Friday, 2014-08-15

atiwariwoodster, there are some test failure due to
atiwariI will fix it tomorrow first thing00:19
*** atiwari has quit IRC00:19
*** alee has joined #openstack-barbican01:14
*** crc32 has quit IRC01:33
*** woodster has quit IRC01:55
*** woodster_ has joined #openstack-barbican02:06
*** uberj has quit IRC03:14
*** uberj has joined #openstack-barbican03:14
openstackgerritPaul Kehrer proposed a change to openstack/barbican: change CryptoPluginManager to be instantiated in the module scope
reaperhulkcircular import fun. let's see what we get out of devstack now04:36
woodster_I've been fighting C dependencies with my tox env04:43
reaperhulkwhat OS?04:44
woodster_macos, mavericks04:45
reaperhulkdo you have the dev tools installed? try looking at /usr/include/ffi/ffi.h04:45
reaperhulkshould be present by default04:45
rm_workT_T still fighting that woodster_?04:45
rm_workwhat's the full trace of the failure?04:45
reaperhulkI tried installing uwsgi tonight and it reminded me why I hate it04:45
reaperhulkClaiming I don't even have a compiler04:45
woodster_ha, I reinstalled xcode per a forum entry I saw04:46
rm_workwoodster_: I was thinking about reformatting this machine with mavs04:46
reaperhulkwhy bother, wait a few months and go straight yosemite04:46
rm_workif i do, i will have you to tell me how to fix it I guess :P04:46
rm_workreaperhulk: that's an option too i guess04:46
reaperhulkand then keep up to date in the future because the past sucks!04:47
reaperhulkwoodster_ if it's still being weird what happens if you CFLAGS="/usr/include" tox -r -e py27 (or CFLAGS before a pip install if you can replicate without tox's involvement)04:48
woodster_that inline pastebin is pretty cool04:48
reaperhulkso cffi is failing to build because it can't find ffi.h, which is pretty stupid04:48
woodster_this was all working on mavericks I thought, but might have been fooling my self04:48
reaperhulkit definitely should be. I do all dev on mavericks :)04:49
rm_you|yeah... wonder where it's looking04:49
reaperhulkso, first question, is ffi.h present where I said?04:49
reaperhulknext, what version of pip and setuptools are you running?04:49
rm_you|i may have used brew to install cffi / dev04:49
reaperhulknot required for libffi04:49
rm_you|on mine04:49
*** rm_you| is now known as rm_you04:50
reaperhulkand finally, what happens if you do CFLAGS="/usr/include" pip install -U cryptography04:50
reaperhulklibffi in brew installs in cellar only due to the pre-existence of libffi on OS X04:51
reaperhulkso you have to explicitly link against it (/usr/local/Cellar/libffi)04:51
reaperhulksame as the openssl formula and a bunch of other things04:51
woodster_should be libffi correct? 3.0.13 is the version I see there04:52
reaperhulkin the cellar? That's not relevant for the moment :)04:52
reaperhulkdoes /usr/include/ffi/ffi.h exist?04:52
* reaperhulk is actually working on getting cryptography building against libressl right now04:53
reaperhulkjerks changed some function signatures04:53
reaperhulkand cffi sucks at handling that04:53
woodster_ /usr/include/ffi/ffi.h does not exist. Yeah libffi is in cellar04:54
reaperhulkweeell interesting04:54
reaperhulkokay do this for me04:54
reaperhulkxcode-select --install04:54
reaperhulkshould pop up a dialog, download some stuff, and install it04:55
woodster_1.4.1 of pip04:55
woodster_yes, it is downloading stuff04:55
reaperhulkonce that installs check and see if /usr/include/ffi/ffi.h exists04:56
woodster_ha, yep I see it now04:57
reaperhulkokay try to compile cryptography now :)04:57
reaperhulkthat's fine04:59
woodster_2 other 'safely ignored' errors after that one :)04:59
reaperhulkstandard cffi BS04:59
