Almost laser pointer brightness at one point on the screen which you don't consciously see because the your eyes and the hold the image for a few milliseconds but do you really want to be looking at a extremely bright point of light all day. Higher refresh rates on CRTs lowered eye strain since that point could be effectively lower brightness, just refreshing twice as fast.

I honestly don't get eye strain on LCDs nearly as bad under any circumstance. 60hz LCD was a massive improvement over the 120hz CRTs.

Not trying to dismiss your perception, but are you sure the feeling of lag wasn't caused by some other bugs related to the bad config, say, some kind of interval mismatch? The difference between 240hz and 180hz is less than 1.5ms (if my calculation is correct) and I've seen OSU! player demand less precision than this.

But slow computers still need fast and reliable user interfaces, like radial pie menus.

Flight of the PIXIE - Yuja Wang:

https://www.youtube.com/watch?v=jDrqR9XssJI

PIXIE: A New Approach to Graphical Man-Machine Communication:

https://www.donhopkins.com/home/documents/PIXIE%20a%20new%20...

The computer shown in that video is a much more advanced and faster PDP-7, talking to an even bigger and faster Titan mainframe over a custom network!

https://en.wikipedia.org/wiki/Titan_(1963_computer)

It proves that slow computers can still be usable with better software, and that today's unusable computers are pissing away enourmous amounts of power, but still suck.

Edit: I just showed you a video of a computer with a light pen. Did you check that out? It's only two minutes, and I made it to musically highlight PIXIE, because most people don't want to sit through silent hours of the original digitized films from Cambridge.

PIXIE is not one box. It is a network distributed CAD workstation (1967–1972) -- like AJAX, but with P7 two-layer cascade phosphor on a 16ADP7A radar tube instead of HTML:

Interactive front end: DEC PDP-7 + Type 340 CRT. Real-time drawing, light pen, radial "control lightbuttons", graph model in core.

Compute / storage host: Titan (Cambridge’s Ferranti/ICT Atlas 2 prototype). RAINBOW apps: CONN, COMPACT, PLOT, LADAN circuit analysis, file store.

Wire between them: Wiseman link (Cambridge custom) + Lang's supervisor software. Move typed ring-structure models blocklet-by-blocklet.

Word size: 18 bits. Core (PIXIE design target): 8K words — "only 8K of 18 bit words ... no auxiliary storage".

Core (likely installed): 8K–16K (GUIDE: "here likely 16K").

Cycle time: ~1.75 µs. ~571K cycles/s.

Addressing: 13-bit + indirect bit.

Registers: One accumulator + link bit.

DEC made lots of "instruction sets in a chassie" extension boxes, and Wiseman at Cambridge rolled his own network with its own custom instructions.

Extended Arithmetic Element: Type 177 EAE. Multiply, divide, shifts, normalize (64xxxx op family in listing).

Precision incremental display: Type 340. Required — second processor + DMA display file.

Light pen: Type 370. Pen flag IOTs; diagnostic 7-78-M in pdp-7 reference.

Symbol generator: Type 342. 6-bit character codes → stroke sequences in display file.

Subroutine option: Type 347. DJS / DJP — display jump-to-subroutine (subpictures as display subroutines)

Titan link interface: Cambridge custom (Wiseman) Undocumented IOTs: LCF, LSF, LKE, LLB6, LLB18, LRB18, LLAM, LSA, LKD — networking as instructions.

The 340 display was an amazing ensembel of hardware, and could plug into a PDP-1 as well. The SIMH emulator supports it well.

Model: DEC Type 340 Precision Incremental CRT System (H-340, Nov 1964).

Tube lineage: Type 30E → 16ADP7A radar tube, P7 phosphor, 1024×1024 address grid.

CPU coupling: Cycle steal / data break — 340 fetches display words from PDP-7 core while CPU runs.

Vector timing: ~1.5 µs incremental vectors (vs Type 30 ~50 µs per CPU-plotted dot).

