W3C home > Mailing lists > Public > public-device-apis@w3.org > September 2009

Re: ISSUE-4: Versioning Straw Poll

From: <anssi.kostiainen@nokia.com>
Date: Tue, 1 Sep 2009 09:39:19 +0200
To: <robin@robineko.com>, <public-device-apis@w3.org>
Message-ID: <C6C2AA57.6533%anssi.kostiainen@nokia.com>
On 27.8.2009 15.48, "ext Robin Berjon" <robin@robineko.com> wrote:
> Here's the SP:
> 
>    This house believes that explicit version mechanisms on an API, such
>    as have been done elsewhere using for instance hasFeature(), a version
>    attribute on interface object, or a version parameter passed to a
>    constructor are not useful in a web context and should be forsaken.
>    Future revisions to given interfaces should either be strictly additive,
>    change names, or ensure that what limited deltas are made do not
>    break real-world code.

I Agree.

(IMHO object detection -- the best practice adopted among JavaScript
developers for checking feature support -- is a pragmatic and reasonably
robust mechanism. Additionally, based on a quick scan of some major
JavaScript libraries none of them relied on hasFeature() to fight against
DOM quirks [I grepped through the latest of jQuery, Dojo Toolkit, MooTools
and Prototype].)

-Anssi

BTW. What's the netiquette of this WG ML with regard to posting style? I.e.
do we prefer inline, top and/or bottom posting?
Received on Tuesday, 1 September 2009 07:37:26 UTC

This archive was generated by hypermail 2.3.1 : Monday, 23 October 2017 14:53:38 UTC