Camelia, the Perl 6 bug

IRC log for #gluster-dev, 2013-09-06

| Channels | #gluster-dev index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
01:35 awheeler joined #gluster-dev
02:28 aravindavk joined #gluster-dev
02:33 lalatenduM joined #gluster-dev
02:34 ajha joined #gluster-dev
02:58 lpabon joined #gluster-dev
03:08 kshlm joined #gluster-dev
03:11 bulde joined #gluster-dev
03:15 shubhendu joined #gluster-dev
03:40 bharata-rao joined #gluster-dev
03:53 itisravi joined #gluster-dev
03:56 hagarth joined #gluster-dev
04:00 bulde joined #gluster-dev
04:02 mohankumar joined #gluster-dev
04:17 johnmark joined #gluster-dev
04:28 ababu joined #gluster-dev
04:41 ppai joined #gluster-dev
04:45 lala_ joined #gluster-dev
04:45 awheeler joined #gluster-dev
04:46 ndarshan joined #gluster-dev
04:53 kanagaraj joined #gluster-dev
04:55 badone joined #gluster-dev
05:16 spandit joined #gluster-dev
05:17 hagarth joined #gluster-dev
05:21 bala joined #gluster-dev
05:41 bulde joined #gluster-dev
05:52 raghu joined #gluster-dev
06:07 mohankumar joined #gluster-dev
06:13 rgustafs joined #gluster-dev
06:35 hagarth joined #gluster-dev
06:47 deepakcs joined #gluster-dev
07:14 vshankar joined #gluster-dev
07:53 hagarth joined #gluster-dev
08:34 ababu joined #gluster-dev
08:35 bulde joined #gluster-dev
08:38 bulde1 joined #gluster-dev
09:08 bala joined #gluster-dev
09:24 mohankumar joined #gluster-dev
09:48 edward1 joined #gluster-dev
10:04 bulde joined #gluster-dev
10:58 bala joined #gluster-dev
10:59 bala joined #gluster-dev
11:12 ppai joined #gluster-dev
11:37 bala joined #gluster-dev
12:03 kkeithley hagarth_: ping
12:20 kkeithley a2,avati: ping
12:57 awheeler joined #gluster-dev
12:58 bulde joined #gluster-dev
12:58 awheele__ joined #gluster-dev
12:59 hagarth joined #gluster-dev
13:00 awheeler joined #gluster-dev
13:01 jclift joined #gluster-dev
13:02 awheeler joined #gluster-dev
13:18 lpabon joined #gluster-dev
13:41 bulde joined #gluster-dev
13:43 ndevos hey bulde, do you have the permissions to revert changes like http://review.gluster.org/5828 ?
13:44 ndevos bulde: that change seems to break the regression tests :-/
13:46 hagarth joined #gluster-dev
13:49 ndevos bulde: never mind, found a vbellur
13:49 hagarth done
13:57 lpabon ndevos: ping
13:57 ndevos hagarth: thanks!
13:57 ndevos kkeithley: ^
13:57 ndevos lpabon: pong - was just going to respond to your note :)
13:58 ndevos fallocate fails for me as well, but I'm testing on glusterfs-3.4 and I assumed that fallocate has only been included in the master branch. But, that's not something that should be fixed with this commit.
13:58 lpabon :)
13:58 ndevos from gluster import gfapi
13:58 ndevos looks surely nicer, but that requires the /usr/lib/python*/site-packages/gluster/__init__.py to be available when either gluster-swift or glusterfs-api is installed.
13:58 ndevos Louis, what package should contain that file, in your opinion?
13:58 lpabon hmmm
13:58 kkeithley yes, I saw it. thanks hagarth
13:59 lpabon can both have it?  ours is an empty file
14:00 ndevos lpabon: it would be an empty file in both, but I expect that installing both packages would be impossible due to a conflicting file :-/
14:00 lpabon doh
14:00 lpabon bummer
14:01 ndevos lpabon: what glusterfs packages does the swift stack rely upon?
14:01 kkeithley why not put it in glusterfs-api, since gluster*-swift will require glusterfs-api?