Display word types: Parameter, point, vector, vector-continue, increment, character; with 347: DJS/DJP subroutines.

Light pen: Per-word pen-enable bits; tracking uses cross + recovery spiral (~2× cross size) — spec in Ch. 5.

The 340 executes its own instruction set in memory (display file = program). PIXIE’s subpictures are literally display subroutines — Myer & Sutherland’s "wheel of reincarnation".

The Titan mainframe it communicated woth was another story, a big deal at the time! It ran RAINBOW: BCPL + FORTRAN analysis (e.g. LADAN), file store, COMPACT/PLOT/CONN/CONNMAP pipeline.

Identity: Ferranti/ICT Atlas 2 prototype; Cambridge name Titan (1964–Oct 1973).

Word size: 48 bits (8×6-bit chars or 2×24-bit halfwords).

Core growth: 32K → 64K → 128K words.

Addressing: Base/limit registers; user address ORed with base (not added).

“Cache”: Tunnel-diode operand slave store — Cambridge claims first cache.

Software “instructions”: Up to 512 extracodes (supervisor code in main store).

OS: Titan Supervisor / Cambridge Multiple-Access System — public 22 Mar 1967.

Storage: Two Data Products 16M-word discs; tape; card/punch.

Terminals: Cambridge 64-line multiplexor — 73 registered, 26 simultaneous; modems from 1967.

I've been able to gather together and shared all this information about it, thanks to an ongoing discussion with Heinz Lemke, the guy operating the lightpen in the video. He has recently unearthed 128 pages of PIXIE source code in PDP-7 assembly source and octal machine language, and we're scanning it in, and working on reincarnating it with SIMH and a virtual lightpen!

It is a slow, ugly, overcomplicated mess designed by people who do not know how threading works and don't know how to do email properly. Yuck.

My arm would get tired quickly if I tried to point to things by pen as fast as I can by mouse.

Buxton, W. (1983). Lexical and Pragmatic Considerations of Input Structures. Computer Graphics, 17 (1), 31-37.

https://www.billbuxton.com/lexical.html

TL;DR: "It's complicated."

TL;DR-but-not-illiterate:

Claim: How you spell a command (lexical) and how you move your body to do it (pragmatic -- device, gesture, layout) are different problems. Treating all "locators" as the same -- mouse, tablet, light pen, joystick -- is a design error.

Figure 1: Input devices sit in a grid by what they sense (position / motion / pressure) and how your hand moves. Light pens cluster with touch screens; mice with tablets; joysticks with trackballs. Different motor skills, not interchangeable.

Why mrob is wrong here: "Graphics need a mouse" assumes one cheap locator fits every graphics task. Buxton's point (and skydhash's): viewport pan/zoom wants isometric/spacemouse control; tracing wants a pen. A mouse is a compromise, not the default.

What PIXIE has to do with it: I already showed the right transducer: light pen on a vector display for tracing and radial lightbuttons. No 1000 Hz, no mouse. The 1969 demo is a pragmatic match; Buxton (1983) is the theory that says that match matters.

Buxton defines the terms Chunking: Good interfaces fuse actions into one gesture (pick up => drag => release). Pie menus and pen tracking are in that family -- syntax, tension, and closure enforced by the body, not remembered as steps.

Quieter computing isn't only slower phosphor or fewer Hz. It's also picking the right input device for the task instead of defaulting to mouse + glass.

Miyamoto closes the loop from the other end. He designs from the hands: start with the kinesthetics of the controller, how the thing feels before what it does.

Will Wright on Miyamoto: "Miyamoto starts from the kinesthetics of the controller. What is this thing going to feel like in my hands? Will I feel like I'm instinctively connected to this world? As opposed to, I've got sixteen buttons, and I'm trying to figure out which button does the super-thrust power-up, in which case it's very cerebral, kind of like learning to play the piano. Rather than, you know, just picking up a shovel and starting to dig."

