| dfNull | anyone up to speed on the sso API for U1. I have it mostly working buy have some questions about the tokens and how revocation might work for the user | 03:08 |
|---|---|---|
| === karni is now known as karni-holiday | ||
| karni-holiday | dfNull: You still here? | 03:37 |
| dfNull | yeah, Hi | 03:38 |
| karni-holiday | dfNull: If you are, I could probably help | 03:38 |
| karni-holiday | dfNull: What language are you using? | 03:38 |
| dfNull | we can do this later...youre supposed to be away | 03:38 |
| karni-holiday | dfNull: I'm about to get some sleep ;) Come on :) | 03:38 |
| dfNull | java....you sent me your REST api stuff earlier in the week | 03:38 |
| karni-holiday | ah cool | 03:39 |
| karni-holiday | dfNull: If you use a revoked token, The U1FileAPI will throw Unauthorized Failure (failure with getStatusCode() == HttpStatus.SC_UNAUTHORIZED) | 03:39 |
| dfNull | I've got auth tokens working but am puzzled why I have to use "Ubuntu One" and the host name for the app token? | 03:40 |
| dfNull | when I use that it works | 03:40 |
| dfNull | but using anything else fails | 03:40 |
| karni-holiday | dfNull: It's because when you do sso.pingUbuntuOne(), Ubuntu One pulls tokens from sso, and filters them by "Ubuntu One @ " string pattern | 03:40 |
| karni-holiday | dfNull: If they match, the token is saved to U1 token cache. | 03:40 |
| dfNull | but then when I look in U1 web UI under my account I see my android device listed again, since I'm also running the U1 client from youguys | 03:40 |
| karni-holiday | Or at least, that's how you should think about it. | 03:41 |
| karni-holiday | dfNull: If you're asking about the token, you need to use it in the format of "Ubuntu One @ <your stuff here>" | 03:41 |
| karni-holiday | Normaly it should be the host name/device name. | 03:42 |
| karni-holiday | But you could put your application name in there. | 03:42 |
| karni-holiday | dfNull: Did that answer your question? | 03:42 |
| dfNull | hmmm...well I tried that with something other than androids Build.Device (hostname equiv) and your REST api wrapper fails | 03:42 |
| dfNull | fails auth at least | 03:43 |
| dfNull | yeah, that at least lets me know I should be able to vary the host name part | 03:43 |
| karni-holiday | dfNull: HAve you looked at lp:~karni/ubuntuone-files-java-library/cli-ubuntuone ? | 03:43 |
| dfNull | and I can work with that | 03:43 |
| karni-holiday | https://code.launchpad.net/~karni/ubuntuone-files-java-library/cli-ubuntuone | 03:43 |
| karni-holiday | dfNull: There's a line saying: final AuthenticateResponse response = sso.authenticate("Ubuntu One @ CliUbuntuOne"); | 03:44 |
| karni-holiday | It could also be "Ubuntu One @ Cli U1 whatever_you-like" what not :) | 03:44 |
| dfNull | Ok, I will look at that. was looking at | 03:45 |
| dfNull | https://code.launchpad.net/~ubuntuone-android-hackers/ubuntuone-files-java-library/trunk2.x | 03:45 |
| dfNull | most of which is broken, but got me going | 03:45 |
| karni-holiday | dfNull: Broken? | 03:45 |
| karni-holiday | dfNull: https://code.launchpad.net/~ubuntuone-android-hackers/ubuntuone-files-java-library/trunk0.x | 03:46 |
| dfNull | much of the command line stuff is commented out | 03:46 |
| karni-holiday | dfNull: cli-ubuntuone (the link I provided a few lines above) is a fully functional command line client | 03:46 |
| dfNull | the uri for listing volumes has an extra / at the end, etc. | 03:46 |
| dfNull | ok, I will look at that one | 03:47 |
| karni-holiday | dfNull: It has been extracted to a separate project from trunk2.x (which is now trunk0.x) | 03:47 |
| dfNull | I'm in a good place now know that I should be able to vary the hostname part | 03:47 |
| dfNull | which then lets someone revoke access via the web UI | 03:47 |
| dfNull | yeah, I noticed that the original one you sent was 404 | 03:47 |
| dfNull | thanks....hope youre away for work and not away on holiday doing IRC | 03:48 |
| karni-holiday | dfNull: "the uro for listing volumes nas an extra / at the end" - I don't know what you're talking about :< | 03:49 |
| karni-holiday | dfNull: API uri is https://one.ubuntu.com/api/file_storage/v1/volumes | 03:49 |
| karni-holiday | dfNull: And this is what getVolumes is calling: http://bazaar.launchpad.net/~ubuntuone-android-hackers/ubuntuone-files-java-library/trunk0.x/view/head:/main/src/com/ubuntuone/api/files/U1FileAPI.java | 03:49 |
| karni-holiday | line 409 | 03:49 |
| karni-holiday | dfNull: Well, I started holiday yesterday, but I'm very happy to answer questions related to SSO and U1FileAPI :) | 03:50 |
| dfNull | the one at trunk2.0x had extra trailing slash..IIRC | 03:50 |
| karni-holiday | dfNull: Do not hesitate to leave me priv messages or e-mail me at michal.karnicki at canonical dot com, so I can follow up :) | 03:50 |
| dfNull | had to grab source and fork to make getVolumes to work | 03:51 |
| karni-holiday | dfNull: I've rearranged things yesterday. the trunk2.x is now trunk0.x, so it's same code. | 03:51 |
| karni-holiday | hm o_O | 03:51 |
| karni-holiday | this is the last part of the uri: private static final String VOLUMES = "volumes"; | 03:51 |
| karni-holiday | I see no trailing slash (and there was none, if I am not mistaken. I can be mistaken :) ). | 03:52 |
| karni-holiday | dfNull: Anyhow, this should be working :) You can bzr branch the project trunk at lp:ubuntuone-files-java-library - it's most up to date | 03:52 |
| karni-holiday | and points to trunk0.x (tech preview of the library) | 03:52 |
| karni-holiday | It'll become 1.0.0 as soon as we believe is stable and has no further critical feedback from developers. | 03:53 |
| dfNull | hmmm...all I can say is the source I had wouldnt give me volumes until I changed final URI uri = new URI(HTTPS, META_HOST, API_PATH + "/volumes", null); | 03:53 |
| karni-holiday | dfNull: Have fun and if you find any problems, please do file bugs on the project page :) | 03:53 |
| dfNull | yeah, mercurial and git guy but time to try bazaar I guess | 03:54 |
| karni-holiday | hmm | 03:54 |
| karni-holiday | ah :) | 03:54 |
| dfNull | thanks and go away to enjoy holiday! | 03:54 |
| karni-holiday | dfNull: Haha, my pleasure! Keep sending your feedback :) Cheers!! | 03:54 |
| karni-holiday | dfNull: holly shoot, wait a second. haven't you branched the old library version by any chance :/ | 03:56 |
| karni-holiday | There's no such code as you just wrote :O | 03:56 |
| karni-holiday | dfNull: The old code is used by Ubuntu One Files on Android (so it's production code), but it wasn't near enough good to become public (as in, official) | 03:57 |
| karni-holiday | dfNull: Please make sure you are using U1FileAPI class to talk to U1, if not - ouch ouch, please branch lp:ubuntuone-files-java-library and sorry for all the confusion. We were rearranging things for the library release, and I think that bit you! | 03:58 |
| dfNull | huh? I got stuff from: https://code.launchpad.net/~ubuntuone-android-hackers/ubuntuone-files-java-library/trunk2.x | 03:58 |
| karni-holiday | Hmm. In that case, you got the right code. Do you remember where did you change the URI uri = new URI(HTTPS, META_HOST, API_ATH + .... ? | 03:58 |
| karni-holiday | Which file was it? | 03:58 |
| dfNull | which is now gone. | 03:58 |
| karni-holiday | ResourceClient.request(..) line final URI uri = new URI(getHost().getSchemeName(), getHost().getHostName(), path, query, null); ? | 03:59 |
| dfNull | RestApi.java | 03:59 |
| karni-holiday | dfNull: omg. you haven't branched trunk2.x :< | 03:59 |
| dfNull | no, I copied local | 03:59 |
| karni-holiday | you branched lp:ubuntuone-java-files-client few days ago, instead lp:ubuntuone-java-files-client/2.x | 03:59 |
| dfNull | really? | 04:00 |
| karni-holiday | dfNull: Ok, any how - I am terribly sorry. You are looking at the old, un-official code :< | 04:00 |
| dfNull | hmmm | 04:00 |
| karni-holiday | dfNull: The new one is SO MUCH BETTER :) | 04:00 |
| dfNull | well..I got it working mostly, nonetheless | 04:00 |
| karni-holiday | dfNull: I am so sorry. The old library is written quite ad-hoc, and is not that pleasant to work with (static calls with multitue of parameters) | 04:01 |
| karni-holiday | dfNull: I'll say this one last time - I am sorry, this was a misunderstanding. You've branched the old library - which is, well, production code, so it's not bad, but the new one is waaay better. | 04:01 |
| dfNull | yeah, well I've seen worse. Hadnt used signpost before and that was handy | 04:01 |
| karni-holiday | dfNull: :) | 04:01 |
| karni-holiday | dfNull: If you have time and nerve to change your code, you could upgrade to the tech preview of the new version | 04:02 |
| karni-holiday | dfNull: Notably, U1FileAPI class instead of RestApi class | 04:02 |
| karni-holiday | (Just mentioning a key point that differs the libraries. Everything has been written from scratch.) | 04:02 |
| dfNull | yah the jar I have is ubuntuone-files-java-client-rev3 | 04:02 |
| karni-holiday | dfNull: Right, that's the old one. Have a look at https://launchpad.net/ubuntuone-files-java-library | 04:03 |
| dfNull | from which I had to get source for RestApi.java and make my own version to fix the volumes uri, etc | 04:03 |
| karni-holiday | dfNull: The jar is now libUbuntuOneFiles-0.1.0-rc.1.jar | 04:03 |
| karni-holiday | dfNull: Well, you don't need to get source from the jar, there's source on launchpad, right :) | 04:03 |
| karni-holiday | Shoot, 3.5 hours of sleep and then hours on the bus. I should be going now.. | 04:04 |
| karni-holiday | dfNull: You've got nerve to have used RestApi, I'm actually pleased someone other than me (and Ubuntu One Files for Android) has used that code :) | 04:04 |
