#selenium

lukeis: is there anything blocking a java release atm?

kallisti: nothing it calls would result in that exception being thrown

kallisti: it calls _element_if_visilbe

AutomatedTester: i don't think so

ah okay. what things would throw element not visible then?

that too; no exceptions there

kallisti: .text .click .send_keys

kallisti: element.click() or another interaction like send_keys

dude-x: not .text

lukeis: you sure about .terxt?

hmm

yes

that's interersting...

it will return an empty string

so, should we move to 2.48 with this next release/

lukeis: yeah you're right.

yeah so I don't really need an analogous for visibility_of_element_located because the libary already lets you write: expect =<< isDisplay element

all I need to add special functions for are conditions where you need to catch a specific error, mostly StaleElementReference

*analog

*isDisplayed

typing is hard

Was the idea for 2.48 to have Marionette support included?

titusfortner: not necessarily but I am doing a release

i saw wires 0.4

AutomatedTester: https://github.com/jgraham/wires is still the official repo, right?

yes

I created 0.4 today

no linux release? ;)

not yet, will be there soon

learn vagrant or docker :)

AutomatedTester: now, i'm back too!

:)

i think i answered my initial question, which is that wires handles all the port setup goo to get firefox to talk to the proxy. so that's cool.

the next question is should i submit a bug to add a /status end point?

jimevans: submit it to the specification

the bug

copy that.

it's not in marionette because it is not in the spec

right. i figured.

but /shutdown appears to be… :)

or maybe not.

i could be wrong.

I dont think so

does quitting the session cause the executable to terminate?

I think so

the shim manages the firefox binary

ah. that'd explain how the exe terminates properly then.

yeah, but the shim exits too.

let me take a quick look

hmmm... I think I might have broken my toe...

that's suboptimal

indeed

quit() does indeed appear to stop the shim.

Is there a particular Firefox build I should be using for testing? Or just "nightly"

(Marionette testing with wires 0.4)

could use the current release? :)

it 'works'

oh yeah?

does for me...

I didn't know where we were in that phase

ok

maybe not all the features implemented

well, I need to start with getting a session

:-D

AutomatedTester: how do the python bindings make sure the HTTP server is started after launching the wires executable? or do they just assume it's working, and ignore the potential race condition?

I think I ignore the potential race

jimevans: I check the port is free and then assign the port and then launch the executable and wait for the port to still be blocked

ignoring the potential race

[13selenium] 15lukeis pushed 1 new commit to 06master: 02http://git.io/vctjg

13selenium/06master 1401cbd65 15Laurence Rowe: Python 3 compatibility for remote Authorization...

ooh and a new edgedriver

yep. hot off the presses.

lukeis: that just looks weird to encode and decode

yes, yes it does

but it's a py2 and 3 compatible way to deal with the base64 encoding...

maybe it's one of those libs left behind.

AutomatedTester: bug added. :)

jimevans: ta muchly

hi, i'm wondering what value i set a file input to for performing an upload

dmarr: path to the file

seems like there is no easy way to automate finding that

dmarr: huh?

the file is where you put it

i'm gonna be running this test in different environments

you dont need the full path

I dont think...

uploading is rarely something I test with selenium

dmarr, sendKeys(file_absolute_path)

looks like i can use node to resolve the relative path

amazing http://i.imgur.com/BtPaztg.gif

LOL