[Sugar-devel] sugar-web: env.js questions
Rogelio Mita
rogeliomita at activitycentral.com
Fri Nov 22 16:59:25 EST 2013
2013/11/22 Daniel Narvaez <dwnarvaez at gmail.com>
> On 22 November 2013 21:20, Rogelio Mita <rogeliomita at activitycentral.com>wrote:
>
>>
>> 2013/11/22 Daniel Narvaez <dwnarvaez at gmail.com>
>>
>>> Yes the callback is never executed because there is no python
>>> notification. And our test succeeds because it doesn't make sure that the
>>> callback is called.
>>
>>
>> Ok, I see...
>>
>> I am watching
>> https://github.com/sugarlabs/sugar-toolkit-gtk3/blob/master/src/sugar3/activity/webactivity.py#L89and I can see if we are in standalone mode this fragment of code is not
>> executed (obvious because the activity is not started from sugar) then we
>> don't have *window.suga*r global var, and I think that it is good and
>> need be maintained through all cycle (is understood?).
>> I mean, if you are in *standalone mode* you should not exist *"global"
>> sugar var*,
>>
>
> Yeah I think setting that variable is wrong. I probably know why I did it
> but I was just missing something. We should remove that...
>
>
>>
>> and in the client side, the logic for *onStandalone* method *is check
>> that sugar var existence*.
>> Because I noted the behaviour when *sugar var* don't exists on client
>> side, this fragment<https://github.com/sugarlabs/sugar-web/blob/master/env.js#L13>creates this variable anyway and I do not understand why.
>>
>> I propose remove this fragment at the moment and make a method
>> activity.onStandalone (as was already discussed with Manu, I remember) and
>> the logic can be: checks existence of sugar var, make sence?
>>
>
> I think I was the one suggesting to add an isStandalone method. But after
> looking into it a bit more I tend tothink it's better to make
> getEnvironment pass null to the callback, it should make it easier to write
> code that works in both modes.
>
> About the logic, unfortunately checking window.sugar won't work because we
> can't assume window.sugar has been already set when getEnvironment is
> called... the python code you linked might be called after getEnvironment
> has been called and will erroneously think we are in standalone mode.
>
you are right. I thought for a moment that this fragment is executed to
begging before activity.setup(), I'm still in doubt, are you sure about
this?
>
> That's why I suggested to check if window.location starts with
> activity://. It's a bit ugly but it will always work.
>
Seems to be a good start, if I understand what you say, it would result
there is no need to use callbacks in this method, ie: we can return object
environment on sugar environment or null in other case, or better yet, ask
to Environment model about the mode which is...
var environment = env.getEnvironment(), though it sounds redundant...
OR!
var environment = activity.getEnvironment()
sounds better to me, and ask to environment object everything related to him
You likes some these ideas?
--
Roger
Activity Central <http://activitycentral.com/>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.sugarlabs.org/archive/sugar-devel/attachments/20131122/64142b55/attachment-0001.html>
More information about the Sugar-devel
mailing list