14:01 kkeithley will require glusterfs-api anyway, right?
14:01 lpabon kkeithley: good point
14:01 ajha joined #gluster-dev
14:02 ndevos lpabon: I can do that, but there might be a timespan that the conflict exists, releasing both rpms should be a coordinated effort then
14:03 lpabon what about this crazy idea:  What about including gfapi in gluster-swift?  I would like to assign someone from the gluster-swift team to work on hardening gfapi.py
14:03 lpabon we need unit tests, functional tests, etc
14:03 ndevos nah, not everyone would want to install gluster-swift when they want to use the gfapi.py
14:03 lpabon true
14:04 kkeithley like Samba, Hadoop, and eventually Ganesha
14:04 kkeithley oh, and oVirt/qemu-kvm
14:04 ndevos well, they dont use gfapi.py (only libgfapi.so*)
14:04 lpabon true, ok, maybe then call it the python lib:  glusterfs instead of gluster ?
14:05 ndevos sounds good to me too
14:05 ndevos and put it in glusterfs-api
14:05 kkeithley sounds like there ought to be a glusterfs-swift-gfapi RPM???
14:05 lpabon yes
14:05 lpabon oooo  O.o ..maybe
14:06 ndevos we can add 'conflicts: gluster-swift < x.y.z' in glusterfs-api and 'requires: glustefs-api >= a.b.c' in gluster-swift
14:06 lpabon i'm up for whatever makes it easier
14:07 ndevos I dont like a glusterfs-swift-gfapi RPM, there are so many RPMs already
14:08 lpabon ok, im confused on what decision we have come up with... ndevos do you mind summarizing
14:08 ndevos kkeithley: what is your opinion on such a conflicts/requires?
14:09 ndevos lpabon: what package (and latest version/release) does contain that /usr/lib/python*/site-packages/gluster/__init__.py ?
14:10 wushudoin joined #gluster-dev
14:11 lpabon both v1.8.0 (grizzly branch) and v1.9.1(master branch) of gluser-swift  ( https://github.com/gluster/gluster-swift )
14:11 ndevos and those contain .spec files?
14:11 ndevos ah, yes :)
14:12 lpabon yep
14:12 bulde ndevos: i don't have the permissions yet
14:12 ndevos bulde: okay, good to know, I wont bother asking you again ;)
14:13 lpabon SRPMS for gluster-swift are available on build.gluster.org jobs named gluster-swift-build-*
14:15 lpabon kkeithley: quick question, should the gfapi.py be a package in fedora called, something like, python-glusterfs... ?
14:16 ndevos haha, we could do that, but it's still an extra package that will always require glusterfs-api, so I'm for including gfapi.py in that
14:18 lpabon yeah, i'm not sure if we will escape that one
14:20 ndevos hmm, I'll put some notes together
14:20 kkeithley I was suggesting a subpackage of the glusterfs-openstack-swift  package. Kinda like the glusterfs package has glusterfs, glusterfs-server, glusterfs-api, etc., RPMs. But if it's just to have an empty /usr/lib/python*/site-packages/gluster/__init__.py file then no need.
14:22 lpabon *if* it was part of a gluster-swift package, and *if* it was part of the gluster-swift repo, it would benefit from the unittest and functional test python infrastructure already implemented.  There is a lot of python expertise in gluster-swift... just a suggestion
14:23 lpabon but i can definitely see the counterpoint where gfapi.py really does not depend on gluster-swift
14:23 ndevos and, gfapi.py heavily depends on libgfapi.so, splitting those over two projects does not have my preference
14:23 lpabon agreed
14:24 lpabon we can setup a testing infrastructure inside glusterfs.git
14:24 lpabon for gfapi.py i mean
14:25 ndevos lpabon, kkeithley: how about http://titanpad.com/GpLiXcFsxf
14:26 lpabon Aye!
