search

da2x / save-data

WebExtension for sending the Save-Data: on HTTP request header.
https://www.daniel.priv.no/web-extensions/save-data
GNU General Public License v3.0
21 stars 4 forks source link

Implement directly in browser #2

Open techieshark opened 4 years ago

techieshark commented 4 years ago

Hi, thanks for making this extension.

FYI - I've suggested that Firefox implement the functionality directly in the browser.

Bugzilla feature request here: https://bugzilla.mozilla.org/show_bug.cgi?id=1627407

I'd be curious if you have any insights or thoughts you'd want to share on that thread (or even just a +1 comment). Thanks!

da2x commented 4 years ago

Save-Data is part of the Client Hint specification. AFAIK, Mozilla is opposed to the whole spec.

techieshark commented 4 years ago

@da2x thanks for pointing that out! I was not aware.

It looks like they made a decision about the spec in general, knowing that they might make separate decisions for each part later:

At some point we will need to make decisions separately about whether to send DPR, Save-Data, and friends when a server requests them. That decision might be different for each hint, but right now we are looking to see if the spec is good or bad.

And while there was an indication that Save-Data was "fraught" (see below), it appears that was based on this discussion, which rightfully noted that it's difficult to know when the human will want to save data based on the network characteristics alone (e.g. does using mobile data mean you want to save it? Not if you have plenty of it and your plan is cheap).

Unfortunately, though I say that we want a position on the entire specification, the answer is bound up in the trade-offs that apply to each of the hints. For mine, DPR is harmless, whereas Save-Data remains fraught for all the reasons that the network information API was tricky.

So after reviewing these discussions I would think that the proposal to put the 'data saving mode' in the user's control (vs the browser's automatic control) is something totally different, and might be worth a new discussion about.

Thanks again for pointing this out - I'll see how the bugzilla issue goes and might raise some of these points there too.