More people should have read Eisenhower's farewell address. Not just the "military-industrial complex" part, but also the _other_ big problem he was concerned with, which got ignored (in both the facets he mentions!), with predictable consequences: archives.gov/milestone-docum…
Well, I was wondering how long it would take @nytimes to do something with #wordle. As of a few days ago, powerlanguage.co.uk/wordle/ redirects me to this:
. @github @githubstatus
Links like github.com/project-chip/conn… used to work, but don't anymore: doesn't scroll to the right comment, and does not un-hide it. That makes dealing with resolved PR comments a huge pain (e.g. links in e-mails don't work). Broken for almost a month now
I wish people writing their code via copy/paste/search/replace would at least read the result enough to realize their comments (1) make no sense and (2) tell me exactly where they copy/pasted the code from...
I wish I could select a bunch of github action workflow runs and cancel them all at once instead of having to cancel one at a time from the little menu on each one.....
For some people the first step in resolving an interpersonal conflict seems to be to post someone else's name, picture, and contact information on social media.
It's hard for me to treat that as just a "faux pas" or "bad manners". It's just malice: trying to whistle up a mob.
Not impressed with @aclu actions if nytimes.com/2021/02/24/us/sm… is presenting things truthfully. It's one thing to be wrong; that can happen to anyone, especially in the short term. It's another to refuse to admit it when you're wrong.
Trying to reconcile the media coverage from last week. Everyone agrees the Secret Service pretty immediately evacuated Pence. A few places say Pence and Harris. Some say Pence and Pelosi. Some say those two plus Grassley. Anyone know a good source for what really happened?
I thought Puerto Rico finally had a referendum in 2020 with a non-loaded statehood question on the ballot. Then I read senado.pr.gov/Legislations/p… page 31 and discovered that it was loaded in the usual way by officially defining "no" to mean "sovereignty", not "status quo". Ah, well
W3C working group resolution for an issue about a spec not defining behavior "We have a test for this; we're not going to bother putting an actual definition of the behavior in the spec" (my paraphrase). github.com/w3c/webauthn/issu…
@w3ctag
@EmpowerToday Sometime between July and now, downloading a quarterly transaction register as QFX broke on your website in Firefox. For some reason it's showing dates like "7/1/-2020" in the UI, and passing them to the back end, which does not understand them.
@EmpowerToday I tested the current-in-July Firefox version, and it's broken there too, so this looks like a website change. Tried looking for where that string might be coming from, but it's from somewhere in the guts of React, so debugging the minified stuff is a bit hard.
popehat.substack.com/p/no-lo… is a must-read if only to counter at least some of the misinformation that we are all being flooded with all the time...
Another dumb thing that makes -Wconversion hard: `(x >> y)` has type `int` if `x` has type `uint8_t`. Why??? I can see this for left-shifting, but right-shifting really should not increase the size, ideally.
If the `ifr_flags` member of `struct ifreq` is a bitmask, why exactly is it `short` instead of `unsigned short`?
And in general, why do system headers seem to go out of their way to make compiling with "-Wconversion,-Werror" so difficult?