14:26 lpabon +1
14:26 ndevos lpabon: a testing framework could be part of the standard <glusterfs.git>/tests/bugs/...t files, they could call specific .py files when a new operation (like fallocate) gets added
14:27 lpabon i'm not familiar with those, but I can see how they can leverage using 'nose' and 'tox'
14:28 johnmark semiosis: hey
14:28 lpabon I have to check, but there is a new hire coming soon, and we may assign him to stabilize and harden the gfapi.py code
14:29 johnmark lpabon: w00t
14:29 ndevos lpabon: oh, that testing is pretty basic, but it is quite flexible, extending the framework to call nose/nox (whatever they are) should not be too difficult
14:29 lpabon ok
14:30 lpabon i'm good then
14:31 lpabon johnmark: are going to be in the office next week?
14:31 ndevos lpabon: okay, so, if you like the plan, and kkeithley agrees, I'll update the patch to allign the change
14:31 lpabon ndevos: sounds good
14:33 ndevos johnmark: Obama was blocking my way to the trainstation yesterday, almost missed my flight, luckily he delayed that as well
14:33 kkeithley just looking at titanpad. Looks okay to me
14:34 johnmark lpabon: only on Friday
14:34 kkeithley ndevos: the nerve of that guy. ;-)
14:34 johnmark I have to head out to SF for most of hte week
14:34 johnmark kkeithley: lulz
14:34 ndevos kkeithley: yeah, but its impressive how many cars he needs!
14:35 lpabon johnmark: wanted to see if i could spend some time with you to review the openstack meetup presentation
14:36 kkeithley three rings for elven kings, seven for the dwarf lords, eleven for mortal men, and 50 Chevy Suburbans for the POTUS
14:42 kkeithley ndevos, lpabon: the old glusterfs-ufo package used .../site-packages/gluster (including .../site-packages/gluster/swift/*)  I gather that glusterfs-openstack-swift uses .../site-packages/gluster.swift, is that correct?
14:42 kkeithley 50 Chevy Suburbans and two Boeing 747s.
14:42 lpabon yes
14:42 bala joined #gluster-dev
14:42 lpabon it uses gluster/swift/.... as well
14:43 kkeithley so gluster.swift is different than gluster/swift
14:43 lpabon yes is different, gluster-swift uses the same as the glusterfs-ufo package used to
14:44 lpabon which is ../site-packages/gluster/swift/
14:44 ndevos hmm, .../site-packages/gluster.swift doesnt look very pythony to me
15:06 ndevos lpabon, kkeithley: would you like to see a 'provides: python-glusterfs = $version' in the .spec file? or shall I leave that out?
15:07 lpabon i will leave up to kkeithley
15:07 bulde joined #gluster-dev
15:08 ndevos lpabon: you would be the user of that, otherwise you would need to add a 'requires: glusterfs-api' (with optional version)
15:08 lpabon ah, yes, require: python-glusterfs would be nice
15:09 ndevos lpabon: okay, you got it :)
15:11 lpabon awesome
15:14 * ndevos will enjoy the weekend now, see you next week again!
15:55 kkeithley Okay, too bad ndevos is gone, but...   portante and I think lpabon is conflating python 'import gluster.swift' with installed directories in .../site-packages/gluster/swift/.... and that there really isn't a .../site-packages/gluster.swift directory
15:55 portante yes
15:55 kkeithley And....  for gfapy.py, portante and I agree that we should not add a new .../site-packages/glusterfs directory, we should put gfapy.py in  .../site-packages/gluster/
15:55 portante I agree with kaleb
15:55 portante yes I agree with kaleb
15:56 * portante this is actually kaleb typing at my console ;)
15:56 * kkeithley has really long arms
15:56 portante he is part of the fantastic four, actually
15:57 * kkeithley can type at my own console on the third floor and portante's console on the second floor at the same time
16:37 bala joined #gluster-dev
16:58 lalatenduM joined #gluster-dev
16:58 Technicool joined #gluster-dev
17:26 kkeithley ruhro, rots of core files piling up in / on build.gluster.org
17:41 bulde joined #gluster-dev
21:47 JoeJulian If there was any way to be sure that people would get sarcasm, I'd write an NSA translator....

| Channels | #gluster-dev index | Today | | Search | Google Search | Plain-Text | summary