| dfNull | yeah, that what I meant. started with jar and went to get source from bzr for RestApi.java | 04:04 |
| karni-holiday | ah cool | 04:05 |
| karni-holiday | The jar didn't naturally contain the sources, only compiled classes. | 04:05 |
| dfNull | or javadoc in this case :/ | 04:05 |
| karni-holiday | no java doc? uhh that's a bug | 04:05 |
| dfNull | go away and thank you! | 04:06 |
| dfNull | HOLIDAY! | 04:06 |
| karni-holiday | dfNull: I hope you will consider upgrading to the new version, it's written way better and the source contains java doc. it also has the cli-ubuntuone example ^_^ | 04:06 |
| karni-holiday | haha, take care dfNull ! | 04:06 |
| karni-holiday | Catch you in 1.5 week | 04:06 |
| * karni-holiday disappears | 04:06 | |
| dfNull | I will use new and hope to have something to keep our keepass files safe when u return | 04:07 |
| karni-holiday | Rock on, dfNull ! | 04:07 |
| karni-holiday | :) | 04:07 |
| === m_conley is now known as m_conley_away | ||
| JamesTait | Happy Friday - wait, what, again? Already? ;) | 09:22 |
| dfNull | I think time dilates as you age! | 09:25 |
| dfNull | Yay! got a simple POC going on Android using the new ubuntuone-files-java-library-rc.1 | 09:28 |
| mandel | morning all | 09:44 |
| dfNull | morning! | 09:46 |
| dfNull | nice work on ubuntuone-files-java-library-rc.1! thanks! I've got enough of an android client working now that I can sleep tonight. have fun! | 09:50 |
| mandel | dfNull, laters! | 09:51 |
| * mandel rebooting | 09:53 | |
| === webmaster is now known as davidcalle | ||
| ralsina | good morning! | 10:39 |
| ralsina | mandel, can I get a simple review? https://code.launchpad.net/~ralsina/ubuntuone-control-panel/go-native | 10:39 |
| ralsina | mandel: and maybe a not-so-simple one: https://code.launchpad.net/~ralsina/ubuntuone-control-panel/tab-tab-tab/+merge/96687 | 10:40 |
| mandel | ralsina, sure, let me finish first with the ones from alecu | 10:42 |
| ralsina | mandel: cool, those are more urget | 10:51 |
| mandel | ralsina, you mean the proxy ones, right? | 10:51 |
| ralsina | mandel: yes | 10:51 |
| gatox | morning! | 10:52 |
| mandel | gatox, buenas, feeling any better? | 11:01 |
| gatox | mandel, yep.... taking some pills and a special diet :P | 11:01 |
| ralsina | good morning gatox | 11:01 |
| gatox | ralsina, hi | 11:01 |
| mandel | gatox, special diet? do you eat babies now? | 11:02 |
| gatox | ralsina, i'm reading nessita's email.... could you fix the button thing? | 11:02 |
| gatox | yep..... i'm seeing the screensshots right now | 11:03 |
| gatox | ralsina, who have the idea of changing the top and bottom padding on focus? | 11:17 |
| mandel | yay! all of alecus reviews done \o/ | 11:35 |
| mandel | ralsina, moving to yours | 11:35 |
| mandel | ralsina, for https://code.launchpad.net/~ralsina/ubuntuone-control-panel/go-native/+merge/96123 should we add tests for that? | 11:37 |
| nessita | buenas! | 12:19 |
| gatox | nessita, hola | 12:19 |
| gatox | nessita, i answer to your email | 12:19 |
| nessita | hi gatox! | 12:19 |
| nessita | gatox: how are you? | 12:19 |
| gatox | nessita, better | 12:19 |
| nessita | gatox: you *sure*? :-) | 12:20 |
| gatox | nessita, yes.... the doctor gave me some pills, special diet, and drink a lot of water because i was dehydrated because of the fever.... :P so, i was really sick jeje | 12:20 |
| nessita | gatox: and how many rest days did he give you? ;-) | 12:21 |
| gatox | nessita, i feel silly not thinking about reducing the highlighted area during focus.... grrrrrr | 12:21 |
| nessita | gatox: is ok, I did not come up with that, ralsina did | 12:21 |
| gatox | nessita, well..... they always gave two..... but that is for people who HAS to go to work.... i'm feeling better today... and i have all the weekend to rest :D | 12:22 |
| gatox | really! | 12:22 |
| mandel | nessita, morning! one question, landing in trunk, how does that work? I mean, if a branch is approved, when can we set it to be merged etc.. | 12:22 |
| nessita | mandel: hola! yes, releases were made on Wed, so go ahead | 12:23 |
| mandel | nessita, also, we need some reviews for alecu, all the ones related with https://bugs.launchpad.net/ubuntuone-client/+bug/929207 | 12:23 |
| ubot5` | Ubuntu bug 929207 in Ubuntu One Client "[FFE] Proxy "tunnel" for syncdaemon" [High,In progress] | 12:23 |
| mandel | nessita, super! gracias :) | 12:23 |
| nessita | mandel: yes, will continue reviews today | 12:23 |
| mandel | nessita, great, I have done all of them with some comments and a couple of very simple need fixing, if we can land this today it would be great since QA will have time to test it before beta2 | 12:24 |
| nessita | yes | 12:24 |
| nessita | mandel: do you know if qa has env set with proxy? | 12:25 |
| mandel | nessita, I'll be sending an email to them on how to at least set a very simple squid3 on P | 12:25 |
| mandel | nessita, with a sample configuration for auth and non-auth ports | 12:25 |
| nessita | mandel: ack, would you please cc me? | 12:26 |
| mandel | nessita, sure! | 12:26 |
| mandel | nessita, actually, I'll write it now so I don't forget | 12:26 |
| ralsina | good morning again! (for real this time) | 12:28 |
| ralsina | mandel: there are tests, it checks that it gets the constant as argument | 12:29 |
| ralsina | gatox: that's why I am the manager, I know all the dirty hacks | 12:29 |
| gatox | ralsina, jeje groso | 12:30 |
| ralsina | nessita: no, they don't have a proxy setup but a squid+auth is very easy to configure on linux | 12:30 |
| ralsina | in fact, I have a hack that would let us have a proxy where you auth with your canonical credentials in a public IP! Hey, I will do that now :-) | 12:31 |
| mandel | ralsina, ok, I'll do an irl and will add my +1 | 12:31 |
| ralsina | gatox: not for today, but this is going to totally own your weekend: from PyQt4 import QtDesigner | 12:34 |
| gatox | ralsina, do you mean for ninja? | 12:35 |
| ralsina | gatox: yep | 12:35 |
| ralsina | gatox: gives you access to all of Designer to integrate in apps | 12:35 |
| ralsina | gatox: apparently, I have been told, have not tried it ;-) | 12:35 |
| mandel | gatox, you should do the same with ninja | 12:36 |
| gatox | ralsina, we have been looking at that (i don't know if they updated that) but that was reallyyyyyy limited.... because the actual useful things of the qt designer part were private in qt, so riverbank didn't do the wrap | 12:36 |
| mandel | gatox, and then import ninja in ninja! | 12:36 |
| ralsina | gatox: ok, then, look for pynoto, they have a hack using private headers ;-) | 12:36 |
| gatox | ralsina, cool! thanks! :D | 12:36 |
| gatox | ralsina, i totally going to take a look at that! | 12:37 |
| mandel | ralsina, gatox, nessita do you sue evolution for mail? I get errors with the canonical smtp server :( | 12:46 |
| gatox | mandel, thunderbird | 12:47 |
| nessita | mandel: thunderbird | 12:47 |
| ralsina | mandel: thunderbird | 12:47 |
| ralsina | hahaha | 12:47 |
| ralsina | mandel: conclusion: evolution sux0rz | 12:48 |
| mandel | is it meant to appear now next to me, like beatlejuice? | 12:48 |
| mandel | ralsina, a little, just a little :P | 12:48 |
| nessita | ralsina, gatox: could you please review? https://code.launchpad.net/~nataliabidart/ubuntuone-control-panel/focus/+merge/96654 | 12:50 |
| gatox | nessita, yep, on it | 12:50 |
| ralsina | mandel, nessita, gatox, elopio, rmcbride: I have setup a public authenticated proxy. When you want to try using proxy support, if you don't want to setup your own squids, ping me ;-) | 12:54 |
| gatox | ralsina, ack | 12:54 |
| ralsina | nessita: sure | 12:54 |
| nessita | ralsina: nice! | 12:54 |
| ralsina | nessita: the explore button was already disabled. It just didn't *look* very disabled | 12:55 |
| nessita | ralsina: exactly, it was cause by the style being defined for "every state" instead for :enabled | 12:55 |
| ralsina | nessita: ack but then maybe the changelog should be changed a bit? | 12:56 |
| ralsina | nessita: or not, I don't really mind | 12:56 |
| nessita | will edit | 12:56 |
| ralsina | it's casual friday I won't have strong opinions until monday | 12:57 |
| nessita | added (style-wise) clarification | 12:57 |
| ralsina | nessita: good idea to start using declarations for colors | 12:57 |
| nessita | ralsina: can we do that? | 12:58 |
| nessita | ralsina: I just added a comment | 12:58 |
| ralsina | nessita: I think so, let me check | 12:58 |
| nessita | but not sure if "variables" can be defined | 12:58 |
| ralsina | we could pre-process with cpp at least | 12:58 |
| nessita | ralsina: you sure is worth it? | 12:59 |
| ralsina | nessita: well, right now we have ... 78 color declarations | 12:59 |
| ralsina | nessita: if I had a code file with 78 color literals, what would you say? ;-) | 13:00 |
| nessita | ralsina: 78 colors but the uniq of that is much less, no? | 13:00 |
| ralsina | ok, no, it's only 54 colors | 13:00 |
| ralsina | nessita: probably the uniq is much fewer, which is more reason to unify in constants somehow | 13:01 |
| gatox | ralsina, nessita maybe i'm wrong.... but the qss can use things from the resources...... so "maybe" (big maybe).... we can define that colors constants in the qrc | 13:01 |
| ralsina | gatox: that's an idea. It doesn't have syntax for variables or constants, AFAIK, let me dig resources | 13:01 |
| nessita | ralsina: I'm not sure how much is worth the effort, but will not block any work for that... as long as the solution is not more complex that adding a color in a file and using it :-) | 13:02 |
| nessita | ralsina: can we talk a little about the go-native branch? | 13:02 |
| ralsina | nessita: of course, no blocking anything | 13:02 |