woodster_but it ran to success04:59
reaperhulkyep you should be fine now04:59
woodster_running tox -e pep8 now...05:00
woodster_thanks for the help!05:00
woodster_so too much wind brought you back? :)05:00
reaperhulkbtw rm_you, probably doing a 0.5.4 tomorrow including crc32a's backports.05:00
reaperhulkI kited for about an hour or so. Feet getting torn up so I came back. Now they're covered in tape, ha05:01
reaperhulkwarning, boring photo of feet.05:01
rm_youreaperhulk: ok cool05:01
rm_youreaperhulk: (the release, not the feet, the feet are average) :P05:02
reaperhulknot sure what the story is on the pyopenssl release, but there's a bug that needs to be fixed so someone is going to push at exarkun a bit05:02
rm_youthough we need exarkun to get back so we can actually get the pyopenssl changes in05:02
rm_youlol jinx05:02
woodster_oh wow, so even though board is slick, the pressure with your feet is probably the issue?05:02
reaperhulkThere are bindings you put your feet in and the waves can make it choppy so you rub against them05:03
woodster_rm_you: are you becoming a late nite cryptohead too? Poor life choices are contageous!05:03
rm_youi'm just a standard IRC addict05:03
rm_youcan't help myself <_<05:03
rm_youand since i didn't do much today other than get a decent devstack deploy script up... i am looking at the client again05:04
rm_youthough our deadlines are now... Kilo :P05:04
rm_youso whatev05:04
rm_youneutron incubator shit happened05:04
rm_youso now we won't get into Juno like we hopes05:04
woodster_oh man, really? So the pressure is off, or still trying to get your stuff done at least?05:06
rm_youyeah now we're still working but... no code freezes, no rush to merge05:07
rm_youand Kilo is the soonest we could possibly see release of the features :/05:07
reaperhulkIf you stop rushing it won't make it to kilo since that's how openstack works ;)05:09
*** juantwo has quit IRC05:09
woodster_yeah, set your milestone 6 months before you expect it landed/used05:12
woodster_alright folks, I'm heading off to bed. Thanks for you help Paul!05:16
reaperhulkno problem, goodnight05:17
reaperhulkbah, I will look at this more tomorrow. This looks suspiciously like import order dependency, which is super terrible05:24
reaperhulkI hope I'm wrong05:24
*** jamielennox is now known as jamielennox|away05:35
openstackgerritOpenStack Proposal Bot proposed a change to openstack/barbican: Imported Translations from Transifex
*** woodster_ has quit IRC07:45
*** juantwo has joined #openstack-barbican12:01
denis_makogonHye, guys, i've did  barbican setup using DevStack, and trying to examine API  and that's what i get PolicyNotAuthorized: Policy doesn't allow orders:get to be performed. Could someone help?12:25
denis_makogonAlso, i've noticed this bug (on fresh new install of Barbican + DevStack)12:43
*** russellb is now known as rustlebee13:03
*** nkinder has quit IRC13:10
*** akoneru has joined #openstack-barbican13:14
*** akoneru has quit IRC13:18
*** nkinder has joined #openstack-barbican14:00
*** denis_makogon has quit IRC14:36
*** paul_glass has joined #openstack-barbican14:51
*** atiwari has joined #openstack-barbican14:54
*** atiwari has quit IRC15:06
*** rellerreller has joined #openstack-barbican15:13
*** denis_makogon has joined #openstack-barbican15:17
*** atiwari has joined #openstack-barbican15:27
openstackgerritArvind Tiwari proposed a change to openstack/barbican: Add more type in order post
openstackgerritTim Kelsey proposed a change to openstack/barbican-specs: blueprint for adding HP Atalla ESKM crypto plugin.
openstackgerritTim Kelsey proposed a change to openstack/barbican-specs: blueprint for adding HP Atalla ESKM crypto plugin.
