These are chat archives for riot/riot

3rd
Jul 2015
Tsutomu Kawamura
@cognitom
Jul 03 2015 03:37
Sauce Labs.png
is there a way to re-run the saucelabs tests?
this result was due to the error inside the saucelabs, I think.
Gianluca Guarini
@GianlucaGuarini
Jul 03 2015 13:21
@cognitom go here https://travis-ci.org/riot/riot and restart the build
Tsutomu Kawamura
@cognitom
Jul 03 2015 13:22
Oh, thanks. I’ll try it, right now.
Gianluca Guarini
@GianlucaGuarini
Jul 03 2015 13:22
@cognitom @tipiirai could we remove these lines https://github.com/riot/riot/blob/master/lib/browser/tag/update.js#L7-L10?
Tsutomu Kawamura
@cognitom
Jul 03 2015 13:23
riot:riot.png
Gianluca Guarini
@GianlucaGuarini
Jul 03 2015 13:24
great
Tsutomu Kawamura
@cognitom
Jul 03 2015 13:25
Some one (maybe @GianlucaGuarini?) has already reloaded. Thanks.
Gianluca Guarini
@GianlucaGuarini
Jul 03 2015 13:25
I did it some minutes ago
Tsutomu Kawamura
@cognitom
Jul 03 2015 13:25
:+1:
I’m checking that line...
Gianluca Guarini
@GianlucaGuarini
Jul 03 2015 13:26
why do we need to redefine the event object? I don't know why someone should ever do this
probably these lines should be in a try and catch as well
Tsutomu Kawamura
@cognitom
Jul 03 2015 13:43

Hmm, IE8 needs e.target = e.srcElement
https://msdn.microsoft.com/library/ff974946(v=vs.85).aspx

But it seems that we shold this in the polyfill.

Gianluca Guarini
@GianlucaGuarini
Jul 03 2015 13:46
I thought we do not support ie8 anymore
Tsutomu Kawamura
@cognitom
Jul 03 2015 13:46
Then, we don’t need this line at all.
yeah right I would like to check this with @tipiirai first
probably he knows something we do not know
Tsutomu Kawamura
@cognitom
Jul 03 2015 13:50
Yeah.
BTW, MDN says "use key instead of charCode…"
https://developer.mozilla.org/en-US/docs/Web/API/KeyboardEvent/key
Gianluca Guarini
@GianlucaGuarini
Jul 03 2015 13:51
the problem is that with use strict e.which is also read only
Tsutomu Kawamura
@cognitom
Jul 03 2015 13:51
i c
I’m checking supported browser and versions...
Gianluca Guarini
@GianlucaGuarini
Jul 03 2015 13:54
good because I would like to release a new riot release this weekend
the latest has still some important bugs
Tsutomu Kawamura
@cognitom
Jul 03 2015 13:58
IE9 supports e.which and IE8 doesn’t
The problem is that event.which is already deplecated as the web standard.
Gianluca Guarini
@GianlucaGuarini
Jul 03 2015 14:00
what happens then if we keep them in a try catch block?
Tsutomu Kawamura
@cognitom
Jul 03 2015 14:01
I’m not sure, but use strict would be still a problem even if in catch block.
have you try that already?
Gianluca Guarini
@GianlucaGuarini
Jul 03 2015 14:02
hmm... I need to test and try to understand what they tried to achieve when they wrote that code
Tsutomu Kawamura
@cognitom
Jul 03 2015 14:05
Hmm..
All of the key events are readonly.
https://developer.mozilla.org/en-US/docs/Web/API/KeyboardEvent
Gianluca Guarini
@GianlucaGuarini
Jul 03 2015 14:06
so that line makes no sense!
Tsutomu Kawamura
@cognitom
Jul 03 2015 14:09
we can use e.key on FF, Chrome, IE9, but can’t on Safari !!! :cry:
Gianluca Guarini
@GianlucaGuarini
Jul 03 2015 14:10
:worried:
Tsutomu Kawamura
@cognitom
Jul 03 2015 14:11
“safari is next IE”...
if we can drop IE8, it seems the best to remove these lines. I agreed.
https://github.com/riot/riot/blob/master/lib/browser/tag/update.js#L7-L10?
Gianluca Guarini
@GianlucaGuarini
Jul 03 2015 14:13
ok let's give it a try
I will maybe release 2.2.2-beta
Tsutomu Kawamura
@cognitom
Jul 03 2015 14:13
:+1:
Tsutomu Kawamura
@cognitom
Jul 03 2015 14:41
how about exporting setEventHandler to riot.hackable.setEventHandler or somewhere? then, we can tweak that in external file with <!--[if lte IE 8]>
if polyfill script is external, strict mode will not be a matter.
Jens Anders Bakke
@cfenzo
Jul 03 2015 14:50
:+1: for exporting <=IE8 hacks to external files
Brandon Wilhite
@JediMindtrick
Jul 03 2015 20:15
@cfenzo Just saw your comment, will give that a try.