- From: Florian Rivoal via GitHub <sysbot+gh@w3.org>
- Date: Tue, 01 Oct 2019 01:11:34 +0000
- To: public-css-archive@w3.org
> but the only way to get the # of identifying bits down to an acceptable number is to start removing as bits anywhere and everywhere we can (where "can" means "w/o breaking the web").
That's the trick though:
1) Can we eventually remove enough bits without breaking the web to reach a usefully low number? If yes, then we should start where we can. If no, it doesn't matter where we start if we already know we won't get to a useful level ever. As I understand it, @tabatkins 's claim is that he (and Google) are skeptical that we can ever reach a usefully low number, and that they would therefore like to see some attempt at proving that it is possible (or at least plausible) that we can get there at all before we start removing things.
2) What do we count as "breaking the web"? How severely do we need to inconvenience people before it counts as breaking the web? How many people need to be impacted? What if it is a small absolute number that represents a high percentage of a particular demographic?
I am in particular concerned about making the web substantially worse or unusable for people who fall into @hsivonen 's category 5 and 6.
I'll add two more categories:
8. People who pay a substantial percentage of their income on data costs, and wish to reduce the weight of web pages by installing commonly used fonts to avoid them being downloaded over and over by `@font-face`. May overlap with 5.
9. People on linux systems like debian where everything including the base system is installed via a package manager, and there's no distinction between system fonts and user installed fonts.
--
GitHub Notification of comment by frivoal
Please view or discuss this issue at https://github.com/w3c/csswg-drafts/issues/4055#issuecomment-536815201 using your GitHub account
Received on Tuesday, 1 October 2019 01:11:35 UTC