*** akoneru has joined #openstack-barbican16:02
rm_workdenis_makogon: hmm, i'll have to test the client on my devstack today16:10
rm_workwhen did you setup the devstack?16:11
denis_makogonrm_work, like few hours ago16:11
rm_workwe just fixed something regarding API version yesterday… :/16:11
rm_workbut you must have that if you just did it an hour or two agop16:11
denis_makogonrm_work, everything was made according to, except the number of enabled services16:12
rm_workk yeah16:12
rm_work(I wrote the script under "Easy Way", btw :P16:13
denis_makogonrm_work, and also, i've modified the policy.json to accept requests from all users16:13
denis_makogonrm_work, i saw it, as i can see, in that script was added tempest installation, right?16:14
rm_worktesting now16:14
denis_makogonrm_work, i'm up to help to fix it =)16:14
rm_workwell, i'm also rewriting the entire client this next week anyway :P16:15
rm_workswitching from the current method to using objects that you create, set up, and then .save()16:15
rm_workinstead of secret_manager.creat(...)16:15
rm_workit'd be my_secret = Secret()16:16
rm_worksecret.payload = ...16:16
rm_workerr, my_secret.payload = ...16:16
rm_workat least, I will be once redrobot finishes the bP16:16
denis_makogonrm_work, nice =)16:17
denis_makogonwell, i've also noticed that current client doesn't follow common CLI style like other clients do16:17
denis_makogonrm_work, eg. "order create" should be "order-create"16:18
denis_makogonand so on]16:18
*** rellerreller has quit IRC16:18
rm_workyeah, not sure how much we'll be messing with the frontend, but feel free to put those comments on the BP once redrobot gets it posted16:19
denis_makogonrm_work, i've filed a bug-report for that, i wasn't aware about BP being in progress16:21
rm_workit's… not16:21
rm_workhe still needs to post it16:21
rm_workwhich is why you wouldn't have been aware :P16:21
denis_makogonrm_work, hah, sounds amazing =)16:22
denis_makogonrm_work, another question16:23
denis_makogonrm_work, do you plan to add API to query existing algorithms?16:23
denis_makogonalgorithms and its modes (like AES and its 5 modes)16:24
rm_workthat'd be a redrobot question16:24
rm_workI'm just a lackey right now on loan from LBaaS :P16:24
rm_workI get "barbican 2.2.1" from --version16:27
rm_workon my devstack install16:28
redrobotdenis_makogon we've talked about adding querying capability, but we haven't decided on anything yet16:28
redrobotrm_work I think I may have some time to work on the BP today16:28
denis_makogonrm_work, that's strange16:28
denis_makogonrm_work, how it could be 2.2.1 when actual version is 2.0 ?16:29
rm_workthat's from the server i think?16:30
denis_makogonrm_work, server version is 2014.216:30
rm_workthough, dunno for sure16:30
rm_workyeah i see that too...16:30
redrobotdenis_makogon see tags
rm_workah, you do that same tag magic16:30
redrobotwe release to PyPI via openstack infra16:31
redrobotby adding a new tag to the git tree16:31
denis_makogonredrobot, true, github releases shows the same16:31
rm_workdoes that translate through if you use barbicanclient from pypi?16:31
rm_worki'm using the git version16:31
rm_workso it makes sense that mine sees 2.2.116:31
redrobotyeah, I think PBR is looking at the latest tag16:31
redrobotnot sure what the deal is with the outdated version in the setup.cfg though16:32
rm_workyeah, though PBR is royally meessing up on barbican main tho16:32
rm_worksee: your setup.cfg revert :P16:32
denis_makogonwill try to install it again16:33
denis_makogonhere's my local.conf
redrobotrm_work loonch?16:36
denis_makogonomnomnom time, for sure =)16:36
rm_workredrobot: hmm16:36
rm_worki still am not sure on lunch plans16:36
rm_worksaucer got cancelled on me16:36
rm_workneed to get ready to like… got to work tho16:37
