something98

Running Zotero (directly) is the big requirement for my next e-reader. Yeah, I can sync pdfs from my library with webdav or something else. But I want the full Zotero interface.

This is a very misleading statement; most of those physicians are using LLMs to transcribe notes from visits and/or for billing purposes (e.g., proper billing codes).

Don’t spread lies. Several states also allow same day registration. And all of them set the ballot aside until the registration is verified. This means it isn’t even counted until verified.

And, no, I’m not going to “trust me, I’ve seen it before. [I just have no evidence.”

I think your historic trend data is out of date. The inflation-adjusted median wealth crossed the 2007 bar sometime between 2019 and 2022.

See the "CPI-Adjusted to 2022" column https://dqydj.com/net-worth-by-year/

Yes, you are the only one that thinks that.

A terse diff like that, on a separate page, is probably not what they were referring to.

Gotcha. Thank you.

FYI, for anyone else that stumbles upon this: I decided to do a quick check on PyTorch (the most problem-prone dependency I've had), and noticed that they recommending specifically no longer using conda—and have since last November.

Can someone explain a non-project based workflow/configuration for uv? I get creating a bespoke folder, repo, and uv venv for certain long-lived projects (like creating different apps?).

But most of my work, since I adopted conda 7ish years ago, involves using the same ML environment across any number of folders or even throw-away notebooks on the desktop, for instance. I’ll create the environment and sometimes add new packages, but rarely update it, unless I feel like a spring cleaning. And I like knowing that I have the same environment across all my machines, so I don’t have to think about if I’m running the same script or notebook on a different machine today.

The idea of a new environment for each of my related “projects” just doesn’t make sense to me. But, I’m open to learning a new workflow.

Addition: I don’t run other’s code, like pretrained models built with specific package requirements.

This is the correct (2025) paper.

https://arxiv.org/abs/2501.02305

Look you probably think you 'won' because I mis-spoke about a majority of democrats, but these numbers are .. not great. 45% of the country, 11% of democrats? The results have stayed stable across time. Guess what, they get to vote too? You have to convince them. One easy way to do that is to share their incredulity that 1000s of ballots come in 100% for Biden in the middle of the night.

You know, you can approach this like a scientific hypothesis testing, or you can approach it the way everyday voters do.

That's the problem, there is no proof that will convince them. And I don't mean that derogatorily. In generally, one can't be convinced by facts when they arrived at a conclusion by feelings. Especially, when so much astroturfing that was done about "I'm not saying there's fraud, but a lot of people are saying it.". If you have people in high places, that know better (because we have them on record saying they know 2020 was legit), spreading fear for the sake of it, all it does is create a false narrative of some overwhelming consensus, which then just feeds on itself. "See, look at all of these tweets, posts, articles, mentioning other people mentioning that they have a bad feeling about 2020."

There was no evidence that anything was untoward in the 2020 election—and people have had 4 years at this point to present evidence that there was. The problem with your example is that people looked at something that has occurred for almost every election in the past 100 years, that precinct by precinct votes often come in over time and in groups (you cannot be surprised that a precinct, as in a specific small area, is more homogenous than not). You can find pictures of chalkboards and primitive displays of election-night results coming in even from the early 20th century (probably before this too). But people took these static, after the fact, incremental updates to the running total as some sort of horse race. Like the NYTimes was tracking a basketball game. It's not, to pick a computer metaphor, it's like showing a progress bar for a count of all of the items in a database. The computer literally cannot count all of the "red" or "blue" items in one atomic operation; it has add them up, incrementally.

This has been proposed elsewhere, but in addition to actual speeding up the counting in certain states (and removing the barriers to doing counting ahead of of time(looking at you, PA)), is the idea of making it illegal to post results until some representative threshold of the results are in. As in, once you can be statistically certain that a different outcome isn't possible (maybe 99.99% or something), then you can post your first update of the results. Obviously, no one would accept that we should wait until all of the results are in, because oversees/military/absentee ballots might take days to arrive (even if they were mailed the day before election day). And, one obvious solution is maybe to require, at a national level, that all ballots must be received by COB election day, so that they can be tabulated in a timely manner.

Again, I fully understand that most people operate off of vibes and feels. Even highly educated people, outside of their domain (and maybe even inside their domain, if they have a vested interest in not being proved wrong), will default to vibes and feels. But you literally CANNOT prove a negative. So, just like it's impossible to prove no aliens have visited us, no one can prove that the 2020 election wasn't "stolen". Which is why we require evidence to rebut claims (extraordinary or not). No one has provided proof that aliens have visited us. And no one has provided proof that the 2020 election was "stolen"—as in the outcome would've been different. Yup, there's always some shitheads, ne'er-do-wells, and honestly-mistaken-about-their-elibilibity-to-vote-people that are caught each year. (Sadly, especially for the GOP, the number of GOP-voting voters that fall into the category outnumber the Dem-voting voters.) But, not once, at least that I've been able to find record of, in the past 40+ years, has identified fraud been able to change the result of any election from state-representative on up.

