17:08 PM
TE-msmith
visualizing is brianhauer_'s department
17:09 PM
hamiltont
although I'll probably have to up the sampling frequency for any pattern recognition stuff, 1 sample per second is a bit slow for detecting things like CPU bottlenecks
17:09 PM
hamiltont sighs. To have a computer and a free month
17:34 PM
http://imgur.com/Xb5Pi5r - calls into alloc during a JSON benchmark of openresty. You can clearly see the slow warmup, the 256 warmup, and then the actual 256 benchmark
17:52 PM
FWBM-bot
17:52 PM
FrameworkBenchmarks/1172 6c37b11 msmith-techempower: Fixed a bug with installing WT incorrectly
17:52 PM
[FrameworkBenchmarks] msmith-techempower opened pull request #1177: Fixed a bug with installing WT incorrectly (master...1172)
http://git.io/ju9_9A
17:56 PM
17:56 PM
FrameworkBenchmarks/master 89da119 Mike Smith: Merge pull request #1177 from TechEmpower/1172...
18:07 PM
TE-msmith
neat
18:07 PM
I'm doing things!
18:09 PM
I need to spin up an ubuntu12 vm to figure out why everything is busted
18:10 PM
18:10 PM
"cowboy is busted" - that log
18:10 PM
but when I run it on my ubuntu14 hardware, it works perfect
18:10 PM
yay!
18:11 PM
it's like... rebar.beam is the compiled... majig
18:13 PM
hamiltont
TE-msmith: Our vagrant setup can spin up an ubuntu 12 image for you, if your line endings stuff is resolved
18:13 PM
TE-msmith
and you know it's not!
18:13 PM
lol
18:13 PM
hamiltont
;-)
18:13 PM
TE-msmith
sooooooo
18:13 PM
Loading of /home/travis/build/TechEmpower/FrameworkBenchmarks/frameworks/Erlang/cowboy/rebar/rebar/ebin/rebar.beam failed: badfile
18:13 PM
...yeah, that's gibberish
18:14 PM
how did it come up with that file?
18:22 PM
yeah, okay, this is (again) the problem that rebar is built on ubuntu14 and we're testing with ubuntu12
18:22 PM
....SHOOOOOULD we make this another depends?
18:22 PM
fw_depends erlang rebar
18:22 PM
et
18:23 PM
hamiltont
can you update rebar.sh to support both 12 and 14?
18:23 PM
or no?
18:23 PM
TE-msmith
I'm still confused about the whole thing
18:24 PM
okay
18:24 PM
got it
18:24 PM
the version of rebar included with cowboy (wrong, should be a dependency) is built with a newer version of Erlang than exists for ubuntu12
18:25 PM
I think we need to make rebar into a dependency and just build it manually
18:25 PM
hamiltont
ok
18:25 PM
Is it statically including a binary currently?
18:25 PM
like, in our git repo?
18:25 PM
TE-msmith
yes
18:26 PM
18:26 PM
`rebar` is a binary file
18:26 PM
hamiltont
ouch
18:26 PM
yea I can see why that's causing issues
18:36 PM
TE-msmith
welcome back!
18:38 PM
hamiltont: does fw_untar allow me to specify the output filename?
18:38 PM
doesn't look like it
18:50 PM
hamiltont
You can pass it any extra args you'd pass to tar IIRC
18:50 PM
e.f. fw_untar foobar.tgz -C put_it_here
18:50 PM
IIRC...
18:58 PM
TE-msmith
neat
19:03 PM
FWBM-bot
19:03 PM
FrameworkBenchmarks/1178 cad05eb msmith-techempower: Fixed the rebar issue...
19:04 PM
[FrameworkBenchmarks] msmith-techempower opened pull request #1179: Fixed the rebar issue (master...1178)
http://git.io/2y9IsA
19:05 PM
19:05 PM
FrameworkBenchmarks/master 6ad4a0f Mike Smith: Merge pull request #1179 from TechEmpower/1178...
19:21 PM
[FrameworkBenchmarks] stefanocasazza opened pull request #1180: ULib fix - substitute of (#1083 #1122) (master...master)
http://git.io/mVAmgQ
19:22 PM
[FrameworkBenchmarks] stefanocasazza closed pull request #1122: ULib fix - substitute for #1083 (master...master)
http://git.io/LGDGhw
19:50 PM
[FrameworkBenchmarks] msmith-techempower closed pull request #1180: ULib fix - substitute of (#1083 #1122) (master...master)
http://git.io/mVAmgQ
19:50 PM
19:50 PM
FrameworkBenchmarks/master 7326ea3 stefanocasazza: ULib fix - substitute of (#1083 #1122)
19:50 PM
FrameworkBenchmarks/master db9a10c stefano casazza: Update .travis.yml
19:50 PM
FrameworkBenchmarks/master 962207c stefanocasazza: DO NOT MERGE
19:57 PM
acidx
do not merge? whoops?