rm_workif you guys aren't going to lunch until 12:15 or so i'm down, but that's about the soonest i can get in (unless i meet somewhere)16:37
denis_makogonrm_work, trying your "easy way", lets see what it would16:45
rm_workyeah that only gives you essentially barbican and keystone :P16:48
rm_workbut it should work16:48
rm_work(when testing I went to /opt/stack/python-barbicanclient to run the barbican command)16:49
rm_workah and I am going to update that script today to also add some exports to /opt/stack/.bashrc for the endpoints and auth info, forgot to do that oart yesterday16:49
rm_workexport OS_TENANT_NAME=admin16:50
rm_workexport OS_USERNAME=admin16:50
rm_workexport OS_PASSWORD=password16:50
rm_workexport OS_AUTH_URL="http://localhost:5000/v2.0/"16:50
rm_workexport BARBICAN_ENDPOINT="http://localhost:9311/v1/"16:50
rm_workin the meantime, it's those16:50
arunkantHi..looking for review on ..16:52
rm_workthere, fixed16:53
rm_workredrobot: were you going to lunch?16:53
rm_workand if so, was your loonch going to be somewhere i could meet you? :P otherwise I have no idea16:53
*** nkinder has quit IRC16:56
rm_workredrobot: guessing you've already gone then16:57
*** openstackgerrit has quit IRC17:02
*** paul_glass has quit IRC17:02
*** openstackgerrit has joined #openstack-barbican17:03
*** nkinder has joined #openstack-barbican17:08
* rm_work goes to Bigz17:08
openstackgerritKaitlin Farr proposed a change to openstack/barbican: Adds KMIPSecretStore and unit tests
*** paul_glass has joined #openstack-barbican17:46
openstackgerritArvind Tiwari proposed a change to openstack/barbican: Add more type in order post
SheenaG11Hey redrobot, you around?19:05
redrobotSheenaG11 yep, what's up?19:05
SheenaG11Just wanted to kick up to everyone's attention in case you guys hadn't already taken a look at it19:07
SheenaG11From the HP crew19:07
SheenaG11Sorry I'm late to the party19:07
reaperhulklemme drop a comment on that to explain the jenkins failure I guess19:07
SheenaG11Hey reaperhulk!  I didn't know if you'd be around.19:08
SheenaG11Are you waiting for my PR to put a -2?19:08
reaperhulkLying in wait19:08
SheenaG11GO KITING!19:09
SheenaG11I'm waiting until you're gone.19:09
reaperhulkAt least 2 more hours until I go19:10
reaperhulkWant to see the state of my feet?19:10
reaperhulkWhat am I saying, of course you do19:10
reaperhulkWho doesn't like to look at feet.19:10
* redrobot knew better than to click on that link...19:11
* redrobot still clicked anyway19:11
reaperhulkit's not disgusting (unless you dislike feet I guess)19:12
reaperhulkjust...lots of tape.19:12
*** denis_makogon has quit IRC19:12
*** dmakogon_ has joined #openstack-barbican19:13
redrobotlooks like your right pinky toe is about to give out19:13
reaperhulkhaha, actually the tape above that and around the main part of my foot are the biggest issues at the moment19:13
reaperhulkwhole top of my right foot is chafed so anything touching hurts19:13
SheenaG11My legitimate reaction, without filter, was AUGH19:16
SheenaG11What did you do?!19:16
reaperhulkthe bindings on the board eventually start to chafe when you ride around in choppy water and are pushing your feet constantly and bouncing around :)19:25
reaperhulkplus as you get tired your form starts to slip and then things rub more. I really should take a day off but I don't wanna19:26
SheenaG11You don't like having feet very much, do you?19:27
reaperhulkThey should obey me19:27
SheenaG11They aren't computers...19:29
reaperhulksolaris is currently fighting me tooth and nail19:35
reaperhulkbasic things like DNS are ridiculously difficult to configure for some reason19:35
SheenaG11hey hockeynut are you around?19:48
*** nkinder has quit IRC20:26
*** crc32 has joined #openstack-barbican20:58
openstackgerritAde Lee proposed a change to openstack/barbican: Add certificate_manager functionality to dogtag plugin.
