5 ms·
A "real hardware" approach to retrocomputing like displayed here appeals to me because it much better captures parts of the experience lost to emulation. Things
by kitsunesoba 3y ago
A "real hardware" approach to retrocomputing like displayed here appeals to me because it much better captures parts of the experience lost to emulation. Things like insanely low latency thanks to there not being a fractal spiral of hardware and software abstractions sitting between everything. It's much more simple, and you can feel that.
That said I'd also like to see some modern advancements brought in… for example a reproduction of a mid-90s 68k or PPC Mac using components manufactured on a much smaller node would be incredible. It wouldn't even have to be anything cutting edge like 3/5nm — even the now-ancient 14nm or 30nm would be amazing compared to say the 350nm node that the PowerPC 603ev was manufactured on.
- hedora 3y agoOne of the nicest screens I have ever used was on a monochrome ega laptop. Sure, the pixels were huge and slow, but the contrast ratio was really good without it being too bright. It’s hard to explain why it was so good. I suspect it was a bit transflective.
- deaddodo 3y agoThe problem with those old monochrome/greyscale LCDs is that they were horrendous for ghosting. I imagine if you just wanted an old laptop for wordstar or something, they’d be amazing due to their clarity; but I remember trying to play commander keen on them and getting a headache.
- scns 3y agoThanks for the flashback! Got my first own computer from my grandfather. It was a 486 laptop with a bulitin trackball and floppy drive. The display was as you described. Movement on the screen was just a blurry mess.
- nsxwolf 3y agoI'm also bummed out that these sorts of things are all including OPL2/OPL3 Adlib style clones when I'd much rather have Sound Blaster support with the digital audio channel.
- Andys 3y agoFair point, but in this case (4mhz 8088) I think it doesn't really have the guts to play 16 bit stereo digital audio? Update: apparently there's a hacky way to play digital audio through these FM synth chips, and several games and music players for DOS do just this.
- wkat4242 3y agoIt says this card is sound blaster pro 2.0 compatible if I understand it correctly. Of course on.aliexpress a lot is lost in translation.
- hypercube33 3y agoOne thing I'd love to see is a Pentium III class machine with a lot of cache and ram with a pcie SSD in pocket size. I feel like those machines were held back by hard drives of their day.
- riceart 3y ago> mid-90s 68k or PPC Mac using components manufactured on a much smaller node would be incredible. Why not just use an FPGA?
- kitsunesoba 3y agoAs I understand, there’s limits to the efficiency that can be achieved with FPGAs which makes them less than amazing for usage in e.g. laptops.
- Zobat 3y agoThat said, the 8088 wasn't 3nm, it was 3µm.
- grumpyprole 3y ago> fractal spiral of hardware and software abstractions I've often used "onion", but this is much more accurate :)
- myth_drannon 3y agoThe latency issue is very real. I work in Linux, Mac and Windows on fairly fast computers and programming in VSCode is so annoying because of the latency . On my 486 in Borland Turbo C++ ide, I click on a key and I see it instantaneous. Of course if you never tried it, you wouldn't notice the difference and think I have some sort of OCD. But it is really noticable. I don't know if it's 100ms or 200 ms but there is a difference.
- gbear605 3y agoTrying any editor other than VSCode might help; it’s notoriously slow.
- myth_drannon 3y agoEmacs is the same. I'm not talking about slow loading time or changing tabs which I understand. It's the act of clicking on a key.
- 2b3a51 3y agoEmacs? Keyboard lag? Is that with some kind of LSP or other itegration? I find that Emacs keyboard responsiveness on text (on a 17 year old core duo laptop) is indistinguishable from vi or nano.
- kitsunesoba 3y agoWhat they're getting at is that the latency is caused by overhead from the OS and USB, increasing it across the board regardless of editor — each keystroke has more layers to pass through compared to a machine from the 70s, 80s, or 90s, plus USB by the nature of how it works unavoidably adds latency that's variable depending on CPU load, since USB is CPU-dependent. There's also more layers in the display stack, increasing the amount of time it takes for what's actually displayed on screen to match machine state. A Core 2 Duo era laptop specifically might not suffer as much as its desktop counterpart though, because laptops from that era often used PS/2 for their internal keyboard+trackpad connection and thus don't suffer USB latency increases unless the laptop's motherboard was doing something weird like implementing PS/2 via an onboard adapter attached to the USB bus.