Why would you stop a selective recount aimed at finding more votes in a specific set of counties that tended to lean toward one candidate? Hmm...

The above is a one-sided interpretation, but it's not an unreasonable one. The situation was that Florida had voted with punchcard ballots, and there were problems with a percentage of the ballots, both "overvotes", cases where more than one candidate was punched, and "undervotes", cases where no candidate was punched in a race. But it turns out that when you're talked about punched cards, "punched" is an ambiguous term. There were cases where chads (the little rectangular pieces of paper that get punched out) were left barely attached ("hanging chads") and cases where they were partially punched but still fully attached ("dimpled chads"). And everywhere in between.

Gore's team called for the targeted recount, arguing the theory that the targeted counties had more elderly populations who were more likely to have failed to fully punch their choices, resulting in "undervotes" when the scanning machines read them, i.e. the machine saw no holes for a race and decided that the ballot didn't vote for anyone. They argued that human examination of those ballots could clearly see in some cases that a position was punched and that a recount should be done to count those. The Florida State Supreme Court ordered a selective recount of the counties the Gore team thought should be recounted, to correct correctable undervotes.

Bush's team argued to SCOTUS that the selected counties were all Democrat-leaning (they were) which meant that the voters in the Republican-leaning counties that had not been selected were not receiving equal treatment under the law. The initial statewide count was done by machine, but the recount which would be more "permissive", finding votes where the machine wouldn't, was only being done in counties where the newly-found votes were more likely to be for Gore.

SCOTUS stayed the recount while they decided how to handle this, then ruled a few days later that the recount was unequal treatment, that a proper recount would need to be statewide and it would need clear rules for how to count ambiguous ballots, rather than each county defining its own rules. But by then the clock had run out anyway. The court's rationale for staying the count while they decided was that if the count proceeded counties would release updates that would likely show the vote totals shifting, which would make rejecting the recount look really corrupt if it went against Bush and if they decided that's what should be done. "Oh, the conservative court let the count go forward until they saw it wasn't going their way, then they rejected it" was worse than "The conservative court stopped the recount before it produced any results", was the theory.

You can certainly argue that these published positions by the court weren't the real reason, but they're not without merit.

In any event, as a careful, methodical, independent recount determined months later, the Gore-requested recount would still have shown Bush won. An incomplete recount which resulted in no win would have been a Bush win, because the state legislature was voting to send Bush votes for that case. An incomplete recount in which Florida failed to submit a slate would have been a Bush win, because the US House would have picked him. The only scenario in which Gore won was a statewide recount that also tried to tally "overvotes" -- having humans try to discern which of multiple punched chads was "most" punched, but no one at the time thought that would favor Gore, and it wasn't even being discussed.

Aside: "Corruption" is the wrong word. Corruption specifically refers to bribery and other compensation-related schemes (not necessarily monetary). "Partisanship" is a better word.

Indeed it does.

Yep. So all of the F-Droid-distributed apps will be associated with one account. Or maybe it'll be distributed across a handful of accounts.

For open source apps absolutely anyone can package and submit an app under their account.

You think Google is going to verify that an app was submitted by the author/developer rather than some F-Droid guy who submits all of the F-Droid apps?

They just require the code to have been submitted by a registered identity. They aren't going to check copyright ownership... and with open source apps that's rarely only one person anyway. The only risk is that if some of the F-Droid apps turn out to be malware, Google may revoke the permission of the person who submitted them to submit apps. And note that they don't insta-revoke. If it's a legitimate mistake (e.g. someone slipped some bad code in upstream), and it doesn't happen too often, it's fine. But whoever does the F-Droid submissions will want to take some care with what they submit.

That, BTW, is the actual reason for the registration requirement: Being able to block malware authors, at least to the extent of requiring them to find or create some government ID to create a new account. The way many malware authors operate, the $25 fee may actually also pose an obstacle for them.

Similar requirements on the Play store did wonders for reducing malware volume to a level where Google could stay ahead of it. Now they want to extend the same protection to the entire Android ecosystem. If you're curious how I know this: From 2014 to 2025 I was a senior member of the Android security team at Google. While I never worked on anti-malware efforts, I know the senior engineers who do and I chatted with them about stuff. The security engineers have been pushing for this change for years but it has been blocked by management because it was expected that it would generate exactly the sort of mis-perception that you have. Eventually, the engineers were able to prove their case with sufficient data that management let it happen (not without some pushback from the PR team, I expect).

It really, truly has nothing to do with killing F-Droid. No one in Google has any reason to want to kill F-Droid, and more than a few use it personally. NewPipe is a different story, though even there the Android team doesn't particularly care about it, except to the extent that the YouTube team can convince them to care.

If you're a fact-consumer, you should not say "Bush stole the election".

The best evidence is that the recount that SCOTUS stopped would not have changed the outcome. A different recount that wasn't being done, wouldn't have been done and probably couldn't have been done quickly enough, even if someone had asked for it, which no one did, might have changed the outcome.