Will Wright on Miyamoto, New Yorker profile "Master of Play":

https://www.newyorker.com/magazine/2010/12/20/master-of-play

And Miyamoto judges from the face. GDC 2007 keynote: "What I always try to envision is the face of the player when he or she is experiencing the game, not any individual part of the game."

I've posted about his two GDC keynotes before, with links to all the videos (1999 full, 2007 parts 1-6, the wife-o-meter) and my transcription of the face-of-the-player passage. They bracket an amazing time in game development history, 1999-2007, and every word is profound:

https://news.ycombinator.com/item?id=27918255

Shigeru Miyamoto GDC 1999 Keynote (Full):

https://www.youtube.com/watch?v=LC2Pf5F2acI

The grandfather and granddaughter moment, GDC 2007 Part 3 at 12m29s:

https://youtu.be/SY3a4dCBQYs?t=12m29s

Buxton is the theory, Miyamoto is the practice: pick the transducer with your hands, verify it with the player's face. Watch Heinz Lemke's hands and face in the 1969 PIXIE film and you can score the interface without reading a word of the thesis.

Flight of the PIXIE, 1969 light pen and radial menus on the PDP-7:

https://www.youtube.com/watch?v=jDrqR9XssJI

Heinz (the guy in the the video demonistrating the PIXIE system he wrote) recently wrote me:

>Hope it doesn’t surprise you, that the use of a light pen in combination with the radial menus of PIXIE are very much related to the activity of a typical technical draftsman of the 1950’s. During that time and in the beginning of the 1960’s, I was such a technical draftsman in a number of companies in Germany, see attached a typical drawing from 1960 (few days ago digitised from a 1m by 2,5m drawing).

>What you may observe in this drawing is the frequent use of special symbols such as circles, squares, arrows, valves, nuts and bolts, etc. These are usually entered into the drawing (on a vertically positioned but slightly tilted drawing board) with a pencil at one hand and a special template with symbols held at the other hand. In PIXIE, I simply translated this into the light pen with the templates being simulated by a radial menu of symbols (PIXIE’s control light buttons) that can be dynamically adapted to the technical content of the drawing. Both, the light pen and the template can then be operated with one hand, leaving the other hand free to enter data, for example on the TTY.

>If at the same time the drawing gestures and drawing components are also translated into a computational model on the small computer, which after model completion can then be analysed or simulated on a larger computer, it becomes a winning situation all round. This was the aim of the PIXIE-Rainbow (PDP7-Titan) framework in those times. What appears to be missing now in our times is the “vertically positioned but slightly tilted drawing board” that used to be in the 1950’s and 60’s a social meeting point for all the colleagues in the drawing departments. This may be the loosing part of the advances in computer graphics.

Pretty hard to think of anything I want to happen upon returning home that depends on barometric pressure. I can't really understand what you're getting at here more generally either, though.

When I return home, I (1) put down anything I might be carrying, (2) greet the dog, (3) maybe turn on a couple of lights, (4) maybe take off a coat or jacket, then (5) just sort of go about my life, doing whatever it is that I need or want to do.

So genuine question: what complicated series of events do you want to happen when you return home?

A fair bit of this can already be implemented with a workflow that looks like:

Sensors -> zigbee or zwave -> home assistant

Or Internet based data -> home assistant (weather, external data, the humidity sensor at your local airport might be more accurate than the one on your deck, etc).

And then open source extensions/tools built into home assistant to cause things to happen, send external signals, run scripts, call APIs, talk to a local LLM, etc.

It's very much not a black box of opaque closed source software, rather the opposite, and it has a very active and technically knowledgeable enthusiast development community behind it.

If we posit for a minute that the data acquisition into locally-run system problem is not impossible to solve, then all kinds of things can be built atop that to serve your own unique home preferences.