| ralsina | nessita: I am just filing it in my "think in the background" folder | 13:02 |
| ralsina | nessita: sure, let's | 13:02 |
| ralsina | nessita: mumble or here | 13:02 |
| nessita | here | 13:02 |
| nessita | ralsina: so, I was cool with having qt in utils, but then I realized that it will break the ubuntu packaging of the project, let me explain further | 13:03 |
| nessita | ralsina: deja-dup depends on the backend of the cp which has utils, so deja-dup pulls the backend of the control panel in the CD | 13:03 |
| nessita | ralsina: so our backend can not pull qt :-/ | 13:03 |
| ralsina | nessita: ack | 13:04 |
| ralsina | nessita: ok, I can just add the constants in folders.py | 13:04 |
| nessita | ralsina: sorry I did not realized this sooner | 13:04 |
| ralsina | nessita: or create a platform-speciic module in gui/qt | 13:04 |
| ralsina | nessita: no problem, it smelled bad to me too | 13:04 |
| nessita | ralsina: could we please have an utils inside gui/qt/ that can be multiplatform | 13:04 |
| ralsina | nessita: one small argument for putting the constant in folders: this is temporary because it's a workaround for a Qt bug | 13:05 |
| nessita | ralsina: do we know if the bugs was already fixed? | 13:05 |
| ralsina | nessita: AFAIK, it's meant to be fixed in Qt 4.9 | 13:05 |
| nessita | ok | 13:05 |
| nessita | ralsina: not sure what's your suggestion about the fix then :-) | 13:06 |
| ralsina | nessita: I would rather put it in folders.py with a big note about removing it when qt 4.9 arrives | 13:06 |
| nessita | ralsina: works for me, as long as we don't keep adding the same in other modules (ie as long as we don't add more if platform == win in other modules) | 13:07 |
| ralsina | nessita: no, it's just a one-time thing | 13:07 |
| ralsina | nessita: I can put a "don't do this" in the comment, too | 13:07 |
| nessita | ralsina: please | 13:08 |
| ralsina | nessita: cool,fixing then | 13:08 |
| nessita | dobey: when you come in, I wanted to mention I hit this yesterday on an updated precise bug #853060 | 13:14 |
| ubot5` | Launchpad bug 853060 in ubuntuone-installer (Ubuntu) "ubuntuone-installer crashed with GError in function(): Failed to execute child process "ubuntuone-control-panel-gtk" (No such file or directory)" [High,Incomplete] https://launchpad.net/bugs/853060 | 13:14 |
| nessita | dobey: of course the failure was regarding the -qt control panel | 13:15 |
| nessita | dobey: installer version was 2.99.90 | 13:15 |
| gatox | nessita, +1 to your branch | 13:22 |
| nessita | gatox: thanks! | 13:22 |
| ralsina | nessita: updated go-native as we discussed | 13:26 |
| nessita | ralsina: ack | 13:26 |
| nessita | approved alerted | 13:26 |
| ralsina | nessita: thanks! | 13:27 |
| alecu | hello! | 13:27 |
| gatox | alecu, hi | 13:27 |
| alecu | hi gatox! feeling better? | 13:28 |
| gatox | alecu, yep, thanks :D | 13:28 |
| nessita | hola alecu | 13:28 |
| mandel | alecu, morning! | 13:33 |
| alecu | hola mandel! thanks for all the reviews | 13:33 |
| mandel | alecu, I did all the reviews, there are some comments that you might or might no follow and a small lint error in one of them | 13:34 |
| mandel | alecu, all should be very easy to fix :) | 13:34 |
| alecu | mandel, I'm fixing those so you can re-review in a few minutes. | 13:34 |
| alecu | great! | 13:34 |
| mandel | alecu, superb! I'll microwave some pasta and will do them asap | 13:34 |
| * mandel lunch | 13:34 | |
| nessita | ralsina: you wanna remove the code in ubuntuone/controlpanel/utils/__init__.py now, in go-native? | 13:41 |
| ralsina | nessita: ugh, sorry | 13:42 |
| ralsina | nessita: I am getting a bunch of fails on tests now, will ping you when that branch is good again | 13:44 |
| nessita | ralsina: thanks | 13:49 |
| ralsina | sigh, I have been doing much harder stuff, why is this stupid "change one constant" branch driving me nuts | 13:55 |
| dobey | nessita: what was the actual full traceback? | 13:57 |
| nessita | dobey: the exact same traceback as the bug report changing ubuntuone-control-panel-gtk by buntuone-control-panel-qt | 13:58 |
| nessita | dobey: and also, after the crash, the qt control panel was not installed, so seems like the spawning was made too early | 13:58 |
| nessita | dobey: I had to install it by hand, looked like no package was installed at all | 13:59 |
| ralsina | and as usual, the answer is, because sometimes I am dumb | 14:00 |
| dobey | nessita: well i guess it's not easy to debug now since you installed the package by hand | 14:01 |
| nessita | dobey: well, I needed it :-/ | 14:01 |
| nessita | alecu: approved with a fix request https://code.launchpad.net/~alecu/ubuntuone-client/proxy-tunnel-client/+merge/95077 | 14:09 |
| alecu | nessita, great! thanks. | 14:10 |
| nessita | ralsina: any update in my focus branch review? | 14:10 |
| ralsina | nessita: yes, running tests now | 14:10 |
| nessita | ralsina: oh neat! | 14:10 |
| ralsina | nessita: works nice IRL, and code review | 14:10 |
| ralsina | nessita: +1 | 14:11 |
| nessita | ralsina: thanks! wanna merge that in your branch, so I then review it? (referring to tab-tab-tab) | 14:12 |
| ralsina | nessita: sure, will do that now. Here is the much smaller go-native | 14:12 |
| nessita | where? I can't see it! :-P | 14:12 |
| ralsina | nessita: https://code.launchpad.net/~ralsina/ubuntuone-control-panel/go-native/+merge/96123 | 14:12 |
| nessita | (thta's small) | 14:12 |
| ralsina | nessita: I feel stupid about that branch, I hope now it's simple enough that it can be correct ;-) | 14:13 |
| ralsina | nessita: I'll wait for focus to merge and then will merge trunk | 14:13 |
| nessita | sure | 14:13 |
| nessita | ralsina: looks good, running tests and about to test IRL | 14:14 |
| gatox | ralsina, LOL.... love the comments here: https://code.launchpad.net/~ralsina/ubuntuone-control-panel/go-native/+merge/96123 | 14:15 |
| nessita | me too :-P | 14:15 |
| ralsina | gatox: if someone did that and suffered the wrath of the nessita because I failed to warn him, I would feel bad for 10 minutes at least. | 14:15 |
| gatox | jejeej hey! you have a typo :P | 14:15 |
| gatox | temporary | 14:16 |
| ralsina | NOOOOOOOOOOO THAT BRANCH IS CURSED! | 14:16 |
| gatox | jejeje | 14:16 |
| nessita | ralsina: approved on this end | 14:16 |
| gatox | i only realize because it was so funny that i read it twice | 14:16 |
| gatox | jeje | 14:16 |
| ralsina | gatox: pushed "temporary" | 14:16 |
| gatox | :P | 14:16 |
| nessita | ralsina: focus is merged now | 14:17 |
| nessita | dobey: trying to reproduce in my clean french precise vm | 14:18 |
| nessita | gatox: speaking of which, in that ^ vm, the "forgot password" is unclickable | 14:19 |
| nessita | gatox: let me file a bug for you :-) | 14:19 |
| gatox | nessita, ok | 14:19 |
| nessita | dobey: ugh, tarmac is requesting daily build for all the ubuntus for u1cp | 14:20 |
| gatox | nessita, when you have a moment please re-review this one: https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/too-verbose/+merge/96364 and let me know if those loggings are correct | 14:21 |
| dobey | nessita: ok, will fix | 14:21 |
| nessita | gatox: ack! | 14:21 |
| ralsina | nessita: tab-tab-tab ready for you | 14:26 |
| nessita | ralsina: thanks! | 14:27 |
| * gatox needs to resolve conflicts in a .ui file.... :S | 14:28 | |
| ralsina | gatox: ouch | 14:29 |
| gatox | yap | 14:29 |
| nessita | gatox: oh... reset password? | 14:29 |
| gatox | nessita, aja | 14:29 |
| nessita | perdón | 14:29 |
| nessita | need help? | 14:29 |
| gatox | nessita, no... i was reviewing my branch to see if it was ok, and i found that.... | 14:29 |
| gatox | i'll let you know when it's ready | 14:30 |
| * mandel back | 14:34 | |
| mandel | I think I ate too much.. | 14:34 |
| mandel | alecu, please let me know when ever you need a second review | 14:37 |
| nessita | mandel: can you squeeze a fix if any ussoc branch of yours? | 14:38 |
| nessita | mandel: in ssl_dialog, there is a docstring with 4 " | 14:38 |
| nessita | 65 class SSLDialog(QDialog): | 14:38 |
| nessita | 66 """"Dialog used to show SSL exceptions.""" | 14:38 |
| nessita | mandel: and that's busting the syntax highlight :-) | 14:39 |
| mandel | nessita, the other branches have been merged, but I certainly can add it in the next one | 14:39 |
| nessita | mandel: thanks | 14:39 |
| mandel | nessita, sorry for that, vim does not go crazy if that happens.. | 14:40 |
| nessita | mandel: it should :-P | 14:40 |
| mandel | nessita, it would be helpful to avoid this things, that is true | 14:40 |
| dobey | i guess i need to figure out how to set up/use some vms :( | 14:44 |
| dobey | nessita: hrmm, why can't i remove the other devices i have connected, in the qt panel? | 14:45 |
| nessita | dobey: design decision | 14:45 |
| dobey | GRRRRRRRRRR | 14:45 |
| nessita | dobey: I don't like it either | 14:45 |
| dobey | that is so broken | 14:46 |
| nessita | dobey: but they say is kinda a "security" measure, if someone different from you access the controlpanel in a single devicew | 14:46 |
| dobey | nessita: right, so they click on the big orange button, and it loads up the web page, and they click remove on there instead | 14:46 |
| nessita | right (?) | 14:46 |
| nessita | dobey: adding the remove button in the remote devices is trivial, wanna talk with design? :-) | 14:47 |
| dobey | no. my doctor says i need to keep my blood pressure down | 14:47 |
| ralsina | I don't eve like having the "remove *this* device" button since it leads 50% of the time to rot_mismathc | 14:48 |
| nessita | dobey: ah, ok. Mine says I need to keep my gastric juice down :-D | 14:48 |
