fluid-work IRC Logs-2013-10-29

Owned by Fluid IRC Log Bot
Last updated: Oct 29, 2013
10 min read

[08:29:20 CDT(-0500)] <cindyli> Justin_o: i've sent a pull request to fix the problem with aux builder at handling multiple composite panels per schema: https://github.com/fluid-project/infusion/pull/428

[08:29:26 CDT(-0500)] <cindyli> can you take a look when you have a chance?

[08:30:59 CDT(-0500)] <Justin_o> cindyli: will do, thanks

[08:31:14 CDT(-0500)] <cindyli> thanks, Justin_o

[08:35:14 CDT(-0500)] <colinclark> Hey yzen, do you think we're ready to put up a testing instance of the online Flow Manager up in our IDI cloud?

[08:56:52 CDT(-0500)] <yzen> colinclark: we have it there but it's not all up to date with kettle but it s the same thing

[08:57:00 CDT(-0500)] <yzen> http://flowmanager.gpii.net/carla/settings/%7B%22OS%22:%7B%22id%22:%22linux%22%7D,%22solutions%22:%5B%7B%22id%22:%22org.gnome.desktop.a11y.magnifier%22%7D%5D%7D

[09:06:21 CDT(-0500)] <colinclark> ok

[09:07:17 CDT(-0500)] <colinclark> So we could update it at some point without much trouble, yzen?

[09:07:26 CDT(-0500)] <colinclark> nice URL-escaped JSON, by the way (smile)

[09:07:28 CDT(-0500)] <yzen> correct

[09:07:32 CDT(-0500)] <colinclark> ok cool

[09:07:42 CDT(-0500)] <yzen> can't copy a non-encoded one from the browser (smile)

[09:07:59 CDT(-0500)] <colinclark> lol

[09:08:04 CDT(-0500)] <colinclark> good browser

[09:40:26 CDT(-0500)] <Justin_o> cindyli: your FLUID-5197 branch has been merged

[09:41:11 CDT(-0500)] <cindyli> great! thanks, Justin_o, anastasiac, you might be interested ^

[09:41:46 CDT(-0500)] <anastasiac> thanks

[09:44:18 CDT(-0500)] <Justin_o> anastasiac, cindyli: anything else that we might need to for GPII.. I want to do another upgrade to the prefsEditor repo..

[09:45:35 CDT(-0500)] <cindyli> nothing else from my side, Justin_o

[09:46:54 CDT(-0500)] <anastasiac> can't think of anything, Justin_o

[09:47:06 CDT(-0500)] <Justin_o> great thanks

[09:48:46 CDT(-0500)] <anastasiac> cindyli, nit-picky question: I notice we've renamed fluid.prefs.panels to fluid.prefs.panel, but we didn't make a comparable change to fluid.prefs.enactors? (not that I'm looking forward to another sweeping namespace change!!)

[09:50:47 CDT(-0500)] <cindyli> good finding, anastasiac. i agree that they should be consistent. Justin_o, thoughts?

[09:59:35 CDT(-0500)] <Justin_o> cindyli, anastasiac: i would prefer consistency

[10:00:21 CDT(-0500)] <cindyli> anastasiac, Justin_o, another jira to rename fluid.prefs.enactors?

[10:00:49 CDT(-0500)] <Justin_o> cindyli, anastasiac: i guess so, and lets get this in before we do the update to prefsEditors

[10:01:06 CDT(-0500)] <anastasiac> I'll file it

[10:02:46 CDT(-0500)] <anastasiac> cindyli, Justin_o: enactors renaming task created: http://issues.fluidproject.org/browse/FLUID-5198

[10:03:07 CDT(-0500)] <cindyli> thanks, anastasiac

[11:06:41 CDT(-0500)] <clown> jhernandez: do you time for a quick question?

[11:15:19 CDT(-0500)] <jhernandez> clown: sure!

[11:18:13 CDT(-0500)] <clown> just a quick question, jhernandez: did you see my pull request? looking up url

[11:18:26 CDT(-0500)] <jhernandez> of course

[11:18:27 CDT(-0500)] <clown> https://github.com/javihernandez/linux/pull/3/files

[11:18:39 CDT(-0500)] <clown> good!

[11:18:48 CDT(-0500)] <clown> I imagine you are very busy with the pilot stuff.

