Re: AppCache post-mortem?

I think everyone's being nice, and in the interest of not ruffling
feathers, trying to make this about vague-sounding dynamics.

That is missing some very important points:

   - Tobie, I, and several other web developers engaged repeatedly on the
   topic of AppCache.
   - We raised our concerns through the Bug Tracker[1], with support from
   Microsoft, and were repeatedly berated and told that we didn't understand
   the feature, and that the HTTP cache was "good enough". This continued
   through dozens and dozens of comments by many in-the-wild developers over a
   long period of time. I highly doubt *any* developer would be willing to
   put up with that level of abuse indefinitely. If you have not already read
   through the comments on this bug, I would encourage you to do so.
   - We also attended an in-person "Fixing App Cache" workshop sponsored by
   the W3C. We raised the same issues, and weren't able to get enough traction
   to get past the executive veto.
   - Eventually, we were able to get our specific issue fixed, but after an
   ordeal like that, no involved web developer could seriously think that a
   bigger fix (a la NavController or Jonas' AppCache) had a serious likelihood
   of being accepted. The disconnect between the web developers (and
   implementers!) presenting well-reasoned use-cases and the very strong
   pushback we got from the specification was plain wrong.

I say this as somebody who has continued to participate in standards, and
who has done work with both Alex on NavController and Jonas on his
declarative AppCache. What happened in the discussion on the bug tracker
was unacceptable, and will not contribute to an atmosphere that encourages
web developers and implementers to raise issues with the current state of
the specification.

It's fine to blame any number of things on inadequate involvement or
generic communication mishaps (and often that will be an accurate
explanation for some situation) but that is emphatically *not* what
happened here.

[1]: https://www.w3.org/Bugs/Public/show_bug.cgi?id=14702


On Thu, May 9, 2013 at 8:52 PM, Tobie Langel <tobie@w3.org> wrote:

> On Thursday, May 9, 2013 at 7:33 PM, Marcos Caceres wrote:
> > I just don't think we (WebApps participants) have the bandwidth to
> support developers.
>
> This isn't about supporting developers. This is abut making sure that the
> features that are standardized are actually the ones developers need to
> build stuff upon, AND that the APIs to access these features don't suck.
> > It would be great if the W3C could provide resources there (e.g., help
> with adding documentation for each API on the WebPlatform.org (
> http://WebPlatform.org) site).
>
> Let's face it, W3C doesn't have resources for this. Members might.
>
> While documentation is fantastic (and is best done at the same time as the
> spec is developed, as documenting an API usually informs really well on its
> inconsistencies and lack of clarity), that's only part of the picture.
>
> We need an instance that reviews specs that come out (much like we have
> the Privacy IG) and provides feedback on API design (and which can be
> consulted early in the process so last call reviews don't immediately send
> back spec authors to the writing board).
>
> --tobie
>
>


-- 
Yehuda Katz
(ph) 718.877.1325