov_joined
21:01:32  <isaacs>TooTallNate: `make pkg` does the fat binary
21:01:44  <tjfontaine>isaacs: but you don't do anything special wrt cflags for a release?
21:01:51  <TooTallNate>isaacs: right, right, been a while since i looked at that code :p
21:02:53  * EhevuTovquit (Ping timeout: 265 seconds)
21:06:28  <wolfeidau>isaacs: TooTallNate: Thanks will have a look around and see what i find.
21:07:57  <isaacs>tjfontaine: nope
21:08:20  <tjfontaine>isaacs: k
21:08:22  * isaacsok away for reals now
21:09:18  <tjfontaine>wolfeidau: how do you actually get your node binaries: nodejs.org, platform/distribution provided, 3rd party binary vendor?
21:09:29  * EhevuTov_quit (Quit: Leaving)
21:09:47  * EhevuTovjoined
21:10:06  <wolfeidau>tjfontaine: Used nvm initially then just to be sure grabbed them directly from the site and tried it
21:11:34  <wolfeidau>tjfontaine: To explain myself fully, I was testing on ubuntu box which i had downlaoded the sources and ./configure && make ect but once I wanted to test someone else ran it on a system which had used npm to install node, hence a bit of my initial confusion
21:11:35  <TooTallNate>different compiler versions probably have different default flags
21:11:38  <TooTallNate>wolfeidau: what OS?
21:11:42  * `3rdEdenquit (Remote host closed the connection)
21:11:46  <TooTallNate>oh, ubuntu
21:11:58  <wolfeidau>TooTallNate: I replicated the issue on both 1004 and 1210
21:12:29  <TooTallNate>wolfeidau: you can check what gcc version was used at least
21:12:32  <TooTallNate>in the process.config object
21:12:54  <wolfeidau>TooTallNate: I will use the make target isaacs mentioned and be sure my assumptions about the flags are correct
21:12:54  <tjfontaine>TooTallNate: ya I suspect the build flavor vs deploy flavor is where the difference is
21:13:44  <TooTallNate>anyways, it sound like the precompiled binary "does the right thing" wrt buffer overflows, so that's good at least
21:13:53  <wolfeidau>Yes this only happened so far on linux, tried solaris and osx and they didn't have the same buffer overflow protection as far as i could tell
21:13:54  <TooTallNate>damn compiler flags
21:13:59  <TooTallNate>so many possibilities :D
21:14:22  <wolfeidau>I was deving on OSX so hence further confusion
21:14:56  <wolfeidau>And the exit is pretty blunt, "buffer overflow detected" with very little periphal information
21:14:59  <tjfontaine>wolfeidau: to make it worse some people may use platform provided binaries, which can have an entirely different configuration
21:15:25  <TooTallNate>especially with 0.9.4
21:15:34  <TooTallNate>essentially all deps can be configured to be shared libs
21:15:37  <wolfeidau>tjfontaine: Yeah especially when they take it on themselves to "tweak" the build process
21:15:42  <TooTallNate>which the debians *will* do
21:15:46  <wolfeidau>haha
21:15:56  <tjfontaine>TooTallNate: and rh/fedora
21:19:41  * EhevuTov_joined
21:22:25  * TooTallNatequit (Ping timeout: 252 seconds)
21:23:00  * EhevuTovquit (Ping timeout: 264 seconds)
21:25:34  * EhevuTov__joined
21:28:50  * EhevuTov_quit (Ping timeout: 250 seconds)
21:36:41  * AvianFlu_joined
21:36:59  * AvianFluquit (Read error: Connection reset by peer)
21:37:05  * AvianFlu_changed nick to AvianFlu
21:41:48  * AvianFlu_joined
21:41:59  * AvianFluquit (Disconnected by services)
21:42:00  * AvianFlu_changed nick to AvianFlu
21:45:16  * AvianFlu_joined
21:45:27  * AvianFluquit (Disconnected by services)
21:45:31  * AvianFlu_changed nick to AvianFlu
21:49:09  * Benviejoined
21:55:57  * bradleymeckquit (Ping timeout: 252 seconds)
21:58:45  * rendarquit
22:03:23  * bulatshakirzyanojoined
22:03:24  * bulatshakirzyanoquit (Excess Flood)
22:10:36  * indexzerojoined
22:15:44  * brsonquit (Ping timeout: 260 seconds)
22:16:48  * Chip_Zeroquit (Ping timeout: 245 seconds)
22:19:08  * Chip_Zerojoined
22:29:45  * c4miloquit (Remote host closed the connection)
22:49:57  * brsonjoined
22:50:03  * brsonquit (Client Quit)
22:50:10  * brsonjoined
22:59:36  * TheJHquit (Ping timeout: 245 seconds)