[11:18:48 CDT(-0500)] <jhernandez> didn't get the chance to review it

[11:18:59 CDT(-0500)] <jhernandez> yes

[11:19:08 CDT(-0500)] <clown> that's fine. I wanted to make sure you got the email.

[11:19:16 CDT(-0500)] <jhernandez> yes

[11:19:24 CDT(-0500)] <jhernandez> and thanks, btw

[11:19:26 CDT(-0500)] <jhernandez> :]

[11:19:49 CDT(-0500)] <clown> cool. the email did not go missing. All right, when you get a chance…. Thanks.

[11:19:49 CDT(-0500)] <jhernandez> it's a great improvement

[11:20:05 CDT(-0500)] <clown> still needs work, though (I'm not satisfied).

[11:20:11 CDT(-0500)] <clown> but it's a start.

[11:20:51 CDT(-0500)] <jhernandez> yes

[11:21:49 CDT(-0500)] <jhernandez> we're making little progresses, but with a lot of them, we'll end up with a big one

[11:21:52 CDT(-0500)] <jhernandez> :]

[11:22:27 CDT(-0500)] <clown> yes. incremental development. (smile)

[11:25:22 CDT(-0500)] <jhernandez> yeah

[11:25:23 CDT(-0500)] <jhernandez> and

[11:26:02 CDT(-0500)] <jhernandez> in terms of device reporting, we can start thinking on different ones

[11:26:31 CDT(-0500)] <jhernandez> ie, looking for manually installed software

[11:26:41 CDT(-0500)] <jhernandez> sounds funny, isn't it?

[11:26:41 CDT(-0500)] <jhernandez> (tongue)

[11:27:51 CDT(-0500)] <jhernandez> the funny thing is, look for an installed software which was installed without the package manager

[11:29:48 CDT(-0500)] <jhernandez> ie: looking for a well-known binary, under one of the paths from PATH environmental variable

[11:59:21 CDT(-0500)] <clown> jhernandez: sorry, I had to step away. catching up

[11:59:57 CDT(-0500)] <jhernandez> np

[12:00:17 CDT(-0500)] * clown tries that scenario (looking for a package installed by yum, not by packagekit).

[12:00:18 CDT(-0500)] <jhernandez> I'm playing with android'd GPII UI

[12:00:28 CDT(-0500)] <jhernandez> (busy)

[12:01:12 CDT(-0500)] <jhernandez> clown: well, by without package manager, I was meaning not installed by yum

[12:01:13 CDT(-0500)] <jhernandez> :]

[12:02:31 CDT(-0500)] <clown> oh… installed "manually" then? Like when I clone, build, and install accerciser?

[12:04:20 CDT(-0500)] <jhernandez> clown: got it! (smile)

[12:06:06 CDT(-0500)] <clown> jhernandez: something to think about, then. How to handle that case. Hmmm….

[12:06:12 CDT(-0500)] * clown steam comes out of clown's ears.

[12:06:27 CDT(-0500)] <jhernandez> xDD

[12:06:49 CDT(-0500)] <jhernandez> yeah, maybe looking for the well-known binary in differnt well-known OS's paths

[12:12:27 CDT(-0500)] <clown> yeah.

[12:12:55 CDT(-0500)] * clown wonders if well-known paths are the same across different linux desktop distros. Also about GNOME vs. KDE.

[12:15:26 CDT(-0500)] <jhernandez> clown: PATH environmental variable sounds cross-distro enough to you? (wink)

[12:15:45 CDT(-0500)] <jhernandez> echo $PATH

[12:15:48 CDT(-0500)] <jhernandez> :]

[12:16:03 CDT(-0500)] <clown> jherenandez, yup. Also, "which" is your friend, e.g., "which accerciser".

[12:16:10 CDT(-0500)] <jhernandez> well, another way is by using which

[12:16:17 CDT(-0500)] <jhernandez> of course!

[12:16:30 CDT(-0500)] <jhernandez> you were faster than me!

[12:16:59 CDT(-0500)] <jhernandez> but yes, "which" command looks for the given binary in system's PATH

[12:19:55 CDT(-0500)] <clown> true...

[12:20:09 CDT(-0500)] <clown> well, there's always "find". Whoo hoo!

[12:22:58 CDT(-0500)] * jhernandez loves find|grep

[12:29:30 CDT(-0500)] * clown notes that many commands have a "--version" option for determining the version (if found).

[12:31:58 CDT(-0500)] <jhernandez> nice catch!

