Re: getting rid of callers

On Fri, Oct 29, 2010 at 1:26 PM, David Singer <singer@apple.com> wrote:
>
> On Oct 29, 2010, at 12:54 , Robert O'Callahan wrote:
>
>> On Sat, Oct 30, 2010 at 7:27 AM, Ian Hickson <ian@hixie.ch> wrote:
>>>
>>>
>>>
>>> I agree that
>>> if the content using this is insignificant, that we should drop the
>>> feature, but then why would Microsoft not want to simplify their code and
>>> drop it also?
>>
>>
>> There are many possible reasons. Perhaps the feature has significant usage
>> in IE-only arcologies. Perhaps they don't know, but they fear it might, and
>> the risk outweighs the benefits to them of removing the feature. Perhaps
>> they have adopted principles of never changing anything that aren't always
>> optimal for the Web platform. Perhaps they haven't considered it at all.
>
> I have no idea what the actual state here is, but plenty of companies have a
> policy of not removing features from shipping products unless it's really
> really necessary (e.g. a security fix), or after a *long* period of
> deprecation.

I think that is fine. I can understand if vendors want to first
release one or two versions that warn in error console about usage
before they remove a feature. I've definitely wanted this myself a few
times.

However I don't think the spec needs to be bound by that. So far no
one has shown *any* content which relies on callable collections, much
less enough such content on the public web that it would put a browser
at a disadvantage not to support it. On the contrary we have data, in
the lack of bugs filed on Firefox, showing that this feature doesn't
seem depended on by content.

I think we should remove this from the spec for now. It can always be
added back later if data turns up which shows that callable
collections are needed.

/ Jonas

Received on Friday, 29 October 2010 21:36:34 UTC