Is set basically syntactic sugar for deepcopying a struct, mutating the specified field, and then returning that deepcopy? Seems like it could be quite slow.
Yes. I’m not sure how slow it is in Julia, but pure functional languages do tend to generate more garbage for this reason. Hopefully the compiler can optimize it away in simple cases.
Edit: It’s not deepcopying the whole struct, just the parts that need to point to something new. So if you update a.b.c, it will shallow-copy a and a.b, but nothing else.
There is an underlying precarity in the academy that is so deep it almost feels like a natural part of science. This is part of what makes reform so difficult. Early on, rocking the boat feels like career suicide. Later on, if you are lucky enough to become established, you are much less likely to feel like deep reform is necessary: after all, the system benefited you. And even then, the precarity doesn't go away. You still compete for grants like everyone else, and your trainees are attempting to become established. Why should they be the ones to shoulder the risk of, e.g., ignoring the glam journals and exclusively putting out preprints?
This seems like a tautology, by definition most people are not so unique that they can’t be replaced in a large enough organization.
This was different in the past because back then academia was vastly smaller, so practically everyone from post—doc on up was a literal genius, or close to it. And thus much harder to replace.
It seems to me that something like eLife's model is the best solution to this [0]: you still have a minimal amount of curation, but generally if a paper is written well enough and within the field it won't be desk rejected. Then, it gets published on the site and sent off for peer review. Peer reviewers assess how sound the paper is and pass a judgement which readers can view, as well as provide some recommendations to the authors make it stronger. The authors can then either revise the paper, or do nothing at all. In either case, papers don't languish in reviewer hell and the larger scientific community gets to see it.
This behavior and macOS’s terrible default window management ultimately drove me back to Windows for serious work. I suppose I never fully adapted to the macOS way of doing things, but I never understood what they envisioned users were supposed to do here. Is the intended behavior that the user minimize any VSCode windows that are not currently in use?
But then what if I have multiple windows open because the other has some reference codebase? Do I keep that in a separate desktop instead? It’s just perplexing.
This is where you see the NeXTSTEP under the hood :-). A very long time ago, this made perfect sense, as applications from that era were multi-window. You can sort of see how it worked here: https://www.paullynch.org/NeXTSTEP/NeXTSTEP.screenshot.jpg (no affiliation, just the first relevant Google result). The Mailboxes window (to the right) and the email display window (in the background, just behind the File Viewer window) were separate, but they were part of the same application. That's why moving all of the application's windows to the foreground when switching to that application was the right call back then, the expectation was that you'd usually run a single instance of that application, and that application would have any number of windows. IIRC if you didn't need one of them too frequently, you would just iconify it.
Contemporary applications (especially cross-platform lowest common denominator applications, like most Electron apps) don't really do that anymore. Pretty much every modern email client has the mailboxes view in a side pane or something, whereas NeXT would have the email view, the mailbox view, address book views etc. all in separate windows. If VS Code has multiple windows, they act pretty much like fully separate instances. It's just not the kind of multi-window application that (what eventually developed into) the modern macOS UI was built for.
Edit: this has been, at various times, been retrofitted onto various contemporary design notions in terms of simplicity or intuitiveness. That's 100% ivory tower bull: this interaction model was pretty common on late 80s/early 90s interfaces, especially on Unices, and everyone gradually moved away from it precisely because it was anything but simple or intuitive, it was confusing as hell. Even as early as the early '00s it had gone out of fashion, and holdouts were just plain weird. E.g. GIMP used to have this mode (and just this mode) in its 1.x-releases and if you asked anyone why they hated it, that was their first answer, before they got to everything Photoshop did and GIMP didn't.
It is a windowing system built for one monitor and one desktop. Its metaphors broke completely once they added multi monitor multi desktop per monitor support.
Even the menu bar does not make any sense when you have two monitors. Let alone the useless command tab that picks up whichever window (or not even a window) it wants
Only Gnome does multi-desktop really well in my opinion.
Unless Windows has fixed things, I found that windows alerting on an alternative virtual desktop could lock everything until I found the popup. There were plently of other sharp edges.