[12:33:13 CDT(-0500)] <clown> jhernandez: alas, accerciser does not support —version. Odd. But not really all that important.

[12:34:12 CDT(-0500)] <jhernandez> xDDDDD

[12:34:17 CDT(-0500)] <jhernandez> sorry for that

[12:34:18 CDT(-0500)] <jhernandez> (tongue)

[12:34:36 CDT(-0500)] <jhernandez> I think it never had this commandline option

[12:35:01 CDT(-0500)] <clown> jhernandez: do a "man accerciser". For fun...

[12:35:31 CDT(-0500)] <jhernandez> xDDDD

[12:35:53 CDT(-0500)] <jhernandez> then I should review the commits log

[12:36:00 CDT(-0500)] <clown> however, in general, for apps that do support 'version' option, it (nicely) does not actually launch the program, it merely returns the version number and exits.

[12:36:10 CDT(-0500)] <jhernandez> yup

[12:36:22 CDT(-0500)] * clown makes notes

[12:45:22 CDT(-0500)] <clown> jhernandez: this looks promising — a way to run shell scripts from within node.js: http://www.2ality.com/2011/12/nodejs-shell-scripting.html

[12:45:46 CDT(-0500)] <clown> maybe this is old news to you.

[12:46:48 CDT(-0500)] * jhernandez looks

[12:49:44 CDT(-0500)] <jhernandez> clown: yeah

[12:49:54 CDT(-0500)] <jhernandez> I know all these tips

[12:50:37 CDT(-0500)] <jhernandez> in fact, we use some of them in GPII

[13:15:09 CDT(-0500)] <clown> jhernandez: re: "in fact, we use some of them in GPII": can you give an url to one one of these uses? Only do this, when you want to take a break from Android — much more important.

[13:23:42 CDT(-0500)] <jhernandez> clown: np

[13:24:24 CDT(-0500)] <jhernandez> regarding android, I'm integrating my prototype into GPII's android app code

[13:24:39 CDT(-0500)] <jhernandez> http://snag.gy/OCr8n.jpg

[13:24:49 CDT(-0500)] <jhernandez> :]

[13:25:10 CDT(-0500)] <jhernandez> and, regarding node os related stuff

[13:25:22 CDT(-0500)] <jhernandez> ie: https://github.com/GPII/universal/blob/master/gpii/node_modules/lifecycleActions/src/LifecycleActions.js

[13:25:44 CDT(-0500)] <jhernandez> we make a wrapper for child_process.spawn

[13:26:46 CDT(-0500)] <jhernandez> ie2: https://github.com/GPII/linux/blob/master/node_modules/orca/orcaSettingsHandler.js

[13:27:17 CDT(-0500)] <jhernandez> there you'll find examples of readFileSync, writeFileSync, and IIRC, some operation with paths

[13:27:51 CDT(-0500)] <jhernandez> https://github.com/GPII/linux/blob/master/node_modules/orca/orcaSettingsHandler.js#L34-L36

[13:28:41 CDT(-0500)] <jhernandez> in the highlighted code from the last link, you can see how we make the path to orca's user-settings.conf

[13:38:42 CDT(-0500)] <clown> thanks jhernandez. Off to another meeting. Later...

[14:36:14 CDT(-0500)] <anastasiac> cindyli1, Justin_o, I have a question about the new string handling. Which one of you should I ask?

[14:42:15 CDT(-0500)] <colinclark> oh hey jhernandez, are you still around?

[14:47:21 CDT(-0500)] <Bosmon> Hi cindyli1

[14:47:40 CDT(-0500)] <Bosmon> Did you have ideas about the modelRelay work?

[14:47:43 CDT(-0500)] <cindyli1> hi Bosmon

[14:51:20 CDT(-0500)] <cindyli1> yes, Bosmon. anastasiac, Justin_o and i talked about it and agreed modelRelay should be the way to go. but since with model relay, the internal path "value" needs to be specified in rules. so this change can wait until the model relay goes into the framework. I'm going to close my pull request. what do you think?

[14:54:50 CDT(-0500)] <jhernandez> colinclark: yeah

[14:55:02 CDT(-0500)] <colinclark> I just thought of a question for you

[14:55:10 CDT(-0500)] <colinclark> We were chatting with kasparnet yesterday

[14:55:24 CDT(-0500)] <colinclark> about how we can a) make master more stable and b) cut releases much more frequently

[14:55:35 CDT(-0500)] <colinclark> since we seem to struggle to cut releases at all (smile)

[14:56:01 CDT(-0500)] <colinclark> jhernandez: Do you think there might be some way that we could distribute a VM with an emulator or something running Android on it?

