Engineer on @googlechrome. Involved in CSS and W3C standards. Previously @mozilla, @w3ctag. Mastodon: @dbaron@w3c.social
Rockville, Maryland, USA
Joined March 2008
- Tweets 7,521
- Following 1,244
- Followers 4,296
- Likes 18,519
This sounds like an excellent use of the data Google has. It prevents real people from getting sick. Consent shouldn't be needed for this. The only harm I see is encouraging us to like that one company has all this information.
you know:
- google knows where you are
- google knows what you search
you probably don't know: when you search for food poisoning symptoms, google looks back in your location history the time that food poisoning takes to incubate and guesses which restaurant poisoned you
1
1
This is quite the map (courtesy @NHC_Atlantic) for still being in July. It's well before the peak of the Atlantic hurricane season, already up to H, with one Cape Verde tropical storm now and maybe another one coming.
Well, given the lack of other responses, I may as well point out that you were correct anyway.
Mandela I saw intentionally, though it was an accident that I saw him waving from his car from only 20 feet away as he was leaving the venue. (Was much further during the speech.)
1
1. Bill Clinton
2. Al Gore
3. Nelson Mandela
4. Richard Nixon
5. Barack Obama
(I'm being a bit generous with "a few feet"; I'm including "in the same auditorium as" and "saw their motorcade go past from the sidewalk". But that makes it a little more interesting.)
List 5 famous people you've either met or have been within a few feet of, but ONE is a lie. Then let your friends guess which one they think is a lie.
1. Mickey Rourke
2. Al Pacino
3. Bobby Orr
4. James Earl Jones
5. Charlie Sheen
Who is my lie?
1
Sounds a lot like what happened to the Russian consulates in San Francisco (nytimes.com/2017/08/31/us/po…) and Seattle (nytimes.com/2018/03/29/world…) a few years ago.
1
2
Also, I'd note that for Web NFC in Chrome, I don't think the permissions are scoped to devices. I think it's just permission to use NFC or not.
1
I agree executables are often dangerous.
However, as long as the Web is a multi-vendor multi-OS platform, there will be things you can do in native apps on at least some platforms that the Web can't do. So some level of unusual or new things will always need native code/apps.
2
Yubikeys happen to be well-known devices; given that there are lots of less-well-known devices out there, it seems likely they're not the only ones with that sort of problem.
(I think that issue was first raised in github.com/mozilla/standards… and led to devices being blocklisted.)
1
One concern from Mozilla folks is that it exposes to the Web things that weren't designed with the knowledge that they'd be exposed to arbitrary Web content. (There should be straightforward ways to address this.)
This isn't theoretical; see blocklisting of Yubikeys in WebNFC.
2
4
Seems pretty easy in Firefox, where the default for camera/mic permissions is not to remember the permission (though it's easy if you want to).
1
8
Old code (before tokio/futures update):
github.com/dbaron/wgmeeting-…
Old code (after tokio/futures update):
github.com/dbaron/wgmeeting-…
New code:
github.com/dbaron/wgmeeting-…
Much nicer.
(I admit I don't write a whole lot of rust, so I'm sure my code is non-idiomatic in various ways.)
5/5
3
That said, I'd note that there was also one unfair advantage with the new way, which is that one of my dependent libraries, crates.io/crates/irc , has restructured to get along a better with the new tokio/futures infrastructure than it did with the old.
4/5
1
It feels like a big improvement -- and probably a bigger difference than the introduction of async/await in Javascript.
3/5
1
1
Using futures and streams with chained callback functions (the old way) instead of async/await (the new way) led to lots of struggling over ownership and lifetimes (and fighting with the borrow checker). In the same code restructured to use async/await, it's much easier.
2/5
1
1
The past two days I've had a chance to use async/await in Rust for the first time. It was a useful comparison against the old way since I was refactoring existing code to use it.
It made things *much* easier and simpler.
1/5
1
6
Seems like a big difference between doing this to the leadership versus doing it to about a quarter of the population of the country.
It also seems unlikely that the rationale for this order is genocide in Xinjiang *or* that it will be perceived in China as the rationale.
Seems like it would help if county health departments were publishing summary statistics on how they thought people were actually being infected with COVID-19.
If they can't do so (or think they can't) because of medical privacy laws... perhaps those laws are excessive?
All the “beach shaming” & “park shaming” is likely going to backfire, isn’t at all backed by health research or evidence on how virus spreads. There is continuing huge risk from the virus, and we are backtracking, we have to be careful—it isn’t bc families are going to parks.
1
1
4