20:50 PM
TE-msmith
lol
20:51 PM
I tested it out; seemed like it worked fine
20:51 PM
oh, that history is from the front, not the back
20:52 PM
weird
21:18 PM
codygman- has quit
21:18 PM
codygman has quit
21:19 PM
the Golang tests are insane
21:19 PM
the setup is wacky
21:24 PM
hamiltont
acidx: I think he's copying a convention I use :-)
21:25 PM
If a commit does something like change .travis.yml to make it easier for me to debug, I name it "DO NOT MERGE"
21:25 PM
helps me remember to rebase and delete it before i send a PR
21:28 PM
TE-msmith
meh, i merged it because yolo
21:28 PM
hamiltont: it seems that FWROOT is always going to be ~/FrameworkBenchmarks
21:28 PM
right?
21:28 PM
hamiltont
ah
21:28 PM
no
21:28 PM
that's the entire ponit of FWROOT, that's it not always ~/FrameworkBenchmarks
21:28 PM
TE-msmith
my confusion
21:28 PM
sorry, but that's the intent
21:28 PM
for my setup, it will always be that value, right?
21:28 PM
hamiltont
huh?
21:29 PM
On your testing machine, that's probably true
21:29 PM
but did you merge somethign into master that hardcodes that?
21:29 PM
TE-msmith
no
21:30 PM
hamiltont
ah ok
21:30 PM
TE-msmith
I misunderstood something you said a while ago
21:30 PM
hamiltont
sorry, had me worried
21:30 PM
TE-msmith
I THOUGHT it was supposed to be ~/FrameworkBenchmarks/frameworks/[lang]/
21:30 PM
hamiltont
Ohhh
21:30 PM
TE-msmith
so I could always do something like $FWROOT/revel
21:30 PM
(cannot do this)
21:30 PM
hamiltont
Ok, my bad. Misunderstood you
21:30 PM
TE-msmith
honestly though
21:30 PM
help me with the $GOPATH stuff
21:30 PM
it makes no sense
21:30 PM
hamiltont
Yea, you cannot. Need to use $FWROOT/frameworks/lang/etc
21:30 PM
TE-msmith
21:30 PM
yeah, that's fine
21:30 PM
hamiltont
Dude I had the *same* problem
21:31 PM
I tried to fix go
21:31 PM
got lost with gopath too
21:31 PM
TE-msmith
I've got a bunch of go frameworks working
21:31 PM
hamiltont
it doens't seem to work like it claims
21:31 PM
TE-msmith
falcon's busted forever (might remove it because no one is maintaining it)
21:31 PM
RIGHT?
21:31 PM
here's some err output: /home/techempower/FrameworkBenchmarks/frameworks/Go/revel/src/bin/benchmark (from $GOPATH)
21:32 PM
but I `print os.environ["GOPATH"]` before that line and it spits out: /home/techempower/FrameworkBenchmarks/Go/revel/
21:32 PM
so... it's looking in $GOPATH/src/bin ?
21:32 PM
(it's looking for `benchmark`)
21:32 PM
why would it do that!?
21:33 PM
and from that link I just sent, it says
21:33 PM
The source for a package with the import path "X/Y/Z" is in the directory $GOPATH/src/X/Y/Z
21:33 PM
The binary for a command whose source is in $GOPATH/src/A/B is $GOPATH/bin/B
21:33 PM
LIES GOLANG
21:33 PM
LIES
21:33 PM
hamiltont
looking to see if I spot anything
21:33 PM
TE-msmith
ALL LIES
21:34 PM
21:34 PM
it has the same layout there
21:34 PM
what's the deal?
21:35 PM
hamiltont
21:36 PM
TE-msmith
yeah, that's wrong
21:36 PM
lol
21:36 PM
hamiltont
but I don't think I set GOPATH correctly there, seems like it should be export GOPATH=$TROOT
21:36 PM
TE-msmith
(what's TROOT?)
21:36 PM
hamiltont
TROOT = $FWROOT/frameworks/lang/your-test
21:36 PM
TE-msmith
nice
21:36 PM
yes
21:36 PM
exactly that
21:36 PM
I think
21:36 PM
I'm working on that
21:36 PM
didn't know TROOT was a thing
21:36 PM
hamiltont
e.g. /home/techempower/fwbm/go/revel
21:37 PM
Yea, these vars need docs. Sorry bro
21:37 PM
TE-msmith
s'okay bro
21:37 PM
you've helped so much that I will bench the knee to these ones
21:37 PM
FIXED IT
21:37 PM
YAY!!!
21:37 PM
hamiltont
ha ha really?
21:37 PM
sick, that was fact
21:37 PM
TE-msmith
yar
21:37 PM
hamiltont
fast*
21:37 PM
TE-msmith
i've been working on them
21:37 PM
lol
21:37 PM
hamiltont
'synergy'
21:38 PM
TE-msmith
indeed