| nessita | gatox: too-verbose comment added, branch is looking great, we really need this done (I know is boring though) | 14:48 |
| dobey | nessita: also, when i move the mouse over the tabs, it *feels* like it should switch the cursor to the pointy-finger url cursor instead | 14:49 |
| gatox | nessita, ok, i'll review it as soon as i finish with the ui thing | 14:49 |
| nessita | dobey: because they get underlined? | 14:49 |
| dobey | nessita: and the tabs light up. if feels like it's a web page. | 14:49 |
| dobey | nessita: there is a bug about the underlining already, right? | 14:50 |
| nessita | dobey: yes, but as far as I know we are not changing that... sorry | 14:51 |
| nessita | dobey: I honestly don't have a strong opinion about that, I don't like underline at all there, but we need to have the same UI across OSes... so | 14:51 |
| nessita | dobey: if you feel strong about this, I welcome you to talk with design, I will build the branch | 14:51 |
| ralsina | gotta go, school run then lunch, so back in about 80 minutes, I suppose | 14:51 |
| dobey | well we can remove it on windows too | 14:51 |
| nessita | dobey: yes | 14:52 |
| dobey | nessita: do you know the bug #? | 14:52 |
| nessita | dobey: sure | 14:52 |
| nessita | bug #939839 | 14:52 |
| ubot5` | Launchpad bug 939839 in Ubuntu One Control Panel "Qt control panel shouldn't underline tab title on hover" [Undecided,Triaged] https://launchpad.net/bugs/939839 | 14:52 |
| dobey | yay bug bot | 14:52 |
| nessita | dobey: let me know if you get new info on this | 14:52 |
| dobey | i just commented on the bug | 14:56 |
| alecu | mandel, "use sys.exit in main specially in the case were there is no proxy" <- why? | 14:59 |
| alecu | mandel, the main function is already returning in that case | 14:59 |
| mandel | alecu, to state that the process did not have a normal termination or something like that | 14:59 |
| nessita | me | 14:59 |
| mandel | alecu, it seems logical to say, hey we finished and did nothing because there is no proxy | 14:59 |
| mandel | me | 14:59 |
| alecu | mandel, but it's perfectly ok to exit if the settings say "no proxy" | 15:00 |
| alecu | mandel, so that case is a normal termination | 15:00 |
| mandel | alecu, if is normal termination ok, I can change my vote :) | 15:00 |
| alecu | mandel, the bits that start that process (in the following branch) already handle all those cases | 15:00 |
| alecu | mandel, ok, great. | 15:00 |
| nessita | gatox, dobey, ralsina, alecu, urbanape, me? | 15:01 |
| gatox | nessita, ok, the ugly reset one is ready | 15:01 |
| gatox | me | 15:01 |
| dobey | nessita: what about you? | 15:01 |
| mandel | alecu, ok, I did not review that taking into account the following ones | 15:01 |
| nessita | dobey: I said me already :-) | 15:01 |
| nessita | dobey: (11:59:50 AM) nessita: me | 15:01 |
| dobey | nessita: and just filed bug #950873 about device removal | 15:01 |
| ubot5` | Launchpad bug 950873 in ubuntuone-control-panel (Ubuntu) "Cannot remove other devices from control panel" [Undecided,New] https://launchpad.net/bugs/950873 | 15:01 |
| alecu | mandel, oh, and regarding "return settings" vs "if settings: return True" <- I like it more to return a boolean instead of the dictionary. | 15:01 |
| dobey | meh | 15:01 |
| alecu | mandel, I think it's cleaner... | 15:01 |
| nessita | alecu: in that case do: | 15:01 |
| nessita | alecu: return len(settings) > 0 | 15:02 |
| alecu | nessita, sounds fine. | 15:02 |
| urbanape | me | 15:02 |
| nessita | alecu: did you say me? | 15:02 |
| alecu | me | 15:02 |
| * nessita got lost with so many messages | 15:02 | |
| nessita | ok, let's! | 15:02 |
| nessita | DONE: did *not* start with bug #933697, instead tackled more reviews and bug #942020 and bug #949035 | 15:02 |
| nessita | TODO: more reviews and start with bug #933697 this time | 15:02 |
| nessita | BLOCKED: nopes | 15:02 |
| nessita | NEXT: mandel | 15:02 |
| ubot5` | Launchpad bug 933697 in Ubuntu One Control Panel trunk "[FFE] Integrate missing pages to Qt Control Panel" [High,Triaged] https://launchpad.net/bugs/933697 | 15:02 |
| ubot5` | Launchpad bug 942020 in Ubuntu One Control Panel "The selected tab has a weird orange highlight" [Medium,Fix committed] https://launchpad.net/bugs/942020 | 15:02 |
| mandel | DONE: Finished Qt webclient code. Reviews, reviews, reviews. | 15:02 |
| mandel | TODO: IRL tests for contorl panel, libsoup ssl. | 15:02 |
| mandel | BLOCKED: no | 15:02 |
| mandel | gatox, please | 15:02 |
| ubot5` | Launchpad bug 949035 in ubuntuone-control-panel (Ubuntu) "Explore buttons for folders not synced to the machine should be inactive" [Undecided,New] https://launchpad.net/bugs/949035 | 15:02 |
| gatox | DONE: | 15:03 |
| gatox | Took a sick day, investigate the focus button problem, fixed the logging branch, resolve conflicts in the ugly-reset branch. | 15:03 |
| gatox | TODO: | 15:03 |
| gatox | Keep fixing bugs from my queue. | 15:03 |
| gatox | BLOCKED: | 15:03 |
| gatox | No | 15:03 |
| gatox | dobey, go | 15:03 |
| dobey | λ DONE: finished packaging new rb, rb-u1 upload, twisted fix upload proposed | 15:03 |
| dobey | λ TODO: installer | 15:03 |
| dobey | λ BLCK: none. | 15:03 |
| dobey | urbanape | 15:03 |
| urbanape | DONE: pushed an initial branch (still needs cleanup) | 15:03 |
| urbanape | TODO: more refinement, out for the afternoon, getting lasik exam | 15:03 |
| urbanape | BLOCK: None | 15:03 |
| urbanape | alecu: go | 15:03 |
| alecu | DONE: worked on webclient tunnel branch #929212, reviews for mandel, fixes on the reviews | 15:03 |
| alecu | TODO: more fixes, finish #929212 | 15:03 |
| alecu | BLOCKED: no, because mandel and nessita are helping with the reviews. | 15:03 |
| * nessita will keep helping | 15:04 | |
| nessita | any comments anyone?> | 15:04 |
| alecu | nop | 15:04 |
| nessita | NOTE: starting next week, I will be leaving every Tuesday at noon ART, university classes start next weeks | 15:04 |
| nessita | next week* | 15:04 |
| dobey | no comments, only complaints ;) | 15:04 |
| alecu | dobey, complain away! | 15:05 |
| dobey | oh i will. i will. | 15:05 |
| nessita | and next Tuesday in particular, I will be taking half od day off, so I will not be at all :-) | 15:05 |
| nessita | ok, I guess is eom then | 15:05 |
| * nessita needs to buy some eggs | 15:05 | |
| nessita | eom! | 15:05 |
| * nessita -> groceries hunt | 15:06 | |
| nessita | brb | 15:06 |
| dobey | por medialunas? | 15:06 |
| nessita | dobey: nopes, to make an omelet :-) | 15:06 |
| dobey | medialunas de eggs | 15:07 |
| gatox | jeje | 15:07 |
| * gatox lunch | 15:07 | |
| === gatox is now known as gatox_lunch | ||
| mandel | ralsina, we have or 1-1 now, right | 15:08 |
| mandel | ? | 15:08 |
| mandel | alecu, so, I have been getting stupid errors with some of the tests because libsoup is reading my proxy settings, we should do something about it, maybe something like the dbus test but a gsettings tests or something | 15:21 |
| dobey | mandel: ugh, it's reading the real settings? | 15:23 |
| mandel | dobey, yes, terrible.. | 15:23 |
| dobey | :( | 15:23 |
| dobey | we need to fix that | 15:23 |
| mandel | dobey, I was wtf for a while.. | 15:23 |
| * nessita is back | 15:28 | |
| dobey | mandel: hrmm. i wonder if it's getting those values from the environment, rather than gsettings :) | 15:30 |
| mandel | dobey, I'm in the process of debugging the bloody thing.. | 15:32 |
| dobey | mandel: export GSETTIGS_BACKEND=memory in the run-tests script | 15:34 |
| mandel | dobey, what does that do? | 15:34 |
| dobey | mandel: makes gsettings use an in-memory db, which will be initially empty. so basically, not break the tests | 15:35 |
| mandel | dobey, awesome! | 15:35 |
| dobey | mandel: if it doesn't work, or breaks in some other way, let me know :) | 15:35 |
| ralsina | mandel: 1:1 in about 45' is ok? | 15:41 |
| ralsina | mandel: I am about to have lunch | 15:41 |
| mandel | ralsina, sure | 15:41 |
| alecu | mandel, pushed fixes to *-process branch | 15:48 |
| mandel | alecu, looking right now! | 15:48 |
| === JanC_ is now known as JanC | ||
| mandel | alecu, can it be that the mock server cancels a request? I getting a canceled request with libsoup + ssl cert | 15:50 |
| alecu | mandel, what? do you have a branch I can take a look at? | 15:51 |
| mandel | alecu, give me a sec and I'll pus it | 15:51 |
| mandel | alecu, lp:~mandel/ubuntu-sso-client/libsoup-ssl-dialog | 15:53 |
| mandel | alecu, try running: u1trial -r gi -t test_ssl_fail_dialog_user_accepts ubuntu_sso/utils/webclient/tests/test_webclient.py | 15:53 |
| alecu | mandel, http://pastebin.ubuntu.com/876189/ | 16:00 |
| alecu | mandel, it looks like libsoup is saying that. | 16:00 |
| mandel | alecu, yes, but I wonder why, ssl-strict is set to false so it should be doing something besides a cancellation | 16:01 |
| === gatox_lunch is now known as gatox | ||
| alecu | mandel, where are you setting "ssl-strict" ? | 16:07 |
| alecu | nevermind, found it | 16:07 |
| mandel | alecu, there :) | 16:07 |
| mandel | alecu, I also added a full no secure debugging with all the info | 16:08 |
| mandel | alecu, it seems that it does not event create the connection | 16:08 |
| alecu | mandel, found it! | 16:11 |
| alecu | mandel, _assert_ssl_fail_user_accepts returns a deferred | 16:11 |
| alecu | mandel, and trial is not getting at that deferred. | 16:11 |
| nessita | ralsina: tabtabtab reviewed | 16:11 |
| alecu | mandel, so: two options: | 16:12 |
| alecu | mandel, have all 4 tests return the deferred returned by _assert_ssl....... | 16:12 |
| alecu | mandel, or make each test a @defer.inlineCallback, and yield on the value returned by _assert_ssl.... | 16:13 |
| mandel | alecu, let me try it out :) | 16:13 |