*** juantwo has quit IRC21:06
*** paul_glass has quit IRC21:11
rm_workhey redrobot you around?21:16
*** atiwari has quit IRC21:16
redrobotrm_work yep, what up?21:16
rm_workredrobot: lacking a BP, i'm doing a best-shot at what I think the client should look like (basically a rewrite from scratch)21:16
rm_workis there a reason the barbican Client couldn't be a singleton?21:16
redrobotrm_work I'm workign on the blueprint as we speak21:16
rm_workI guess maybe people will need multiple clients from the same app...21:16
rm_workthough it makes it a whole lot cleaner if it is a singleton :P21:17
redrobotcleaner how?21:18
redrobotI don't think any other OS clients work that way21:18
rm_workso let's say I want to make a new secret21:18
rm_workmy_secret = barbican.secrets.Secret()21:18
rm_workmy_secret.payload = "some data"21:18
rm_work... etc21:18
rm_workI either have to pass the Connection or Client or SecretManager or something21:19
rm_workto that object somehow21:19
rm_workif it were a singleton, the object could just... grab it :P21:19
openstackgerritSheena Gregson proposed a change to openstack/barbican: Edited docs to improve context
rm_work(I suppose Singleton in this case is just a less dirty way of saying "global" though <_<)21:20
rm_workgogo Sheena21:20
* rm_work reviews21:20
SheenaG11Keep in mind on this one, guys: I'm not trying to fix *all* the problems in the doc21:20
redrobotwtf just happened?21:20
SheenaG11It's an initial pass to add some sections and improve context21:20
chellygelshit got real!21:20
SheenaG11DOUG I'M A CODER21:21
rm_workdat XML21:21
*** nkinder has joined #openstack-barbican21:21
redrobotGonna -2 so fast you won't know what happened21:21
rm_workthough yeah, -1 right now cause that won't pass gate :P21:22
rm_workyour commit message T_T21:22
*** alee is now known as alee_afk21:22
rm_workunless Barbican's gate doesn't check length on additional lines past the title21:22
*** crc32 has quit IRC21:22
*** akoneru has quit IRC21:23
rm_workguess maybe it does on docs21:23
SheenaG11Lines are too long?21:24
SheenaG11They're under 80 char per line21:24
SheenaG11Pretty sure.21:24
rm_workI just ran one through wc and got 81 :P21:24
SheenaG11Looks like I left whitespace everywhere though21:24
SheenaG11Okay maybe I can't count21:24
rm_workin the commit message, though like i said apparently Barbican doesn't care21:24
rm_workcause tox is passing it locally O_o21:24
rm_worki got my shit kicked back in neutron for that T_T21:25
SheenaG11I ran my tox tests!21:25
rm_workyep, barbican appears to be more forgiving21:25
rm_workand vim says 80 O_o21:26
redrobotrm_work sooooo.... what could we do instead of a global client... ?21:27
rm_workpass the client in :P21:27
redrobotrm_work Maybe instatiation of objects still needs to be done via client?21:27
*** crc32 has joined #openstack-barbican21:27
rm_workmy_secret = my_conn.secrets.Secret()21:27
rm_workah and then get is21:28
rm_workmy_secret = my_conn.secrets.Secret(secret_ref="blag")21:28
redrobotmy_new_secret = barbicanclient.Client(...).secrets.Secret()21:28
redrobotbah... not as pretty21:28
rm_workyou see my dilemma :P21:28
rm_workcould also pass it to the save() >_>21:28
rm_worknot unheard of21:29
chellygelhow do we get around the long lines in xml?21:29
rm_workI've seen things handled that way before21:29
rm_workchellygel: i mean, i think the actual XML is fine :P21:29
redrobotrm_work yeah, I think I prefer that.21:29
rm_workmy comment was just about the commit message21:29
rm_workwhich in neutron they care way too much about, rofl21:30