Same. I got a company provided MacBook and used it as my daily driver for a few years. Really tried to buy into the ecosystem. But some things just felt so clunky, so when next laptop replacement came after ~3 years it was back to Windows.
Which I'm not also always a big fan of. But the basic interactions just make more "sense" in my head. I thought it was just familiarity in the beginning, but even after getting familiar with my OS X (at the time) I didn't become effective.
In the same vein, I upgraded to Win11, but downgraded after a month because of the new useless task bar. It couldn't "ungroup" stuff. Had to wait for over a year before that came and I upgraded again. Not being able to see all my open windows on the task bar felt exactly as using MacOS again, just stupidly unproductive, why change a winning formula, Microsoft? (My guess is their designers use Mac and don't "get" Windows..)
This does not scale well with the number of windows, e.g. going to the next pdf open might mean cycling through half your windows. Hence you get grouping to make this simpler. Whether it should be grouping by app, by virtual desktop, by time opened, "smart", manual, whatever is another discussion.
I suppose this makes sense if you have relatively few windows that each themselves encompass self-contained task state: one terminal window with a bunch of tabs, one VS Code window with a bunch of tabs, and so on.
I've always found this rather, uh, un-simple to deal with because I both suffer from attention problems and because I tend to want to consider a whole box of windows together as a complete task. macOS's per-appbundle model, inherited from NeXTstep, isn't perfect, but it does at least get closer to what I want.
I used to write software in Delphi, and Alt-Tabbing through several forms and code windows and panels and hypertext documents and compiler error listings and whatnot was just a gigantic pain. It really, really soured me on Windows' way of doing things: I could just never get the mental model down for how windows would stack in the Alt-Tab order given my need to flip between six or seven of them somewhat randomly in the course of working on something.
I've arrived at something similar after going through a lot of different solutions: Evernote, Quiver notes, Apple Notes, Logseq, Tana: now I just keep everything in one big Journal.md file in Obsidian. I added a datestamp shortcut that inserts the date as a title in "2024 February 19 (Mon)" format, and get to writing. I use subheadings sometimes if I'm writing a lot on a particular day and it gets messy, but most days it's just a hodge podge of everything, and that's fine.
It works. A big issue with computer notetaking software, I've realized, is that I was spending too much time trying to figure out where to put things: what note should this be connected to, which folder should this be in, etc. Dumping everything into a single document, under today's date, gets rid of that. The other issue this solved was that I never looked back at what I'd written previously: opening a bunch of files was too tedious to ever do unless I was explicitly looking for something. With this, I can just scroll down and see what I was doing last week, immediately.
This is one thing where a Roam / Logseq / Amplenote style scrollable list of daily note files works wonders. It's one of the biggest things Obsidian lacks. Dumping things into daily notes is a really powerful workflow.
I don’t disagree with you that life can seem bleak in the present, but people have brought children into the world under much harder conditions. There is something particular about post-industrialization that has caused this attitude shift.
I recently decided to open up Edge and see how it was, given that it is the only browser on Windows which features hardware isolation [0]. Though it seems like this feature is deprecated now.
The sheer amount of garbage I had to turn off made me give up and close it half way through. Which is a shame, because there seems to be good tech underneath all the garbage.
Post-covid, it's hard for me to imagine replaying today how, as a kid, I used to press my face into that periscope viewfinder. I remember it sometimes still smelled of cigarette smoke from the previous adult player before I popped in a quarter. Was that salt on my lips?
Awesome 3D game, amazing cabinet design. Holy smoke super gross, too.
Which is fair. Scratch marks around the cave were said to be attempts at art. The fact that the bones were in a cave made them conclude that they were "burying" their dead rather than the more plausible conclusion that they died in the cave as a result of some water or gas flooding.
Tried reading the paper and it was surprisingly hard to understand, probably because I don't know anything about the field. That said, I think the critiques make sense. Still, both the peer review and the response in the articles seem so extreme, I wonder if there is something else going on here. Is there some political or ideological divide in this field?
