Re: additional download on screen shrinking

Thank you, Yoav, that's fantastic.

On Thursday, 23 October 2014, Dustin Hoffmann <dustintheweb@gmail.com>
wrote:

>
>
>
> On Thu Oct 23 2014 at 5:02:52 PM Yoav Weiss <yoav@yoav.ws
> <javascript:_e(%7B%7D,'cvml','yoav@yoav.ws');>> wrote:
>
>> OK, since you guys ask so nicely: https://codereview.
>> chromium.org/674923004/
>>
>> On Thu, Oct 23, 2014 at 11:06 PM, Dustin Hoffmann <dustintheweb@gmail.com
>> <javascript:_e(%7B%7D,'cvml','dustintheweb@gmail.com');>> wrote:
>>
>>> I've been curious about this for a long time. John, thanks for asking -
>>> and Yoav, thank you for your response.
>>>
>>>
>>> On Thu Oct 23 2014 at 3:56:23 PM Yoav Weiss <yoav@yoav.ws
>>> <javascript:_e(%7B%7D,'cvml','yoav@yoav.ws');>> wrote:
>>>
>>>> Yeah, that behavior is (intentionally) not specced, and left to the
>>>> browser to do as it pleases.
>>>> Current Blink behavior where smaller resources are downloaded when
>>>> larger ones are already in the cache is a bug
>>>> <https://code.google.com/p/chromium/issues/detail?id=425701> and
>>>> should be fixed. Unfortunately, I doubt I can get to that before Chrome 40
>>>> branches, so I don't think it'd be released before Chrome 41.
>>>>
>>>>
>>>> On Thu, Oct 23, 2014 at 10:18 PM, John Holt Ripley <
>>>> john.holtripley@googlemail.com
>>>> <javascript:_e(%7B%7D,'cvml','john.holtripley@googlemail.com');>>
>>>> wrote:
>>>>
>>>>> Hi all,
>>>>>
>>>>> I'm working with non-art directed images in a responsive layout, and
>>>>> noticing that when shrinking the viewport, both Chrome and Opera are then
>>>>> additionally requesting the smaller image instead of shrinking the already
>>>>> downloaded asset.
>>>>>
>>>>> Is this the intended behaviour? I couldn't see anything in the spec
>>>>> that determines this behaviour, so is it up to browser vendors to determine
>>>>> how to handle this? (Or the browser itself to look at network speed and
>>>>> make a decision from there?)
>>>>>
>>>>> I can see the need to request the new asset in art-directed cases
>>>>> (within the Picture element for example), but in a simple srcset and sizes
>>>>> situation is this additional download beneficial? Is the performance
>>>>> implication of rescaling a large image offset by the network request?
>>>>>
>>>>> Thanks
>>>>>
>>>>
>>>>
>>