| alecu | mandel, does it make sense? ^^^^ | 16:13 |
| mandel | alecu, thx! | 16:13 |
| alecu | no prob... it was a hard one to see :-) | 16:14 |
| gatox | need to restart | 16:14 |
| gatox | no need to restart anymore.... suddendly everything get fix :P | 16:15 |
| mandel | alecu, it fixes it!!! sweet I'll get the ssl dialog spawn in no time :) | 16:16 |
| alecu | mandel, lovely! | 16:16 |
| mandel | alecu, FYI running the tests in the branch you updated | 16:16 |
| alecu | mandel, btw: I've found this: http://valadoc.org/Soup-2.4.gir/Soup.Message.get_https_status.html | 16:16 |
| alecu | mandel, while looking for the issue. | 16:16 |
| alecu | mandel, perhaps you already think of using that method? | 16:17 |
| mandel | alecu, no I did not, specially because I don't seem yo find it in the c docs.. | 16:18 |
| * mandel looks | 16:18 | |
| alecu | mandel, I found it's present in the python object (I was looking at the dir() of that object, when it caught my eye) | 16:19 |
| mandel | alecu, in the soup session? | 16:20 |
| alecu | in the soup message | 16:20 |
| alecu | mandel, here you go: http://developer.gnome.org/libsoup/stable/SoupMessage.html#soup-message-get-https-status | 16:20 |
| mandel | alecu, oh, in the message! yep that one I new about :) | 16:20 |
| alecu | mandel, ok, cool. | 16:21 |
| mandel | alecu, +1 on https://code.launchpad.net/~alecu/ubuntuone-client/proxy-tunnel-process/+merge/96212 | 16:22 |
| dobey | need lunch, bbiab | 16:22 |
| mandel | alecu, did you fix the pylint issue for https://code.launchpad.net/~alecu/ubuntuone-client/proxy-tunnel-useit/+merge/96213 ? | 16:22 |
| alecu | mandel, I'm looking at it right now. | 16:23 |
| alecu | mandel, I'll push in 5' | 16:23 |
| ralsina | mandel: 1:1 on mumble? | 16:23 |
| mandel | alecu, cool, let me know | 16:23 |
| alecu | make that 10 | 16:23 |
| mandel | ralsina, sure, let's do it! | 16:23 |
| mandel | ralsina, can you hear me? | 16:25 |
| ralsina | mandel: you can't hear me? | 16:25 |
| ralsina | mandel: I can hear you | 16:25 |
| mandel | ralsina, nop, let me restart, the same happened yesterday | 16:25 |
| ralsina | mandel: ok | 16:25 |
| === zyga is now known as zyga-afk | ||
| pmatulis | nessita: re bug #944110, i'm trying to restart the client but it doesn't seem to work remotely | 16:34 |
| ubot5` | Launchpad bug 944110 in Ubuntu One Client "Permissions changed on unmodified files" [Wishlist,Confirmed] https://launchpad.net/bugs/944110 | 16:34 |
| pmatulis | dbus.exceptions.DBusException: org.freedesktop.DBus.Error.NotSupported: Unable to autolaunch a dbus-daemon without a $DISPLAY for X11 | 16:34 |
| nessita | pmatulis: hi there | 16:35 |
| nessita | pmatulis: yes, U1 is not supported on headless install or headless access, you need a dbus session running and without an X you don't get one | 16:35 |
| pmatulis | nessita: maybe i'll reboot my home 'puter then | 16:36 |
| nessita | pmatulis: is this to gather DEBUG logs? | 16:36 |
| pmatulis | nessita: yes, as per your instructions | 16:36 |
| pmatulis | nessita: i noticed on Home 'puter the syncdaemon.conf file was different from my Work 'puter | 16:36 |
| nessita | pmatulis: right, sorry :-/ | 16:36 |
| nessita | (about the need of reboot) | 16:37 |
| nessita | pmatulis: note that syncdaemon will no be started unitl your login to your x session with your username | 16:37 |
| pmatulis | gah! | 16:37 |
| pmatulis | nessita: i have this on Home. sounds like a default: | 16:38 |
| pmatulis | [__main__] | 16:38 |
| pmatulis | files_sync_enabled = True | 16:38 |
| nessita | pmatulis: yes, that's correct | 16:38 |
| pmatulis | k | 16:38 |
| alecu | mandel, fixed lint issue, now pushing. | 16:39 |
| alecu | mandel, it's pushed now. | 16:40 |
| mandel | alecu, ok, taking a look | 16:41 |
| alecu | nessita, ralsina, gatox, dobey, thisfred, urbanape: plis, reviews? | 16:43 |
| alecu | https://code.launchpad.net/~alecu/ubuntuone-client/proxy-tunnel-process/+merge/96212 | 16:43 |
| alecu | and https://code.launchpad.net/~alecu/ubuntuone-client/proxy-tunnel-useit/+merge/96213 | 16:43 |
| thisfred | sure | 16:43 |
| nessita | alecu: I was waiting for you to fix the needs fixing from mandel | 16:43 |
| gatox | alecu, i'll review the first one right now | 16:43 |
| nessita | thisfred: I'll take use-it | 16:43 |
| alecu | nessita, great, thanks. | 16:43 |
| nessita | gatox: I think between thisfred and I got those 2 | 16:43 |
| gatox | nessita, ok | 16:44 |
| nessita | gatox: but thanks! | 16:44 |
| thisfred | nessita: k, I'll do the first | 16:44 |
| ralsina | alecu: got it | 16:44 |
| ralsina | stupid question, on what file is the list of UDFs? | 16:44 |
| pmatulis | nessita: k, i'm good. i forgot that i have auto-login on Home | 16:44 |
| alecu | ralsina: eric and natalia are already on it. | 16:44 |
| nessita | pmatulis: that's handy :-) | 16:44 |
| alecu | ralsina, probably in some metadata. Only verterok may know :-) | 16:45 |
| pmatulis | nessita: yeah | 16:45 |
| ralsina | alecu: even better :) | 16:45 |
| ralsina | alecu: ok, so I will just delete the metadata, because I think I just turned Ubuntu One into a UDF ;-) | 16:45 |
| alecu | ralsina, cleanest way is "u1sdtool --list-folders" but it's likely you already knew that. | 16:45 |
| pmatulis | nessita: i presume you only need logs from the 'puter that lost the permissions? | 16:45 |
| alecu | ralsina, "Ubuntu One" is usually a UDF, only with an empty volume uuid | 16:46 |
| pmatulis | (when it happens again) | 16:46 |
| nessita | pmatulis: well, ideally from both, since I'm betting there was a modification triggered by the other computer... but is just a guess that allows me to explain the perm lost :-) | 16:46 |
| pmatulis | nessita: ok, will send both if it happens again | 16:46 |
| nessita | pmatulis: unless we have a very dark and hidden corner case, syncdaemon never modifies a files unless the server notifies of a modification happening in another device | 16:47 |
| pmatulis | nessita: sorry to hear that 'cause i did not change 35 shell scripts | 16:48 |
| nessita | pmatulis: sometimes, some programs, will modify your files without you noticing. I know the following example does not apply to your use case, but for example many music players edit tags in mp3 files | 16:48 |
| nessita | and the user himself never "changed" the file. So, with your logs, I can detect if there was an actual modification, and when | 16:48 |
| pmatulis | nessita: k, thanks for your help | 16:49 |
| nessita | pmatulis: thank you! | 16:49 |
| nessita | gatox: https://bugs.launchpad.net/ubuntu-sso-client/+bug/950953 | 16:51 |
| ubot5` | Ubuntu bug 950953 in Ubuntu Single Sign On Client trunk "I can not click the "Password forgotten" link" [High,Triaged] | 16:51 |
| gatox | nessita, weird..... but ok :P | 16:51 |
| nessita | gatox: this happened in a clean vm install, no nightlies, french lang | 16:52 |
| nessita | gatox: 100% reproducable | 16:52 |
| thisfred | alecu: +1, two small comments, but nothing that needs changing | 16:54 |
| alecu | \o/ | 16:54 |
| alecu | thisfred, good point on str.format() | 16:58 |
| alecu | thisfred, I'll change that, since I don't see it used everywhere else in u1-client. | 16:59 |
| ralsina | nessita: replied on tab-tab-tab, the loop in test_focus_order is tricky to get right, I would rather have it fail if someone changes FAKE_VOLUMES_INFO (and I'd rather we don't change it anyway ;-) | 16:59 |
| thisfred | alecu: yeah, it's a shame, because I like the new syntax much better, but it seems to be a failed experiment at this point | 17:00 |
| alecu | thisfred, anyway, I'll do it in the last branch of this series, since it's used there too. | 17:00 |
| thisfred | also having it be the same as in C has its advantages | 17:00 |
| thisfred | alecu, sure | 17:00 |
| mandel | alecu, sorted! | 17:05 |
| mandel | ok, EOD for me, need to go and visit the brother, havea great weekend! | 17:05 |
| alecu | mandel, thanks for all the reviews. | 17:06 |
| alecu | mandel, hope your brother gets well soon! | 17:06 |
| alecu | mandel, and have a great weekend too. | 17:06 |
| alecu | ok, time for lunch and some errands. See you guys laters! | 17:09 |
| === alecu is now known as alecu-away | ||
| === zyga-afk is now known as zyga | ||
| ralsina | nessita: I am looking at lisette's focus doc, and specifically at the settings tab (devices looks ok to me!). Currently, we open focused on "Get more storage" but "Apply these settings" is a orange default button. So, users may be surprised, think they are applying settings and get sent to buy stuff instead | 17:23 |
| ralsina | nessita: how do you feel about making "get more storage" not orange? I know we have to ask design, but it really makes everything harder | 17:23 |
| ralsina | nessita: or at least not start focused there | 17:24 |
| verterok | ralsina: there is no file with the UDFs, it's all in tritcask | 17:31 |
| ralsina | verterok: thanks | 17:32 |
| nessita | ralsina: trying to follow up in that... | 17:36 |
| nessita | ralsina: so which one is the default button that shouldn't be from your POV? | 17:37 |
| ralsina | nessita: having "get more storage" be a default button is weird | 17:37 |
| ralsina | nessita: or at least we should focus on "apply changes" when we switch to the settings tab | 17:38 |
| nessita | ralsina: so, I'm not sure I follow. The app as a whole will have the focus on whatever widget the user last tabbed in, no? | 17:38 |
| nessita | ralsina: when the user swicthes to folder we don't change the focus, same for devices or account, I don't find consistent changing the focus for settings | 17:39 |
| ralsina | nessita: well, the default button is supposed to be bound to "Enter" | 17:39 |