rm_worki think I like that barbican isn't as snooty21:30
rm_workredrobot: kk21:30
redrobotrm_work  how about my_secret = Secret(client=conn) ?21:30
rm_workthat also works21:30
rm_workbut require it?21:31
rm_workor let them make a secret without a conn21:31
redrobotis there any use in a Secret that can't be saved?21:31
rm_workfunsies? :P21:31
* redrobot shrugs21:32
rm_workunit testing? :)21:32
openstackgerritSheena Gregson proposed a change to openstack/barbican: Edited docs to improve context
SheenaG11Sorry - re-push to fix the white space21:32
SheenaG11rm_work:  you want me to shorten the commit msg or should I try to see if it'll fly?21:32
rm_workit'll fly21:32
redrobotSheenaG11 nice!  I didn't even have to downvote it21:33
rm_workman, i got like 10 lines into this review and remembered how much I hate XML21:33
SheenaG11I got like 10 lines into writing this shit and remembered how much I hate XML21:34
* rm_work does not envy SheenaG11 21:34
reaperhulk-2 so hard21:35
rm_workredrobot: so you think you'll have a draft (of at least one of the object types) by EOD?21:35
SheenaG11I thought you were gone21:35
reaperhulkI am always lurking in the bushes ready to crush dreams21:35
redrobotrm_work yep... you're welcome to come hover and give input21:35
reaperhulkIt's actually my job description21:35
rm_workredrobot: heh I'll just wait for you to pastebin me a draft or something21:35
rm_workI want to get this fleshed out and then see if I already coded what you put in the BP :)21:36
redrobotrm_work well, I was making some good progress until you brought up that client ref :-\21:36
rm_worksorry, pesky real life details21:36
redrobotwhy can't shit just magically work!?!21:36
rm_workthis is why I don't think BPs will ever actually look like the final code21:36
rm_workyou get halfway into coding it and realize whoever wrote the BP was living in fairy-land21:37
reaperhulkXML diffs are the worst21:37
rm_works/diffs are/is/21:37
redrobotthe more I think about it, the more I like client.secrets.Secret(...) for both new and get21:37
redrobotdespite the ugly21:37
reaperhulkSheenaG11, for future reference I'd suggest submitting many smaller separate CRs. (This one is fine for now, but a 1200 line diff is very difficult to review)21:38
redrobotthat way you can pass the unsaved Secret around without passing the conn reference around as well21:38
rm_workit's technically only 600 lines of changes or so21:38
rm_workjust XML diffs are shitty21:38
SheenaG11reaperhulk: it's mostly just stuff that got moved down - I didn't actually write that much21:39
SheenaG11Oh and I deleted A WHOLE BUNCH OF STUFF21:39
rm_workredrobot: yeah, well you can do that with Secret(client=conn) too21:39
reaperhulkSheenaG11: yeah I know, but gerrit's diff tools aren't smart enough to identify moves so there's a lot of noise due to the moves21:39
redrobotrm_work yeah, but then instantiation would be different than getting existing objects21:40
redrobotrm_work unless we want to do  Secret(client=conn, ref="http://blah")21:40
rm_workSecret(client=conn) vs Secret(client=conn, secret_ref="blah")21:40
redrobotneed a tie breaker here21:40
SheenaG11reaperhulk: fair enough21:41
rm_worklike I said, i honestly don't care either way21:41
SheenaG11reaperhulk: sorry for the gigantor CR :-(21:41
reaperhulkSheenaG11: not really a big issue this time around, just something to note for the future21:41
redrobotI suppose client.secrets.Secret() is more in-line with other OS clients?21:41
* reaperhulk -2s it21:41
reaperhulkjust kidding21:41
