Re: Web/Native: gap analysis

On 23 Oct 2013, at 15:03, Marcos Caceres <w3c@marcosc.com> wrote:

> 
> 
> 
> On Wednesday, October 23, 2013 at 2:51 PM, Bruce Lawson wrote:
> 
>> On Mon, 14 Oct 2013 13:30:19 +0100, Bruce Lawson <bruce@brucelawson.co.uk (mailto:bruce@brucelawson.co.uk)>  
>> wrote:
>> 
>> 
>>> IMO these are the immediate priorities, so we don't spread ourselves too  
>>> thin. They also appear to have the most developer and implementer  
>>> traction behind them, so appear a good start
>>> 
>>> 1) Service Workers are being implemented in Blink and Gecko
>>> 
>>> 2) Bookmarking - aka application shortcuts - are implemented in various  
>>> browsers and OSs. Now's the time to push for standardisation
>>> 
>>> 3) Permissions are being debated at the moment.
>>> 
>>> Let's start with these.
>> 
>> replying to myself is wierd, but i've been thinking deeper. The three  
>> things above are already known problems and are already being looked at by  
>> other groups. So, while it's important that we document these as part of  
>> the list of Stuff That Makes People Write Apps Not Web, I don't there's  
>> much that this group can add to others' on-going efforts.
> 
> Sorry Bruce, but I have to disagree. I’m not saying we can do better, but last time we turned our backs on this stuff we ended up with AppCache, IDB, and a whole bunch of shitty APIs. Can’t afford to just say “oh, don’t worry! those guys will handle it” again. Fool me once, and all that - we gotta be all over that stuff *early* making sure it’s done right(tm).  

And I agree strongly with Bruce. 

It's not a secret that some APIs have abject failings. Like Marcos, I expect, I'm part of this group because I am deeply worried about the absence of effective offline mechanisms, etc. etc. etc.  However, saying an API is shitty (rather than ineffective for some particular purpose) is a privilege reserved to the TAG, afaik, since we have no working model to base an opinion of that sort on, and are not chartered to create one.

It is our job to offer opinions on why we do or don't consider an API "effective" from a Web and Mobile standpoint and to offer our opinions on what the requirements are from the standpoint we are chartered to espouse. I believe that a framework for evaluation as suggested by Dom, and as referred to by him [1], is the right way of assessing requirements and for moving forward in offering our opinion, in a structured, organised and coherent way to the groups working on these issues.

[1] http://lists.w3.org/Archives/Public/public-web-mobile/2013Oct/0072.html



>    
>> Something that I hear a lot about, but haven't seen documented in-depth,  
>> is the UX gap (as opposed to the feature gap above). Why do consumers like  
>> native apps more? (or, indeed, do they?). What is missing from the UX of  
>> web? For example, if a site has swipe gestures to go from one page to  
>> another, how do users on non-touch devices navigate? If the developer adds  
>> next and back buttons, they are superfluous (arguably) on swipe-able sites  
>> and their superfluity (arguably) detracts from UX as "it doesn't feel  
>> native".
>> 
>> Should we document the UX gap and see how that could manifest itself as  
>> CSS/ JS specs?
> 
> Yes, this was raised on separate thread also. See:
> http://lists.w3.org/Archives/Public/public-web-mobile/2013Oct/0061.html
> 
> I’ve yet to respond to that thread separately.   
> 

bis re Dom's analysis.