| ralsina | nessita: and space is used to activate the focused button instead | 17:40 |
| nessita | yes, when you're "in" a set of widgets with a default button. So you say that clicking on a setting does make hitting enter trigger the apply of the new settings? | 17:40 |
| ralsina | nessita: I see I was extra-annoyed because I was using u1cp-qt --switch-to settings to test the tab ordering | 17:40 |
| gatox | brb.... reboot | 17:40 |
| ralsina | nessita: right | 17:40 |
| nessita | let me try that | 17:41 |
| gatox | everytime i said reboot: the problem get fix | 17:41 |
| ralsina | nessita: or clicking on "settings", then enter, should apply (even if it's useless) | 17:41 |
| dobey | nessita: ugh. also, the window title of the sso qt dialog is "ubuntu-sso-login-qt". that's not good | 17:41 |
| nessita | ralsina: so, before try it (I'm logging in to U1), it smells like a focus issue, not a "default button" thingy, no? cuz even if we make the buy more storage non -default, the same problem is there, no? | 17:42 |
| nessita | dobey: yesm already reported and assigned to gatox | 17:42 |
| ralsina | nessita: I think it's both. We have two "extra-special-orange" buttons, and enter does the wrong thing | 17:43 |
| nessita | ralsina: but the buttons are located in different "widget sets" | 17:43 |
| dobey | nessita: and no icon :-/ | 17:43 |
| ralsina | nessita: doesn't matter, it's a dialog | 17:43 |
| nessita | ralsina: so they are default for different scopes | 17:43 |
| nessita | ralsina: is not a dialog :-) is a mainwindow | 17:43 |
| ralsina | nessita: that's nerdtalk there :-) | 17:44 |
| ralsina | nessita: basically, that sounds like an artificial distinction, when you consider the UI as a whole | 17:44 |
| nessita | dobey: yes, that's a very old bug for ussoc, but we never got the proper name to use for the "ubuntu" icon | 17:44 |
| gatox | nessita, which one is that bug? i can't find it in my queue | 17:44 |
| nessita | gatox: https://bugs.launchpad.net/ubuntu-sso-client/+bug/949744 | 17:45 |
| ubot5` | Ubuntu bug 949744 in ubuntu-sso-client (Ubuntu) "The ‘Sign In’ window does not have a title" [High,Triaged] | 17:45 |
| dobey | nessita: i'm not sure we should use the "ubuntu" logo icon exactly. | 17:45 |
| nessita | dobey: for the sso service, we definitely will not use the u1 icon | 17:45 |
| nessita | dobey: if we would have a sso icon, we should use that | 17:45 |
| dobey | nessita: the u1 icon was not my suggestion :) | 17:45 |
| nessita | dobey: I know, I was being pro active in answers :-P | 17:45 |
| nessita | ralsina: not sure what you mean with artificial distinction... | 17:46 |
| dobey | hmm | 17:47 |
| ralsina | nessita: that we know it's a separate widget hierarchy, but to theuser it's a window with two orange buttons in it (or 3 in accounts) | 17:47 |
| dobey | and login.ubuntu.com of course just uses the same thing as the rest of ubuntu | 17:47 |
| ralsina | nessita: nevermind, I will try to write a coherent email about this and send to you and design :-) | 17:47 |
| nessita | ralsina: ok, sounds good | 17:47 |
| nessita | the UI thingy is getting annoying, too many people expressing their opinions (which is fine) but not much we can do without a clear path from the design team | 17:48 |
| ralsina | g able to move from twitter to facebook using arrows) | 17:48 |
| nessita | ralsina: I would ignore that and let the user move with tab | 17:48 |
| ralsina | nessita: agreed, I was being part of the problem there | 17:48 |
| nessita | is easier and more intuitive? | 17:48 |
| ralsina | nessita: pressing enter to activate the highlighted button is a reflex to many users | 17:49 |
| nessita | ralsina: any idea why only space works? | 17:49 |
| nessita | (for some buttons) | 17:49 |
| ralsina | nessita: but as I said, I will try to write a reasonable email | 17:49 |
| nessita | ok | 17:49 |
| ralsina | nessita: like which ones? | 17:49 |
| nessita | ralsina: twitter and facebook, get help | 17:49 |
| dobey | nessita: they aren't buttons? | 17:50 |
| ralsina | nessita: you are supposed to use space to activate buttons. Having enter work is the strange one :-) But let me do some checks | 17:50 |
| nessita | dobey: they are | 17:50 |
| ralsina | dobey: yes they are | 17:50 |
| nessita | ralsina: if it's expected, that's fine by me | 17:50 |
| ralsina | nessita: I am double-checking | 17:51 |
| nessita | I expect a focused button to be activated by enter, but I'm a power user | 17:51 |
| ralsina | nessita: toy app with buttons in it, they activate by space, not by enter | 17:51 |
| nessita | ack | 17:52 |
| dobey | ugh | 17:52 |
| dobey | this is nasty | 17:52 |
| ralsina | dobey: I assume ugh means "on gtk they activate with enter and not space"? | 17:52 |
| dobey | no | 17:52 |
| dobey | ugh means "ugh, they *are* buttons" | 17:53 |
| ralsina | dobey: we could make them labels, if you want. You would not be able to see the difference, and you would not be able to activate them by kbd though. | 17:53 |
| dobey | ugh and now i have seen the border around the fb/twit icons | 17:53 |
| dobey | ralsina: you can see the difference. though i don't know where any labels are in there which you can tab to | 17:54 |
| ralsina | dobey: the border when focused was added so you know you tabbed there | 17:54 |
| ralsina | dobey: it was invisible before | 17:55 |
| dobey | ralsina: the one that covers up the text? | 17:55 |
| dobey | ralsina: and i see a border on the fb/twit icons when they aren't focused | 17:56 |
| nessita | dobey: can you share a screenshot please? | 17:56 |
| nessita | alecu ping | 17:57 |
| nessita | no alecu? bo | 17:57 |
| gatox | nessita, ralsina here is the branch for the title thing: https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/window-title/+merge/96802 is reallyyyy really trivial | 17:58 |
| ralsina | nessita: I see a visible border when unfocused but it doesn't cover any text | 17:59 |
| ralsina | oops, that was for dobey ^ | 17:59 |
| nessita | there's a border? but they both have background: transparent | 18:00 |
| dobey | ralsina: no, the get help and the giant white button above the tabs, have the bottom part of the text covered up though, when they are focused | 18:01 |
| ralsina | dobey: not here for the get help, and what "giant white button"? | 18:02 |
| dobey | ralsina: "go to the web for blah blah blah blah" | 18:03 |
| nessita | dobey: can you show what you refer to with a screenshot please? | 18:03 |
| dobey | nessita: http://ubuntuone.com/67YpqFQR8wPvK7cS20QUqT shows the borders around the icons | 18:04 |
| ralsina | dobey: nope, no overlapping there, either. Are you using current trunk? | 18:04 |
| dobey | nessita: you might need xmag | 18:04 |
| nessita | dobey: I have it | 18:04 |
| dobey | ralsina: i'm using nightlies | 18:04 |
| nessita | dobey: debugging now | 18:04 |
| nessita | dobey: ok, I see them as well, would you please file a bug for me? | 18:05 |
| dobey | ralsina: oh, get help is a bit weirder. get more storage covers up text on the boottom though | 18:07 |
| dobey | as does go to web blah blah | 18:07 |
| ralsina | dobey: I would love some screenshots of that, please | 18:07 |
| dobey | ralsina: yes, am working on it it | 18:08 |
| dobey | it even | 18:08 |
| ralsina | dobey: because get more storage now should have a black halo outside when in focus | 18:08 |
| dobey | i can't focus them all at the same time and just take one screenshot, you know :) | 18:08 |
| nessita | ralsina, gatox: about https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/window-title/+merge/96802, I'm now considering that perhaps we should not put the app_name as title, but the ussoc name (Ubuntu Single Sign On), since the window is for that service, and there is a trade off between customizing sso and maintaining the branch of sso | 18:08 |
| ralsina | dobey: and "go to web for..." should have a gray halo | 18:08 |
| ralsina | nessita: +1 on generic Ubuntu Single Sign On | 18:09 |
| gatox | nessita, as you prefer | 18:09 |
| nessita | branch -> brand | 18:09 |
| nessita | gatox: the thing is that is not my call :-/ | 18:09 |
| gatox | ahhhhh | 18:09 |
| gatox | nessita, who? design? | 18:09 |
| dobey | ralsina: and they do | 18:09 |
| nessita | gatox: not really, ISD | 18:09 |
| ralsina | dobey: ok, I will just wait for the shots then :-) | 18:10 |
| nessita | gatox: is like we re using, as a dumb example, paypal, and we were not saying that we're using paypal | 18:10 |
| gatox | nessita, i understand | 18:10 |
| nessita | gatox: so we're using a service that should have a brand of its own, and we can't just remove all references to sso | 18:10 |
| nessita | gatox: let's change to "Ubuntu Single Sign On" for the title, non translatable string | 18:11 |
| nessita | ralsina: agreed? ^ | 18:11 |
| ralsina | nessita: yes | 18:11 |
| gatox | ok, fixing | 18:11 |
| nessita | dobey: YEY u1sdtool finishes now! | 18:13 |
| nessita | dobey: may I see the patch you proposed yesterday? (link to the mp please) | 18:13 |
| dobey | nessita: for twisted? | 18:14 |
| nessita | yes | 18:15 |
| dobey | https://code.launchpad.net/~dobey/ubuntu/precise/twisted/fix-935756/+merge/96617 | 18:15 |
| gatox | nessita, done | 18:15 |
| nessita | gatox: thanks | 18:17 |
| * dobey ponders just writing a gtk+ control panel in vala | 18:18 | |
| dobey | oh | 18:19 |
| dobey | speaking of gtk+ | 18:19 |
| nessita | ok, I need to *start coding* today, so will finish title review and then go underground | 18:19 |
| dobey | blah, my scrollback is cut off | 18:19 |
| nessita | spent the whole day so far answering style complaints ;-) and reviews | 18:19 |
| nessita | gatox: will finish too-verbose on monday, ok? | 18:20 |
| nessita | no rush with that one, you may seek another review in the mean time | 18:20 |