rm_workSheenaG11: no, you have to redo the whole thing. REDO IT.21:41
redrobotthough not quite21:41
rm_workmoar XML pain for sheena21:41
rm_workredrobot: could be, need to look21:42
rm_workwe model after keystone right now, right?21:42
redrobotrm_work yep21:42
SheenaG11rm_work: I didn't even get any cake!21:42
chellygel+1 no cake, wtf21:42
rm_workyeah sad21:42
rm_workyou and chellygel were OoO21:42
rm_workvery sad21:42
redrobotdon't worry SheenaG11 and chellygel.  I ate y'alls pieces for you.21:43
redrobotyou're welcome21:43
rm_workchellygel still has a lunch IOU tho21:43
* chellygel dances21:43
redrobotI hear chellygel is a huge fan of Erick's21:43
rm_workyeah I need to do that next week21:43
chellygelgerrit sucks -- its showing whitespace on her stuff when she has none there21:43
chellygelwtf wtf wtf21:43
rm_workit has been way too long, keep missing people going21:44
rm_workneed one of those things... that I can never remember the name of... the torta with the red sauce21:44
chellygellets go to erick's next week, we'll drag 2-pieces-of-cake- redrobot21:44
chellygel3* :P21:44
reaperhulkif he ate both of your pieces he ate 3 pieces21:44
chellygelum you need to /nick bigredrobot21:45
reaperhulkyeah man can't put on all that weight until AFTER the wedding21:45
chellygeloh, yeah /nick bigengagedredrobot21:45
rm_workbleh keystone doesn't really HAVE objects21:46
rm_worknot sure this is really a useful model21:46
rm_workah yeah that21:47
rm_workkeystone API does what current barbicanclient does21:47
rm_workerr, keystoneclient21:47
rm_workso we would need to look at something else as an example21:48
redrobotyep... I wonder what the openstacksdk folks are doing?21:48
* redrobot is Shreddin' for the Weddin'21:50
redrobotor should be anyway... that was the plan21:50
SheenaG11redrobot: you going to start working out with codekobe_ and jraim?21:55
redrobotSheenaG11 They did extend me an invite... the gym is way out of my way, and they meet at obscenely early in the day though...  I dunnno, still debating.21:56
SheenaG11It sounds like a party!  jraim is swimming in his tee shirts now22:03
openstackgerritDouglas Mendizábal proposed a change to openstack/barbican-specs: Refactor Client Entity Models
redrobotrm_work ^^ have at it22:09
*** akoneru has joined #openstack-barbican22:10
SheenaG11redrobot: you can come yoga with me22:13
SheenaG11I took chellygel today22:13
SheenaG11It was HARDCORE22:13
redrobotSheenaG11 at the REC?22:13
SheenaG11Ha, no - I don't drive down for that22:13
SheenaG11She's up in AUS today22:13
chellygeldude, chellygel was dying in yoga22:14
*** nkinder has quit IRC22:14
chellygelchellygel couldn't hold her own weight22:14
*** nkinder has joined #openstack-barbican22:14
redrobotchellygel but you weigh like nothing... o_O22:15
rm_workredrobot: lol that is way more bare-bones than I thought22:15
chellygelthats how week i am22:15
chellygelalso stupid22:15
* rm_work can do like ... one pull-up / push-up22:16
rm_workwtf, our client is throwing an exception that enters into site-packages/keystoneclient/openstack/common/apiclient/exceptions.py22:17
rm_workand it's totally broken22:18
rm_workwe don't handle our own exception code? >_>22:18
rm_workah it's because we wrap our session with a keystone session22:18
rm_workso their exception handling kicks in <_<22:19
rm_workand is broken for our app, obviously, since we don't format our JSON responses the same as they do22:19
rm_worknot sure how to deal with this22:19
rm_workour POST looks like: resp =, data=json.dumps(data), headers=headers)22:20
