Re: Cordova/W3C

[Josh Soref <js...@blackberry.com>]

Lisa wrote:
> I understand what you're asking but I'm not sure if that is something
> that the Cordova team is capable of helping with.


> We don't know how end users are using each of the Cordova APIs and there
>really
> isn't a good way to ask them.


> Cordova is installed using npm (https://www.npmjs.org/package/cordova)
> and the plugin repositories have publically visible download stats
> (http://plugins.cordova.io/#/_stats).

While this is true, it doesn¹t help Marcos. It certainly isn¹t W3¹s job to
try to find our customers.

> We do not have access to contact information for individual developers
> let alone the code they have written.

> There is the Google Group for Phonegap as well as the list of developers
> (opt-in by the developers themselves)

> as well as the Cordova blog where we could submit a poll.

While this is not likely to work well, regular surveys should probably be
done anyway.
The current approach to deprecation while technically fine ³I think it¹s
time to kill X², isn¹t a particularly friendly/useful/responsive approach.

A better approach in general would be to have regular surveys asking
people to submit:
- vendor url, vendor contact, appname, platforms, usage count by platform,
list of plugins they use, with a score for how happy they are for each
platform/plugin.

Building such a survey out of plugins.cordova.io/_stats should actually be
fairly easy, and the resulting data should be useful for Cordova in
general (and even the plugin developers).

> But the responses are unlikely to reach a good subset of users.

If you don¹t build it, they certainly won¹t come.
Let¹s try building it and see who comes.

> There are some case studies and featured applications from the Phonegap
>team (http://phonegap.com/case/).
   
PhoneGap build should be able to survey the things it (PhoneGap Build)
builds, as should other down streams which have access to stores (where
Cordova things get delivered).

Ideally Cordova would be able to reach its consumers and talk to them.
Talking to your consumers is very important, but notably fairly hard to do
(not many vendors manage to do this, or do it well).

Without having this information, Cordova really can¹t usefully sit at the
W3 table, as it clearly has absolutely no idea about what people are doing
and thus clearly has nothing to provide to W3.

Re: Cordova/W3C

[Marcos Caceres <w3...@marcosc.com>]

On Friday, February 28, 2014 at 4:55 PM, Brian LeRoux wrote:

> > Without having this information, Cordova really can¹t usefully sit at the
> > W3 table, as it clearly has absolutely no idea about what people are doing
> > and thus clearly has nothing to provide to W3.
>  
>  
> This is leading a little severe.  
Agree. PhoneGap/Cordova have done amazing work in pioneering this space; and, for years, they've taken a plunge into the deep end where browser vendors have been too scared to go. Even if Cordova can't provide the apps I asked for, their engineering and API knowhow are invaluable to the standardization work at the W3C. The massive success that is Cordova and PhoneGap,  proves how much we need this stuff on the Web.    
> We implemented geolocation before any browser did. That turned out to be a pretty ok thing! I easily spend more than half my time helping people with downstreams of Cordova. We most certainly have a very clear picture of what platforms matter, what APIs people are using, and what APIs they may need in the future. Most of our architecture has been driven by real use---not surveys or pie-in-the-sky ivory towers designed by Enterprise Architects.

I think this is ok. It may just be a pipe dream of mine to actually gather up data - or I just need to get more involved in the Cordova community (I know Moz is getting more involved, so that might make it easier). The anecdotal evidence that people from the Cordova team can provide is extremely valuable - it's not to be understated. And there is nothing wrong with it being anecdotal - particularly because we know how much time people like Brian spend both helping and speaking with actual developers: so I totally trust that if someone like Brian say "developers are all asking for X" then it's worth taking that feedback very seriously.    
>  
> That said, we can definitely do a better job of getting visibility into these things as a project. My only caution is that information is kept private to our group for the sake of burgeoning platforms and to encourage project diversity that makes teh web so resilient to proprietary traps.

Understood.   
>  
>  
> We all have a lot to contribute back to browsers and the web. Indeed is why we're all here working on Cordova today.  
We desperately need you guys! don't go anywhere! :D     

Re: Cordova/W3C

[Brian LeRoux <b...@brian.io>]

>
> A better approach in general would be to have regular surveys asking
> people to submit:
> - vendor url, vendor contact, appname, platforms, usage count by platform,
> list of plugins they use, with a score for how happy they are for each
> platform/plugin.
>
>
I agree we should be looking at surveys. That said, I do not feel surveys
are a good way of driving product design. They are a tool but not the whole
picture.




> If you don¹t build it, they certainly won¹t come.
> Let¹s try building it and see who comes.
>

Yeah, some cheap/easy/quick survey's could be useful for contextualizing
some decisions.


> PhoneGap build should be able to survey the things it

We do. And we also do extensive implicit analytics such as how many builds
a platform gets. Sharing that information would be dangerous to the goals
of the project. Oh look, many people are just building for Firefox OS.
Guess we don't need to support iOS anymore. Or something like that. ;)

> Ideally Cordova would be able to reach its consumers and talk to them.

I do. I think many here do too.


> Without having this information, Cordova really can¹t usefully sit at the
> W3 table, as it clearly has absolutely no idea about what people are doing
> and thus clearly has nothing to provide to W3.

This is leading a little severe. We implemented geolocation before any
browser did. That turned out to be a pretty ok thing! I easily spend more
than half my time helping people with downstreams of Cordova. We most
certainly have a very clear picture of what platforms matter, what APIs
people are using, and what APIs they may need in the future. Most of our
architecture has been driven by real use---not surveys or pie-in-the-sky
ivory towers designed by Enterprise Architects.

That said, we can definitely do a better job of getting visibility into
these things as a project. My only caution is that information is kept
private to our group for the sake of burgeoning platforms and to encourage
project diversity that makes teh web so resilient to proprietary traps.

We all have a lot to contribute back to browsers and the web. Indeed is why
we're all here working on Cordova today.