| dobey | JanC: btw, if you want a separate application for ubuntuone music store, it would take all of about 15 minutes to write a python script that does exactly that. | 18:21 |
| dobey | http://ubuntuone.com/5IvaCQhMzY2jQpYliO57TM | 18:21 |
| dobey | http://ubuntuone.com/4HwY018NyJ5IYPbVgVSqzI | 18:21 |
| dobey | http://ubuntuone.com/6tXJfPd5XXlWO2I2lJw0ob | 18:21 |
| dobey | ralsina, nessita: ^^ ugliness in buttons | 18:22 |
| ralsina | nessita: haha, go and code with my blessing! ;-) | 18:22 |
| ralsina | dobey: the blue square is something else, and we are aware of that | 18:22 |
| ralsina | dobey: although for me it's pink-orangish | 18:22 |
| dobey | ralsina: it may be blue for me because of gtk+ theme | 18:23 |
| ralsina | dobey: the "get more storage" button just doesn't do that here | 18:23 |
| dobey | am not using ambiance at the moment | 18:23 |
| ralsina | dobey: right | 18:23 |
| dobey | using the upstream gnome 3 one, so i can read the text on my gnome-panel | 18:23 |
| ralsina | dobey: and I see no overlapping in "Go to the web" | 18:24 |
| dobey | ralsina: i don't know. i'll file a bug i guess | 18:24 |
| ralsina | dobey: not about the 1st or 3rd ones | 18:25 |
| ralsina | dobey: the 1st we know, the 3rd I don't see the wrong | 18:25 |
| dobey | ? | 18:25 |
| dobey | ralsina: you don't see the bottoms of the "p" and "g" characters cut off? | 18:26 |
| dobey | ralsina: it's the same bug as the get more button. it's just a different color | 18:26 |
| ralsina | dobey: ha, chrome was showing me like 30% of the image | 18:26 |
| ralsina | dobey: I only saw "Go to the web" | 18:26 |
| nessita | ralsina: the "get more storage" button must be our padding hack | 18:26 |
| nessita | ralsina: perhaps we need to make the padding bigger | 18:27 |
| dobey | nessita: i think all these problsm are "our <insert something here> hack" | 18:27 |
| ralsina | nessita: 1000px should be enough ifit was going to work | 18:27 |
| nessita | dobey: maybe | 18:27 |
| nessita | ralsina: yeah, but that triggers another issue... | 18:27 |
| nessita | ralsina: if the "there was a problem window" is raise, the "show deatils" button has that padding applied | 18:28 |
| ralsina | nessita: ok, so let's use 100% padding then | 18:28 |
| nessita | ralsina: you can try it by putting an "assert False" inside the folders load() function | 18:28 |
| ralsina | or 50% | 18:28 |
| nessita | ralsina: wanna try it in a branch? will review *not today* but on monday for sure :-P | 18:28 |
| ralsina | nessita: sure | 18:28 |
| nessita | gatox: approved with a comment https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/window-title/+merge/96802 | 18:30 |
| ralsina | dobey: do all the problems go away if you use -style windows? other than it looking like a windows app, mind you ;-) | 18:30 |
| dobey | ralsina: it *already* looks like a windows app :( | 18:31 |
| dobey | maybe you missed the e-mail where i said that at greater length :) | 18:31 |
| ralsina | dobey: yadda yadda | 18:31 |
| gatox | nessita, ok | 18:32 |
| dobey | ralsina: some of the buttons are weird in a slightly different way, but the text isn't really getting cut off. the underline on "Get help online" gets covered up it looks like, and the focus dotted line intereferes with the p character, but i don't think it's actually "hidden" beyond the obvious confusing white pixels coalescing in the same area | 18:33 |
| ralsina | nessita: the border around facebook and twitter is about those buttons not being set to flat | 18:33 |
| nessita | ralsina: oh my bad then | 18:34 |
| ralsina | nessita: np, trivial fix | 18:34 |
| gatox | nessita, should i look for another aprroval? i already push the changes with the constant | 18:36 |
| nessita | gatox: yes, please ask ralsina (mainly for the nature of the change, not the diff itself) | 18:37 |
| gatox | nessita, ok.... ralsina please when you have a minute, really trivial branch: https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/window-title/+merge/96802 | 18:37 |
| ralsina | gatox: got it | 18:37 |
| urbanape | off to get my lasik consultation. see you guys on Monday | 18:38 |
| gatox | ralsina, thanks | 18:38 |
| ralsina | gatox: +1 | 18:38 |
| dobey | cheers urbanape; don't go blind | 18:38 |
| gatox | ralsina, great | 18:39 |
| urbanape | here's hoping | 18:39 |
| urbanape | yay, science | 18:39 |
| dobey | fml. this weather is hell on sinuses | 18:42 |
| ralsina | dobey: you have 2' to see if lp:~ralsina/ubuntuone-control-panel/tweaks fixes help / twitter / facebook buttons for you? | 18:49 |
| dobey | sure | 18:50 |
| ralsina | dobey: I added some padding around the get help because the halo felt too tight, but that's not final | 18:50 |
| dobey | since now is the time when millbank falls off the earth | 18:50 |
| ralsina | dobey: yeah, we are all alone us americans | 18:50 |
| dobey | ralsina: that branch fixes those issues, yes. the arrow circle thing moves when i focus/unfocus the buttons with them though | 18:52 |
| ralsina | dobey: yes, that's a bug in the gtk theme for qt | 18:52 |
| ralsina | dobey: it draws the icon as if it were not a flat button | 18:53 |
| ralsina | dobey: I can do a fix, though | 18:53 |
| dobey | eh | 18:53 |
| dobey | i'm not as bothered about that | 18:53 |
| ralsina | dobey: me neither | 18:53 |
| === jalcine_ is now known as jalcine | ||
| dobey | and i still think the buttons are ugly, but at least the text doesn't get covered up now | 18:53 |
| ralsina | dobey: and there is no blue highlight anymore, right? | 18:53 |
| dobey | ralsina: right | 18:53 |
| ralsina | dobey: how about on the "go to web" one? | 18:54 |
| dobey | it didn't have a blue hilight before | 18:54 |
| dobey | and still doesn't | 18:54 |
| dobey | but the text is visible now | 18:54 |
| ralsina | dobey: ok, good | 18:55 |
| dobey | come back alecu | 19:01 |
| gatox | nessita, about the forgotten passoword link not clickeable: https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/forgotten-link/+merge/96814 | 19:08 |
| nessita | ack | 19:08 |
| nessita | gatox: nice catch | 19:08 |
| gatox | :D | 19:09 |
| nessita | ;-) | 19:09 |
| ralsina | gatox: care to take a look at bzr+ssh://bazaar.launchpad.net/~ralsina/ubuntuone-control-panel/tweaks/ and tell me if you like the changes? | 19:10 |
| gatox | ralsina, ok, looking | 19:11 |
| ralsina | gatox: very minor tweaks to make the gotoweb buttons a bit less weird (text doesn't wiggle on focus) | 19:11 |
| gatox | ralsina, maybe you can change also the padding of the QTabBar::tab:focus { | 19:18 |
| gatox | from: | 19:18 |
| gatox | padding-left: 1000px; | 19:18 |
| gatox | padding-right: 1000px; | 19:18 |
| gatox | to use with percentaje | 19:18 |
| gatox | percentage | 19:18 |
| gatox | ralsina, everything looks fine! | 19:19 |
| ralsina | gatox: sure | 19:19 |
| ralsina | gatox: I am trying to hack gotowebbutton to set the icon via background-image to remove it from the code | 19:20 |
| ralsina | gatox: also so it stops fidgeting when I tab ;-) | 19:20 |
| gatox | ralsina, and fixing the icon-size from the qss instead? | 19:23 |
| gatox | fixing i mean..... set the icon-size of the button | 19:24 |
| gatox | or it would be the same situationn as now? | 19:24 |
| ralsina | gatox: the problem is not size, it's position | 19:24 |
| gatox | ahhhhh | 19:24 |
| ralsina | gatox: when the button gets focus, the clipping area gets pushed 2 pixels down, and there goes the icon | 19:24 |
| gatox | ahhhh i see | 19:25 |
| ralsina | gatox: I blame elves | 19:25 |
| gatox | jejej | 19:25 |
| gatox | does anyone know where i can report a bug about the hud? | 19:29 |
| dobey | https://code.launchpad.net/~dobey/ubuntuone-client/transient-notify/+merge/96815 | 19:33 |
| dobey | whee | 19:33 |
| gatox | ralsina, i think you forgot to add your +1 here: https://code.launchpad.net/~diegosarmentero/ubuntu-sso-client/window-title/+merge/96802 | 19:52 |
| ralsina | gatox: oops, there | 19:52 |
| gatox | ralsina, thanks | 19:52 |
| gatox | ok, EOD here!! see you on monday people! have a great weekend! | 20:05 |
| alecu | nessita, ping | 20:16 |
| alecu | nessita, "the whole ubuntuone.proxy namespace will be distributed in another binary package" | 20:16 |
| alecu | nessita, in fact, the tunnel_client.py *needs* to be in the ubuntuone-client package where syncdaemon is. | 20:16 |
| alecu | nessita, tunnel_client.py and bin/ubuntuone-proxy-tunnel can be in the different package. | 20:18 |
| alecu | sorry | 20:18 |
| alecu | nessita, tunnel_server.py and bin/ubuntuone-proxy-tunnel can be in the different package. | 20:18 |
| alecu | since tunnel_server is the only one that depends on Qt. | 20:18 |
| alecu | and tunnel_client can work just fine if bin/ubuntuone-proxy-tunnel is not found (minus proxy support, that is) | 20:19 |
| nessita | alecu: ah, ok... | 20:27 |
| nessita | that's a little more complex than I thought, packaging and concept wise | 20:27 |
| nessita | alecu: why does u1client need tunnel_client if no proxy nor proxy process available? | 20:28 |
| alecu | nessita, tunnel_client is pure twisted. There's a class in it named TunnelRunner, (that starts the qt tunnel process). | 20:30 |
| alecu | nessita, a TunnelRunner has a get_client method that returns either a tunnel client or the default reactor. | 20:31 |
| === zyga is now known as zyga-weekend | ||
| alecu | in action queue, in the bits where reactor.connectTCP or connectSSL is called, a "client" is used instead, gotten from the tunnel runner. | 20:31 |
| dobey | alecu! | 20:32 |
| nessita | alecu: why not just having the tunnel runner in the default install, and not the tunnel_client | 20:32 |
| nessita | ? | 20:32 |