[14:56:16 CDT(-0500)] <jhernandez> mmm

[14:56:26 CDT(-0500)] <colinclark> kasparnet is in the midst of creating pre-built VirtualBox VMs for both Windows and Linux for quick testing

[14:56:36 CDT(-0500)] <jhernandez> what about just providing the apk ant the js.tar.gz?

[14:56:39 CDT(-0500)] <colinclark> if the acceptance tests pass there, you've got some reasonable indication that the system works

[14:56:53 CDT(-0500)] <colinclark> jhernandez: That's better than nothing, yes

[14:57:23 CDT(-0500)] <colinclark> And then it'd be as simple as an "adb install" and an "adb push"

[14:57:33 CDT(-0500)] <jhernandez> yes, but the point is that we would like to run it as a prvivileged app, into a rooted phone

[14:57:51 CDT(-0500)] <colinclark> yes

[14:57:56 CDT(-0500)] <colinclark> which is why a VM or something would be great

[14:58:01 CDT(-0500)] <jhernandez> yup

[14:58:21 CDT(-0500)] <Bosmon> cindyli1 - what about the idea of using your own modelRelay for this implementation?

[14:58:28 CDT(-0500)] <jhernandez> but, the only way could be by providing a system image to run in an android emulator

[14:58:42 CDT(-0500)] <jhernandez> but, not sure about the complexity of doing that

[14:58:46 CDT(-0500)] <colinclark> ah, interesting

[14:58:54 CDT(-0500)] <colinclark> I can do a bit of research

[14:59:58 CDT(-0500)] <jhernandez> aaand, about the linux image, it's ok, but we can use our fedora repos to create a spin with it

[15:00:08 CDT(-0500)] <jhernandez> it == the GPII packages

[15:01:01 CDT(-0500)] <jhernandez> no VMs, just grab the iso, flash into an USB drive, and boot your computer with it

[15:02:56 CDT(-0500)] <cindyli1> Bosmon: the goal of the jira is to get rid of the fact that text field slider requires a model path named "value". if this cannot be achieved, it's not a rush to change it. i can certainly use my own model relay. not a big change anyway

[15:03:01 CDT(-0500)] <jhernandez> VMs suck at graphics rendering stuff, you have to deal with USB ports to use them, and/or with the sound output from the VM

[15:03:36 CDT(-0500)] <colinclark> yes

[15:03:40 CDT(-0500)] <colinclark> they do generally suck

[15:03:56 CDT(-0500)] <colinclark> that seems reasonable, too

[15:04:02 CDT(-0500)] <colinclark> jhernandez: is it difficult to do?

[15:04:17 CDT(-0500)] <Bosmon> cindyli1 - surely the goal is to avoid EXPOSING the fact that the textfieldSlider requires a model path named "value" .......

[15:04:33 CDT(-0500)] <jhernandez> colinclark: I can quickly draft up the steps to address this

[15:04:49 CDT(-0500)] <colinclark> I'd be curious

[15:08:00 CDT(-0500)] <cindyli1> Bosmon: with model relay, "value" is not exposed if integrators are using fluid.prefs.textfieldSlider, but if fluid.textfieldSlider component is directly used, the "value" path is required.

[15:08:21 CDT(-0500)] <cindyli1> i think 5190 is to solve the latter

[15:08:22 CDT(-0500)] <cindyli1> Bosmon: btw, your comments for https://github.com/fluid-project/infusion/pull/425 has been address and ready for another round of review.

[15:09:02 CDT(-0500)] <jhernandez> colinclark: Every time we want to create a new spin with the latest code we should:

[15:09:05 CDT(-0500)] <jhernandez> 1.- Update GPII (and related) packages in GPII:Factory repo with the latest code from Github

[15:09:08 CDT(-0500)] <jhernandez> 2.- Once the packages are updated, publish a new version of the GPII:Factory repo

[15:09:11 CDT(-0500)] <jhernandez> 3.- Run a script to build the iso (this script will be done once, and reused every time we want to re-spin)

[15:09:14 CDT(-0500)] <jhernandez> 4.- After tesing the new ISO, Put the image somewhere and available to be downloaded

[15:10:24 CDT(-0500)] <jhernandez> btw, with GPII:Factory I'm talking about -> https://build.opensuse.org/project/show/home:javihernandez:GPII:Factory

[15:11:05 CDT(-0500)] <Bosmon> cindyli1 - thanks for looking at that one, and also for your test case, which looks valid

[15:11:17 CDT(-0500)] <Bosmon> Could you work around the "" issue for now, by replacing the "" with ""fluid.prefs.enactors" ?

