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
If only the bay area would get new RHNA numbers sooner than 2023... although I guess 2023 now seems a lot sooner than it did a few years ago, and ABAG's process has already started...
So the cycle matching day-of-year to day-of-week repeats every 400 years. For any given day of the year (other than February 29), there is either a 56/400, 57/400, or 58/400 chance that it's on a given day of the week.
6/6
However, in the *actual* 400 year cycle of the Gregorian calendar (which is close to matching the length of the mean tropical year), the number of days in the 400 year cycle *is* a multiple of seven. January 1, 2000 is a Saturday. So is January 1, 2400. And 2800, etc.
5/6
1
1
If the number of days in the 400 year pattern isn't a multiple of seven (the 6/7 chance), then matching of day-of-year to day-of-week would form a 2800 year cycle, as the 400 year cycle repeated itself on different days of the week, and probabilities above would be uniform.
4/6
1
Given an arbitrary pattern of year lengths that form a 400 year cycle, there would be a one out of seven chance that the number of days in that 400 year cycle is a multiple of seven.
3/6
1
The Gregorian calendar uses years that are 365 days, but 366 days in years that are multiples of 4, except 365 in years that are multiples of 100, except 366 in year that are multiples of 400. See en.wikipedia.org/wiki/Gregor…
So the pattern of year lengths is a 400 year cycle.
2/6
1
1
I guess it's time to explain this somewhat cryptic poll I posted a few weeks ago.
The correct answer is 14.25% (exactly), or 57/400.
For all days of the week, the answers would be:
58/400 Sun
56/400 Mon
58/400 Tue
57/400 Wed
57/400 Thu
58/400 Fri
56/400 Sat
Why?
1/6
1
3
I think summer time changes are a good idea if you're between about 20 degrees and 45 degrees from the equator. (Closer to the equator and there's not enough seasonal change; further and the seasonal change is so big that summer time's bonus is minor.)
My favorite is when 2038 stuff regresses and nobody cares: bugs.launchpad.net/bugs/1773…
1
1
The non-live version (i.e., take a picture, select the text you want) is often much more reliable than the live version. It also, if you have some understanding of the writing system, lets you correct the characters it got wrong.
2
If Klobuchar wants to find voters who voted for Obama and Trump, there are likely more in a town that went from 70% Obama to 55% Clinton or a town that went from 45% Obama to 30% Clinton than in a town that went from 52% Obama to 48% Clinton.
4/4
1
2
It seems like that exception confuses either candidates or reporters (maybe both) into thinking about elections that way -- in terms of winning pieces of geography rather than getting the most votes.
3/4
1
1
1
In elections you generally want to maximize the number of votes, not the number of towns/cities/regions that you win. The big exception, of course, is states in the US presidential election -- a broken system that means most voters' votes don't matter.
2/4
1
1
1
This seems like either a weird strategy, or an incorrect presentation of a strategy.
To find voters who voted for Obama and Trump, you should look at the difference in votes (or vote %) between Obama and Clinton.
Looking at which towns were won or lost doesn't make sense.
1/4
Amy Klobuchar is focusing on towns that voted for Obama before flipping to Trump in 2016 politico.com/news/2020/01/02…
1
1
1
WebRender is new and scary, so we're rolling it out gradually. Much better than shipping buggy stuff to everybody on the release channel.
If you know how to use about:config you can probably try it out, though, and it may well work.
1
1
I guess when it was originally discussed we got into such a big argument about pseudo-element versus pseudo-class that we never really discussed the styling:
lists.w3.org/Archives/Public…
lists.w3.org/Archives/Public… (twice!)
lists.w3.org/Archives/Public…
1
1
3
The one update I did get was Fenix -- but that's probably because I have Fennec beta installed versus Fenix nightly.
1