For Gore to win (assuming the recount reached the same results as the NORC/media recount), all the following would have had to happen:

1. SCOTUS would have had to remand rather than stay (on Dec 9)
2. FLSC would have had to acted immediately to (a) change the recount to be statewide and (b) direct something very like the maximal standard used by the NORC/media (every other standard found a Bush win). There's no hint that anyone would have asked for either of those things.
3. FLSC would have had to rule that the recount didn't have to be completed by the legislatively-defined Dec 12th "Safe Harbor" date, which most interpreted as the hard deadline. I don't think anyone knows what the probability of that was, but FLSC's previous rulings seemed to imply they wouldn't have.
4. The very careful statewide recount would have to have been completed (including legal arguments and challenges) in at most 8 days. Realistically more like 5, since there absolutely would have been more time-eating litigation.

The most likely outcome if SCOTUS had done nothing is that the recount, if it completed fast enough, would have confirmed Bush's win.

It's also worth noting that another plausible outcome -- Florida just can't make a decision by Dec 18th and sends no electors -- would also have been a Bush win. Without Florida's 25 votes, Gore would have had the most electoral college votes, but wouldn't have had the constitutionally-required 270. In that case, the US House of Representatives would have picked. The result would almost certainly have been a purely party-line vote, 223-211 for Bush.

One final comment: 2000 did not represent some sort of "failure of democracy", and wouldn't even if SCOTUS' intervention actually had changed the outcome. When the electorate is very closely divided, the outcome is determined by random events. Chaos. If a butterfly in Peoria had flapped its wings a couple of months before it could have gone the other way.

Yep. And this just means that F-Droid has to have one person sign up and submit all of the apps for signature, as I said.

Nonsense. There's no reason to expect that mandatory developer certificates will kill F-Droid, at all. F-Droid will need one guy to pay the $25 fee and identify himself. Unless they can use the open source developer exception that Google has talked about (but hasn't announced any details, AFAIK).

That's a bigger issue, because Google's announced policy is to require that apps respect intellectual property, which would include not distributing apps that blatantly violate terms of service. Most likely F-Droid will have to stop distributing NewPipe if they want to be in Google Play. If dropping NewPipe is enough to kill F-Droid, then I guess that'll do it.

Those in the north will complain because the portion of the winter they go to work in darkness will increase, and it will be darker. Those in the center will also complain louder because they'll start going to work/school in the dark.

Those in the south will be confused about why everyone else is complaining, but they'll lose.

It's really not. 20 years ago, yes, but they've grown up and wised up. I know lots of excellent engineers at Microsoft, and I know they do good work, and they report that their colleagues do, too.

And note that I'm no MS fanboy. I hated them with a purple passion in the late 80s and early 90s, and swore off Windows entirely in 2001. I did finally break down and buy a Windows laptop a couple of years ago because I bought a CNC milling machine and the good software is Windows only, but that's the only thing I use it for.

More precisely, "Everyone does".

Microsoft's code isn't especially shoddy, all large codebases that are getting AI review right now are experiencing the same thing.

That would be funny... but also foolish. Iran is thrilled to have Trump right where he is, doing all the dumb things he's doing.

Valid! (Including the gray and the beard, though my hair is still mostly black. Mostly.)

The point, though is that there are lots of grandmas with varying levels of computer knowledge, many of whom can and should be moved to a Linux system that is harder to screw up. Or, honestly even better, a good Chromebook.

On that topic, I did that for my father-in-law around 2010. I got tired of cleaning up the mess he made on his Windows laptop, so (with his permission) I upgraded him to Debian. It was hugely lower-maintenance for me.

The really funny thing about that particular case is that my father-in-law was a retired Full Professor of Computer Science! His problem wasn't that he didn't have the background or ability to manage a system himself, it was that he didn't want to. At his life-stage, the computer was a tool that he used, mostly as a web browser to buy parts for the farming and antique furniture refinishing that were his passions. Also, he had very thick fingers and constantly fat-fingered stuff, literally. What he really needed was a Chromebook with a full-sized laptop keyboard (a bigger-than-normal keyboard would have been even better), but Chromebooks didn't exist yet.

I'm a grandpa and I'm using a computer. My wife is a grandma and she's using a computer, too.

So, $22.50 per day, or $113 per week. How much does one of your people cost, all-in, including benefits? It's unlikely that it's less than $100k per year, and very likely at least double or triple that, if not five times or more. At $200k/year for 50 weeks, that's $4k/week. At the current token price, the AI makes a $200k engineer 3% more productive, the company is breaking even. If it makes them 1.5X or 2X as productive it's a clear and unquestionable win, even with higher token prices.

As for me, think AI makes me about 5X more productive than I would be without it, and that's just considering volume of work. Honestly, I think the quality is a little higher than I'd do myself -- not because the AI writes better code than I do (it definitely does not), but because I'm able to be pickier and do more and larger refactors than I would if I had to do the grunt work myself. Also because I have Claude write documentation that, frankly, I just wouldn't get around to if it were me. The documentation is not nearly as good as if you gave me a dedicated technical writer... but the $2500 I spend per month in tokens would come nowhere close to paying a writer, even if we ignored the coding productivity.