[15:11:21 CDT(-0500)] <Bosmon> I think it would be more correct anyway

[15:11:46 CDT(-0500)] <colinclark> jhernandez: How does one do #1?

[15:11:48 CDT(-0500)] <Bosmon> You really only want to hit enactor subcomponents, not any others that might appear there

[15:13:02 CDT(-0500)] <jhernandez> colinclark: you can do it manually from the OBS's web interface, or (my preferred) use the commandline in a svn-like way

[15:13:34 CDT(-0500)] <jhernandez> OBS it's something similar to github, but for packages

[15:14:55 CDT(-0500)] <jhernandez> you can branch packages from projects, you can create pull-requests (aka: submit-request) back, etc

[15:15:46 CDT(-0500)] <jhernandez> #2 is easy, set the repo's publish flag to on

[15:16:01 CDT(-0500)] <jhernandez> #3 run a script and wait

[15:16:52 CDT(-0500)] <jhernandez> and when we want to freeze a release, we just branch Factory as GPII:STABLE:0.x

[15:17:00 CDT(-0500)] <colinclark> interesting

[15:17:08 CDT(-0500)] <colinclark> Do you need particular credentials?

[15:17:15 CDT(-0500)] <colinclark> And is that something that we can share, for example?

[15:17:22 CDT(-0500)] <jhernandez> it's like github

[15:17:24 CDT(-0500)] <jhernandez> it's public

[15:17:53 CDT(-0500)] <jhernandez> you can fork the code (aka: packages), create sr's, etc ...

[15:18:31 CDT(-0500)] <jhernandez> the funny thing, is that you can build the packages locally before commiting anything

[15:19:06 CDT(-0500)] <colinclark> ah, cool

[15:19:07 CDT(-0500)] <jhernandez> and every time you make a commit, a build will be triggered on the obs

[15:19:40 CDT(-0500)] <jhernandez> you can also add services, ie: to listen to git commits, or tarballs, or more events

[15:20:48 CDT(-0500)] <Bosmon> Hi cindyli - you're back

[15:20:58 CDT(-0500)] <cindyli> yes i am

[15:20:58 CDT(-0500)] <Bosmon> I pasted some comments from whilst you were away onto the pull request : P

[15:21:07 CDT(-0500)] <cindyli> oh

[15:21:50 CDT(-0500)] <jhernandez> colinclark: this is the opensuse's running instance of the OBS https://build.opensuse.org/. But you can also grab the OBS code, and run your own instance of the OBS

[15:22:12 CDT(-0500)] <colinclark> aha, interesting

[15:27:00 CDT(-0500)] <jhernandez> anyway, the "rpm packages" thing is something that I'm trying to not to do by mysef, but a colleague from emergya

[15:28:48 CDT(-0500)] <jhernandez> i have to teach him about how to make it, but once he does it the first time, it's an easy and repetitive task

[15:29:44 CDT(-0500)] <jhernandez> colinclark: ^

[15:29:58 CDT(-0500)] <colinclark> ok

[15:30:05 CDT(-0500)] <colinclark> jhernandez: Will you be at the architecture meeting tomorrow?

[15:30:20 CDT(-0500)] <jhernandez> colinclark: I hope so

[15:30:33 CDT(-0500)] <colinclark> I'm curious to hear more about it then, if you get the chance

[15:30:40 CDT(-0500)] <colinclark> it seems pretty cool

[15:30:49 CDT(-0500)] <jhernandez> heh

[15:30:53 CDT(-0500)] <jhernandez> yes it is!

[15:34:39 CDT(-0500)] <cindyli> thanks for the comment, Bosmon, will take care of that (smile)

[19:13:10 CDT(-0500)] * Topic is 'This channel is logged – https://botbot.me/freenode/fluid-work/ for more details and older logs see: http://wiki.fluidproject.org/display/fluid/IRC+Channel' set by avtar!avtar@nat/ocadu/x-xeesjmijqzzwdkjt on 12:48:45 CDT(-0500)