| dobey | alecu: is bug #703100 still relevant? | 20:32 |
| ubot5` | Launchpad bug 703100 in Ubuntu One Client "pynotify is a non-asynchronous library" [Medium,Triaged] https://launchpad.net/bugs/703100 | 20:32 |
| alecu | nessita, that can be done, sure. | 20:32 |
| alecu | hey thre dobey. | 20:32 |
| nessita | alecu: would you think is cleaner? if we put the tunnelrunner outside proxy we can have the whole proxy namespace in another bianry package | 20:33 |
| alecu | dobey, I think it's still relevant, yes, since the whole twisted loop in SD is waiting every time pynotify is called. | 20:34 |
| alecu | nessita, let me take a better look. | 20:34 |
| nessita | alecu: thanks | 20:34 |
| alecu | nessita, but a priory, yes. | 20:34 |
| dobey | alecu: but it's like 5ms per call, right, and we don't call it very often… | 20:34 |
| alecu | dobey, how did you measure "5ms"? | 20:35 |
| alecu | dobey, I agree we don't call it very often, yes. | 20:36 |
| dobey | alecu: i didn't, hence the "like 5ms" meaning "it's obviously very small" | 20:36 |
| alecu | dobey, perhaps we can lower the priority? | 20:36 |
| alecu | dobey, the bug priority I mean. | 20:36 |
| dobey | alecu: well, given the first number for the bug is 7, and it's not fixed yet, i'm pretty sure it has a low priority :) | 20:37 |
| alecu | dobey, if we measure it to be really fast, then I agree with removing the bug. | 20:37 |
| dobey | alecu: do you know how we can measure it? | 20:38 |
| alecu | dobey, yes. But not by hard :-) | 20:38 |
| dobey | alecu: write a test that logs the times? | 20:39 |
| alecu | dobey, or time.time() - start_time, and logging that. | 20:39 |
| alecu | dobey, are you interested in solving that bug *today*? | 20:39 |
| dobey | alecu: no. i was looking for another bug which i did fix already, and that was in the results list, so i thought i'd ask since you filed it, and it's assigned to you :) | 20:40 |
| thisfred | in general, performance testing in python is best done in pystones, since actual speed is *very* machine dependent. Also usually it's best to call the function you're testing at least a few thousand times, and throw away the outliers, since the first call can be disproportionally expensive | 20:49 |
| thisfred | but if you're looking for a ballpark figure, likely none of this matters ;) | 20:50 |
| JanC | thisfred: timing a function many times is general good benchmarking practice, not only for Python ;) | 20:51 |
| JanC | calling it many times to time it | 20:51 |
| thisfred | sure :) | 20:51 |
| thisfred | http://code.activestate.com/recipes/440700-performance-testing-with-a-pystone-measurement-dec/ for a pystone example | 20:52 |
| alecu | nessita, so, where do you suggest that we should put the runner in? | 20:54 |
| ralsina | I think I have made all the widgets stop moving when focused, and all orangeish squares disappear except for the checkboxes | 20:54 |
| ralsina | As well as getting rid of the icon-in-code in GotoWebButton | 20:54 |
| alecu | nessita, also this should also work if *all* of the code was in the syncdaemon package proper. | 20:55 |
| ralsina | OTOH, I want to poke my eyes with my mate because it was a pain in the arse to get "right" and noone is going to like it on monday ;-) | 20:55 |
| alecu | nessita, since the runner is resilient to the tunnel process crashing because Qt is not installed. | 20:55 |
| alecu | ralsina, we may be able to not like it earlier, if you point us at a branch :-) | 20:56 |
| alecu | ralsina, no need to wait till monday! | 20:56 |
| nessita | alecu: I did not follow you, sorry | 20:56 |
| ralsina | alecu: feel free to hate it as soon as you like ;-) https://code.launchpad.net/~ralsina/ubuntuone-control-panel/tweaks | 20:57 |
| dobey | thisfred: mostly i care about "is this ever going to actually be a problem in any case we care about" | 20:57 |
| alecu | nessita, let's say for a moment that we *do not* create a different binary package for the tunnel process. | 20:57 |
| nessita | alecu: ok | 20:57 |
| alecu | nessita, and stuff all of it into the u1-client package. | 20:57 |
| dobey | thisfred: and i'm pretty sure the answer is "no" and we're calling all kinds of other sync APIs without any similar worry | 20:57 |
| nessita | alecu: aja | 20:57 |
| alecu | nessita, what would happen? | 20:57 |
| nessita | alecu: ImportErrror on Qt is my guess | 20:57 |
| dobey | alecu: i would say no | 20:57 |
| alecu | nessita, the runner will try to start the tunnel process | 20:57 |
| nessita | alecu: before that | 20:58 |
| alecu | nessita, and the tunnel process will fail | 20:58 |
| dobey | unless you are going to handle qt not being there | 20:58 |
| nessita | alecu: ImportError, no? | 20:58 |
| alecu | nessita, but the runner knows how to deal with the tunnel process not starting | 20:58 |
| thisfred | dobey: right, just responding to one of my python "triggers" ;) | 20:58 |
| alecu | nessita, it just ignores it, and uses the default reactor, just like if proxies are not configured. | 20:58 |
| nessita | alecu: we can't release to ubuntu a code that import qt4 and then not list that on Depends | 20:58 |
| alecu | nessita, what if it just "recomends" it? | 20:59 |
| nessita | alecu: the same, recommends pulls packages in | 20:59 |
| alecu | nessita, also, what if we release this on nightlies? | 20:59 |
| dobey | we can't depends/recommends on pyqt in ubuntuone-client | 20:59 |
| dobey | nightlies should be as close to proper ubuntu packages as possible | 20:59 |
| nessita | alecu: I think is much much cleaner to have a separated binary package for the tunnel, where we can set the dependencies properly | 21:00 |
| alecu | nessita, dobey: ok | 21:00 |
| dobey | it should be a separate binary package | 21:00 |
| nessita | alecu: that's also what we commit to with the ubuntu-release team... | 21:00 |
| === jalcine is now known as JackyAlcine | ||
| nessita | alecu: so, I would put *just* the runner, which is more like a chooser, in ubuntuone/tunnel_safe_not_crashing_chooser_runner.py | 21:01 |
| === JackyAlcine is now known as jalcine | ||
| nessita | and import that from action queue. And make that module, check if it has available whatever it needs, and start the super-duper-tunnerl server if available | 21:01 |
| nessita | alecu: and also check if there is a concrete need of using it, no? | 21:02 |
| dobey | how is it starting it? | 21:02 |
| dobey | hard coding the path? | 21:02 |
| nessita | alecu: we should not start it if there is no proxy support need | 21:02 |
| nessita | dobey: no, it uses LIBEXEC | 21:02 |
| alecu | nessita, the only way to know if proxy support is needed is trying to start it. | 21:02 |
| alecu | nessita, because we can't have qt stuff on sd | 21:02 |
| dobey | nessita: from clientdefs? | 21:03 |
| nessita | dobey: yes | 21:03 |
| dobey | ok, good | 21:03 |
| nessita | alecu: but what if I have qt, and no proxy needs? will I get the tunnel proxy running? | 21:03 |
| nessita | dobey: from https://code.launchpad.net/~alecu/ubuntuone-client/proxy-tunnel-useit/+merge/96213 | 21:04 |
| nessita | 319+ from ubuntuone.clientdefs import LIBEXECDIR | 21:04 |
| nessita | 320+ return path.join(LIBEXECDIR, TUNNEL_EXECUTABLE) | 21:04 |
| ralsina | EOD for me, I'll send some emails about the style changes I am trying late tonight | 21:04 |
| nessita | ralsina: ack | 21:04 |
| dobey | nessita: ok. i didn't need the paste :) | 21:04 |
| alecu | nessita, if you have qt installed, and have no proxies configured, the tunnel process will be started, will check if proxies are needed and then die. And the runner will find out about this and not use the tunnel. | 21:04 |
| nessita | dobey: well, I prefer you noticing that code now and not when we're releasing :-) | 21:04 |
| nessita | alecu: ah, ok then | 21:05 |
| dobey | ugh, where are these ubuntuone icons at | 21:05 |
| dobey | grr | 21:05 |
| nessita | alecu: so, I think what I mentioned before may work? what do you think? | 21:05 |
| alecu | nessita, I think I can make it work. | 21:07 |
| nessita | alecu: is it too much overhead for you? can I help somehow? | 21:07 |
| alecu | nessita, I would have liked this to get to nightlies, so we can start to test it. | 21:08 |
| alecu | nessita, and work on this change on a different branch. | 21:08 |
| nessita | alecu: the proxy namespace is not being distributed ATM in mightlies | 21:09 |
| alecu | nessita, ok. | 21:09 |
| nessita | alecu: so this branch, when landing, will just broke nightlies | 21:09 |
| nessita | so, we should also fix the packaging in nightlies as close as ubuntu | 21:10 |
| nessita | as close as how ubuntu will have it, I mean | 21:10 |
| alecu | ok, perfect. | 21:10 |
| nessita | otherwise we're missing one important piece of the testing (not having/having the thing available) | 21:10 |
| nessita | alecu: sorry for the delay... I know you wanted this ready today | 21:10 |
| alecu | nessita, no problem: I'm just one review from landing this, so it's much closer than I've been all week :-) | 21:11 |
| nessita | :-) | 21:11 |
| alecu | nessita, I'll finish this fix tonight, and you can review it first thing in the morning, and then we'll try to get nightlies fixed in the week. | 21:12 |
| alecu | I mean, "first thing *monday* morning" | 21:12 |
| nessita | alecu: +1 | 21:12 |
| nessita | of course ;-) | 21:12 |
| alecu | nessita, thanks! | 21:12 |
| nessita | alecu: thank you :-) | 21:12 |
| dobey | hmm | 21:27 |
| dobey | hrmm | 21:53 |
| dobey | i can't find these icons | 21:54 |
| nessita | bye all! | 22:00 |
| dobey | indeed | 22:04 |
| thisfred | later! | 22:07 |
| dobey | later | 22:16 |
| === yofel_ is now known as yofel | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!