Back in the mid to late 90s tracker stuff was what I considered "how techno was made" (I got an A grade for a presentation in "music" subject at grammar school, mostly because tracker terminology also included "patterns", samples, bpm and such (Future Crew Demos such as Panic and 2nd Reality were my baseline back then).
Boy was it a revelation when I got my hands at a "real" 909, 303+MPC+MIDI setup..
I mean, was it? I always felt like trackers were crappy and “real” producers used real synths and drum machines, but in hindsight we now know that some of the most creative/interesting producers were trackers
It did always feel like the "real" folk had some mysterious room full of hardware
I guess it was true, but so neat that they were sometimes triggering that stuff with trackers or sampling it for use within trackers (or similarly niche software)
Not sure if that dove evolution parody video got me introduced to the flashbulb’s works initially but I sure enjoyed Kirlian Selections/Tape and Our Simulacra and for some reason came across Benn on numerous occasions (mostly HN related) in the past decade (didn’t know about the YT channel until now though)
Always considered the idea/realisation behind Motion Twin (a french game dev studio set up as "cooperative") quite interesting: https://motiontwin.com/faq
Disclaimer: didn’t read the paper and only have a very basic understanding of chemistry (and sorry for simplifying a complex topic), but started wondering whether we should be able to discover useful reactions via pure calculation by now (ignoring quantum effects)?
Despite being a specific domain, wouldn’t reasoning about reactions (and their efficiency in an industrial, large-scale setting) be something that "models" should be able to do quite "easily" (given the fact that bonding forces, energy requirements and catalytic effects are sort of well-known, i.e. just some more dimensions to deal with)?
The only kind of bonds we can model reasonably well while ignoring quantum effects are hydrogen bonds. Hydrogen bonds can be modeled as parameterized effects in Newtonian mechanics, and molecular dynamics simulations using Newtonian mechanics can reproduce some interesting properties of molecules in solution.
The vast majority of chemical reactions involve changes in electronic configuration that are only described by quantum mechanics. The most accurate algorithmic approximations of electron quantum mechanics have terrible scaling properties - O(N^7) or worse. Due to that terrible scaling, the largest high-accuracy calculations that are tractable now are not that much larger than those that could be completed in the 1990s, despite much greater processing power.
There are other ways of approximating quantum effects in chemistry that scale better, but they all have tradeoffs and weaknesses of their own. They can used in limited domains or used to guide experimental design, but they're not accurate enough to discover useful reactions via pure calculation. They need to work in tandem with experimental validation and it requires domain experts in both the experimental and theoretical work. The methods are not simple enough for a bench chemist to use them as a black box for reaction discovery.
Yes, there are many research groups working on this at the moment. We can (roughly) screen through chemical reactions performed in vaccuum. Even this is difficult since you either need to simulate atoms bouncing around until you observe a reaction (extremely slow to perform) [0], or you need to numerically search for a viable reaction pathway (still quite slow to perform) [1]. The main problem is that the best methods scale badly with the number of atoms you're simulating, so you need to trade off accuracy for speed by using less-accurate methods.
Screening through reactions in the real world is particularly hard, since you not only need to worry about the inaccuracy of your simulation method, but you also need to take solvent/environment effects into account. You need to trade off even more accuracy for speed if you want to do so. As computing power advances, there will be less pressure to make these tradeoffs, but a lot of work in comp chem at the moment is focused on either exploring or expanding the speed-accuracy frontier.
Broadly speaking, "computational materials science" has been a thing for quite a while, and has been responsible for the majority of academic supercomputer time allocations for decades.
Not a chemist, but I think trying to do chemistry without modeling quantum effects is like trying to do physics without math. The latter is necessary to express the former.
Yeah, reading the sibling comments made me aware that my mental model when it comes to chemistry (and the role QM plays even in "simple" scenarios) needs an update.
Sporth is a stack-based language I wrote a few years ago. Stack-based languages are a great way to build up sound structures. I highly recommend trying it.
Chorth may need some fixes before it can run again. I haven't looked at it in a while, but I had a lot of fun using when I was in SLOrk.
There are a couple of comments on how search engine x dropped feature y on date z (i time periods after competitor j did).
Can anyone recommend a site that tracks those changes?
I find myself getting annoyed by serps of a given vendor and I might even be adopting to changes but ultimately jumping ship towards the next best thing if I can’t figure out a very low-effort way of influencing the results in my favor using flags that suddenly stop working - all I notice is a significant degradation of result quality.
Tldr: create a new visionOS app from the menu, it'll prompt to download the support files correctly.
Had a similar issue after doing the Xcode 15 beta 2 download with both iOS 17 and visionOS 1 simulator checked.
When creating a new visionOS App it would tell me that I'm missing the device/simulator support and prompted to install it.
After confirming this it would show a download for a build that ended in xxxxxxxg instead of xxxxxxxf as was part of the download, so I assume that part was updated in the background and now is no longer the correct reference.
You’d think it would because of that but you’d also think if it was supported they’d mention it, if only to provide defense against the “only for friendless nerds that live alone” criticism, albeit at an absurd price point.
Got your point. I only saw it from a very limited technical understanding of how ARKit works and how shared experiences can be achieved on a framework level.
Not explicitly mentioning shared experiences other than video calls at all could also indicate this is not the way it should be framed (by focusing on the collaborative aspects that exist "today").
The price sure point prevents me and my family members from casually trying this experience.
I was merely suggesting that the technology for shared experiences already exists in the form of shared anchors.
You could be right wrt "if they didn’t mention it explicitly it's not part of their (currently) intended experience", but it might as well be due to "spatial computing" being sth that primarily will be shaped by their adopters along the way, which is something different than a corporation plotting the experience up front (as might be the case with metaverse?).
We're really missing the point here. Yes, that device can't do better. You can't do that with your friends together. But there's some similar in America right now. You can just jump in and in 2 or 4 minutes, and that kind of experience. Why to bother this kind of thing, this new tech? Doing things together is good, I think that's the main selling point of these devices.
Took me right back to my education and I never really thought about impedance values and why they were chosen.
We used to build 10base2 BNC terminators ourselves with empty connectors, a soldering iron, crimp tool and some hot glue using 47ohm resistors albeit with low tolerance values.
According to my boss they outperformed the industrially produced ones.
Back then, the network problems I was facing was either on the software/configuration side (wrong IPX/SPX network id) or simply faulty pc power supplies (I.e. getting shocked by mains voltage when touching ground, frequently happening after thunderstorms; also the terminators would be hot to the touch in such cases).
His videos definitely helped demystify my prior assumptions about how to incorporate ESP32s or STM32 in my own custom PCB designs (and not deal with external programmers, castellated edge ICs etc or resort to devkits adding quite some vertical space requirements and coming with issues of their own).
Boy was it a revelation when I got my hands at a "real" 909, 303+MPC+MIDI setup..