rm_work_session is a keystone session22:20
rm_workso when the post fails with like, a JSON schema error...22:20
rm_workfor instance, expiration date in the past22:20
rm_workthe exception gets caught and handled by keystone's exception handlers, which expect a very different response structure, so THAT breaks and the resulting error message is totally unhelpful and meaningles22:21
redrobotrm_work so we want to remove our auth layer entirely22:21
redrobotrm_work and while we're refactoring the whole thing, we shoudl fix this. :D22:21
rm_workredrobot: you know what scope creep is, right?22:21
rm_workbecause my scope was "use containers"22:22
rm_workand now it is "rewrite python-barbicanclient"22:22
redrobotit's what happens to the guy who's on loan :D22:22
redrobotI'd love to do it myself... but SheenaG11 won't give me the jira points :-P22:22
rm_workwell I started with a 5-point this week22:23
rm_workerr, this Sprint22:23
rm_workI've adjusted it to 20 T_T22:23
rm_workour sprint burn-down is more like a burn-up22:23
rm_workwell, it was 13 until just now22:23
rm_workNOW it's 2022:24
rm_workI'm going to...22:24
rm_workleave this alone for now22:24
rm_workand not pass invalid expiration dates :P22:24
SheenaG11That SheenaG11 is a taskmaster22:31
SheenaG11I tell you what22:31
rm_workredrobot: is there a better pattern than this:22:38
rm_work    @name.setter22:39
rm_work    def name(self, value):22:39
rm_work        self._check_mutable()22:39
rm_work        self._name = value22:39
rm_work    @property22:39
rm_work    def name(self):22:39
rm_work        return self._name22:39
rm_workthat's what my first pass looks like...22:39
rm_workfor like, every attribute22:39
redrobotlooks legit to me...22:39
redrobotrm_work look at that BP22:39
rm_workwas wondering if maybe there was some cleaner magic to do that22:40
redrobotrm_work DO IT!22:40
rm_workredrobot: CLEANROOM22:40
rm_workbut yeah, i looked at it22:40
rm_worklooks bare-bones, and exactly like what I was doing22:40
rm_workyou didn't really dip into specifics much :P22:40
redrobotrm_work can't remember if that'll work on 2.x though22:43
rm_workI will try it out22:43
rm_workI mean, properties are ostensibly FOR this22:43
rm_worklike, this exact scenario22:44
rm_workit's just very verbose22:44
rm_worki'll pastebin you the secrets class in a sec and see what you think22:45
redrobotbetter make it quick, because my GTFO ETO is 15 min22:45
rm_worklol k22:45
rm_workvery close22:45
chellygeleto? lol22:46
rm_workETA :P22:47
redrobotwhatever.... the GTFO part is what matters :-P22:48
chellygelgo vote on reviews for 15 min or something22:48
chellygel-1's all the way down22:48
rm_workredrobot: how long do I have?22:57
rm_workwas trying to finish my get() impl22:57
redrobot2 min22:58
*** juantwo has joined #openstack-barbican23:01
redrobottoo late holmes23:02
rm_workI can give you this as is23:02
redrobotJust upload as WIP and I'll review it when I get a chance23:02
rm_workredrobot: ^^^^^23:03
*** ayoung has quit IRC23:06
*** juantwo has quit IRC23:27
*** juantwo has joined #openstack-barbican23:28
openstackgerritConstanze Kratel proposed a change to openstack/barbican: :Incorporated review comments from John Wood into WADL and devguide
rm_workwoo, -1'd redrobot's BP :P23:49
* rm_work doesn't actually -1 that often23:49
*** akoneru has quit IRC23:50
openstackgerritConstanze Kratel proposed a change to openstack/barbican: Incorporated review comments from John Wood into WADL and devguide

Generated by 2.14.0 by Marius Gedminas - find it at!