On Fri, Sep 1, 2017 at 3:57 PM Neil Patel Quinn <nquinn(a)wikimedia.org>
wrote:
I understand the desire to avoid playing favorites by
directing users to a
list of browsers rather than a single one, but I think that cuts against
*both
*the goals of doing this in the first place.
The first goal is to nudge users to upgrade from an insecure, less-capable
browser to a modern one. But if we present them a list of 10 alternatives
(or even 2), they're far more likely to get stuck in choice paralysis [1]
and far less likely to actually do what we want and upgrade.
Indeed. A big list of "HEY PICK ONE OF THESE" means we'll end up
fracturing
our users over a bunch of browsers that most of us would never even use
ourselves. I merely suggested Chromium alongside Firefox because it's also
free/open, even if driven by the BIG EVIL GOOGLE.
The second goal is to strengthen non-profit,
open-web-focused browser
makers by increasing their market share. As I see it, the best way to do
this is to nudge all our users towards a single, high-quality browser which
already has significant market share, rather than distributing them across
many different browsers with tiny market shares.
Indeed, like I said above. However high quality is subjective...my
experiences with Firefox have been horrible the last several years, which
is why I stick to Chromium/Chrome mostly. That's why I'd suggest like
basically 2-3 options tops so we don't play favorites :)
I'd suggest that the best areas for debate are (1)
whether these are good
goals, (2) whether their benefits justify interrupting users' browsing, and
(3) which single browser would be the best destination
Obviously, my answers are (1) yes, (2) yes, and (3) Firefox, but some will
disagree :)
(1) Eh, maybe. I care mostly because these older platforms are horribly
insecure and if we can get people on a half-decent browser on those
platforms then that's a win (cf: T118181 and all its various linked tasks).
Javascript is wayyyyy down the list of why I care here :)
(2) We already interrupt some of these users anyway per the TLS migration
stuff I mentioned in (1) above. I think the rollout there--start with small
percentages and slowly ramp up prior to there being a deadline is a good
route to go.
(3) I would *really* like to have 2--maybe 3--browsers to list. There's
zero reason to make users think there's only one option when there's a
couple of valid ones.
-Chad