I also like to keep my anti-tiger rock on me at all times. I don't really care that there's no evidence that it works. All I know is what I see, and I haven't seen any tigers.

I don't think HNs database has the free space to contain such a list.

I haven't heard any mention of the dossier in years, other than as an artifact of the past. A quick search, and I can't find sources trying to claim its truth (or evidence of smoke, for which there might be a fire) in years.

I certainly didn't mention Rogan—I'm aware of his existence, but I've actually never heard him speak nor seen any transcripts of anything he's said. But trying to minimize the flood of absolute obvious shit that comes from right-wing outlets by choosing to point to Rogan specifically is a bit telling.

Anyone and everyone should be called out for lies they manufacture or spread. This includes lies on the left, lest you think I'm granting one side a pass.

Now you're making two claims, neither of which you're providing evidence for.

If "everyone agrees...the election had fraud", I'm sure you can provide multiple reputable polls showing this sentiment of "everyone". (I'll be generous and lower the bar to just a majority of Americans, but I'm not going to accept polls that show -only- a majority of Republicans, since your claim is "across the political spectrum")

Second. Even if, as you've admitted elsewhere, every single court case was lost (or denied due to lack of standing) in 2020, you do realize that evidence can be presented outside courts, right? Where is this evidence? They've had 4 years to collect it. If it's widespread and, as you said elsewhere is a statistical anomaly, then it, almost by definition, should be obvious to spot. Hand waving to vibes and feels and "sure seems obvious to me" doesn't mean jack.

Now, I'll grant you, that vibes and feels certainly mean a lot to the animal natures of all of us. But feels and vibes are not proof of anything.

Don't be coy, please enlighten us as to what this conspiracy is involving Russia that you think the MSM peddled, and what evidence you have that disproves the narrative.

You'll need to provide credible sources for that wild claim.

there is an adaptive insulin resistance that preserves glucose for the [specific] brain and heart cells that can’t survive without it.

To be clear, most brain and heart cells do not require glucose to function. Although, a small subset do.

I'd love to read the critiques (if any) the professor dared to give when grading that.

Research careers. Lol. Since this is HN I’ll hit a nerve, but I bet much less than 0.1% of CS MS degree holders do research as a career.

Having done that, I can answer. I'll be a little vague on the topic to keep some anonymity.

Without a plan, per se, I nurtured a burgeoning personal research interest that was well outside my field. The initial hobby was very rudimentary but it gave me enough feedback and fuel that over a number of years I built more advanced software and hardware to investigate it more scientifically. Eventually, having reached the limits of the data I had access to, I reached out to a local university to start a collaboration.

If you already have an idea, I think, as long as you're sufficiently interested in actually doing research, you'll be able to find some professor that is as well. On the flip side, I know for a fact that most professors have more research ideas than students, so they can certainly provide you with things to investigate.

As a bonus, for the professor, since this is a side gig for you, there isn't an cost for them (or much less of one). Access to facilities and resources depend on the university. But I was certainly surprised by what I was given access to in the service of doing the research.

Sometimes I look at my macos coworkers ... and I really wonder how they get work done

And I read the sentence before and wonder the same:

once you've spent a year tailoring your setup

I try not to let the intrusive thoughts win, but this is exactly what I think almost every time there's an article on the supposed supriority of emacs or markdown (i.e., plain text).

As someone who has degrees in both, I can firmly say this is wildly incorrect.

soapbox The area/function mapping is -far- from nearly universal. Not only can the areas be flipped left/right but they are almost guaranteed to be several cm in some other direction. The maps as we know them are just the ven diagram overlaps for dozens (rarely hundreds) of subjects.

good

I don’t see (from my iPad) how we would coordinate beyond the overlapping times. Specifically, how do we contact each other once a time is agreed upon?

I'm pointing out that everyone is comfortable with most files not being human-readable. And that it's a red herring to suggest that text is the longest-lived format. I'm sure my data could outlive text files if I "stored" it on granite blocks, but it's not very useful in that format. Same with plain text; it just isn't a very useful format.

I found this years ago: http://leoeditor.com/

Not universally true. But, regardless, most files are not text.

This argument always sounds superficial to me.

1) You can have corruption in files of any type. Text is no more robust to random bit flips. Pick a file format and there are often tools that can isolate the corruption to a small portion to recover everything else.

2) The solution to lock-in is not "let's make everything text", but "let's have open file formats". Many types of data, especially for project-management-like tools, aren't particularly well suited as plain text (or a markdown format). So many extra unnecessary parsing/regex steps have to be jumped through for the program to use it, and the sanitizing functions now have to be moved from code to the user's brain to not break things.

By all means, support open file formats and exporting functionality. But shoehorning -everything- into text sounds like some crazy mix of bikeshedding and worldbuilding. Why waste your time? Especially knowing MD will fall out of favor eventually, and a new format du jour will rise. What will you do then?