> Having the logic run local

At the smallest scale people are doing this on raspberry pi, though I have seen people have much more satisfactory results on something with somewhat more system resources like an Intel N150 CPU based mini PC with 12GB RAM.

Going a bit up a bit in budget from there, people are running home assistant and frigate together on the same tiny system that is a KVM hypervisor (again all open source, no black boxes involved), so that Frigate can have a GPU passed through to it for image object detection/video analysis pipelines.

A lot of desires that require computation might not be so "logical" in nature, at least not in the automation kind of things.

I use a lot of computing for entertainment purposes and leisure activities, like reading news/ebooks, watching shows/movies, listening to music, gaming, clothing/furniture shopping, going into all kinds of rabbit holes etc. The HN link from GP was such an example.

How could any automation through other physical input help there? Sure, there is vector similarity and recommendation engines, but how could any sensor interpret my aesthetics and my desires to belong/grasp/play/participate?

That sounds exactly like the empty sci-fi promises that current AI companies are betting on.

But going from, say, a Sony Trinitron CRT [1] to one the first LCD was a huge step backwards, which is what GP is saying.

I did wait for the first LCD with DVI support before switching: for the difference in crispyness between an analog video cable and a DVI one was night and day (I also won many points when I helped friend set up a startup in southern california and we realized one person out of two had picked the VGA cable instead of the DVI one when setting up its desk / computer: so with the CTO late at night we went through every PC and changed all those with VGA cables to the DVI ones).

We're not saying modern days flat monitors aren't better than old CRTs: we're saying the first flat screens were complete suckage.

[1] even though on a Trinitron you could see two faint horizontal lines (Trinitron were still amazing)

At 1600x1200@75Hz. Crystal clear and sharp.

I also had something by ELSA with a S3 and a translucent blue RAMDAC by IBM. Same thing. But only 2MB VRAM, and PCI. So only useful for testing purposes.

Regarding the usability under Linux, or rather support by XFree86, Matrox accelerated some graphics primitives in the DDX-layer, or whatever that was called then. So did the Voodoo3's! Independently of that Glide stuff for 3D.

Really nice to have.

I don't think that's true across EU. Seemingly very non-traditional forums, huge ones, stay online still today, many hosted within EU. Probably Flashback Forum is one of the oldest and biggest forums in Nordics, geared towards "freedom of speech" and still online today.

Sounds like it's very difficult in Germany indeed to host an anonymous or pseudoanonymous website/forum, but Germany is also not the entire EU or Europe.

I don't understand people bitching about Impressum like its some sort of German-only thing.

I am sure if you looked at the legislation of many countries, you would find similar requirements for companies to identify themselves. Off the top of my head you have Austria, France, Italy, Spain, Switzerland, UK ... and probably the rest of the EU/EEA states as well. IIRC it is also a requirement in a number of Asian countries.

In reality requiring an Impressum (or equivalent in other countries) is no different to having to put identification information on a company's old-school letterhead. Its just good manners so that your customers/suppliers can know who they are dealing with.

Personally I dislike the US attitude of trying your damnest to hide any trace of your company's identifying or contact information from your website. ;)

It's bad enough but not quite as bad as you claim. Article 19 of the DSA exempts small websites from the most onerous and expensive rules that big platforms have to comply with.

https://eur-lex.europa.eu/eli/reg/2022/2065/oj/eng

You get a request for which data is processed from a user and suddenly you have to spend 8 hours in paper archives to find if the weekly charity mailing has ever mentioned this person by their full name.

In previous eras we'd just ignore laws like these. But in this era we have become, as they say, cucked.

And I see the EU being corrupted more and more by fascist and extreme right parties from Italy, Hungary, Poland, Czech etc. Even in France and Germany they're on the brink. This is why I think it's important to push back against further integration.

Yet my info is exposed by business registration and I get junk mail left and right.

do you think everything on the web is just e-commerce?

and if there are no customers who buy products why should there be an impressum?