@rher@FailurePersonified I think you learn one assembly language, it makes everything else you write make sense, you start to see things very differently. You learn two and then you can kind of extrapolate: they are largely neither complicated nor too different from each other, so learning an obscure one is easy if you know anything with a similar ISA. This one has fewer registers and it's a pain, this one has some registers that are saved for interrupt handlers, this one saves all registers for interrupt handlers, etc.
ARM is a bit weird (in a fun way): conditional flags on *any* instruction rather than just branches, that's the big one. Then aarch64 and thumb.
Anyway, especially old chips, you can pick up assembly for them in a couple of hours as long as you know a similar chip. For example, if you know x86, you can probably pick up most of the old 8-bit chips right away, Z-80 or 6502 or whichever.
So what I did was I learned x86 (or at least as much of it as a person learns if they want to get by; maybe no one knows the entire ISA), then I ended up with an ARM chip and it was delightful, that chip is a lot of fun. Then anything else, I have just picked up as I went. I'd like to play with RISC-V more, but I look at the compiler's output and I can more or less read it.
Since you are talking about computer history, do you know of anything else like this? >https://www.fourmilab.ch/autofile/e5/chapter2_2.html As in, software industry history related quality reads? Including blogs, quarterly/annuals reports, anything really. It's been difficult to find quality sources for the development of this industry.
> It's been difficult to find quality sources for the development of this industry.
Absolutely. You get pieces here or there, but it's easier to get the history of computers themselves rather than the businesses.
This is one of my favorite interviews. It's about the founding of Control Data and then Cray: https://americanhistory.si.edu/comphist/cray.htm . It's mostly about Cray but he talks a lot about starting the companies up, including how they sold shares.
@ins0mniak@FailurePersonified@rher Ha, I got it from some English kid that lived in France. (He's older than we are but I'm pretty sure he was a kid when he wrote it.) It was at heyrick.co.uk , had tutorials and reference, and enough information that you could hack out some GBA games if you also had a reference to the GBA's memory map.
For sure. A pi can't run ps2. A rock 5b may be able to though.
For ps2, I'd advise buying a refurbished Dell optiplex micro with an i7 6th Gen and 16gb ram. You can grab them on ebay for between $100 and $200. They usually don't have bt, so you'd need to buy a dongle for the controllers.
As far as storage on a pi, you could get a 2TB external SSD. Then you could have entire sets, like dos, amiga, and ps1. Read time from usb isn't a huge issue for those systems.
The Odroid n2+ is another solid option if you want an sbc for emulation. Hardkernel makes quality boards.
I have one, but use it at my office as my main pc. I threw an nvme in it, and I throw it in the safe when I leave. There's literally no client information on any other cpu in my office, so if someone breaks in. 🤷♂️
I've never tried emulating on it.
I have a dell optiplex micro in my living room connected to my TV. It runs arch with plasma. Kodi and retroarch run great. The actual roms / media files are in a box upstairs. It's an old dell tower. I threw 4x8TB HDDs in it. It also runs local irc, mumble, mpd, and kabanboard servers.
I use moca adapters for my network, so I can use the existing coax in the walls.
I have full rom sets from Fairchild channel F through PS2 / Xbox, movies, albums, and an entire copy of project gutenberg, the main and generated collections. 😏
@ins0mniak@FailurePersonified@rher Oh, yeah, GBA is easy to emulate because the hardware was really simple. ARM 7tdmi, an old Gameboy CPU that was used to play Gameboy games directly but also was used for comms (and inaccessible directly) when the GBA was in GBA-mode, sprites, background layers, and low-res, no 3D acceleration. Really fun to hack those things.
I should have been more clear. I still have my actual ps2 so emulating it is pointless for me other than my ex jacking my fucking Ghost in the Shell game.
@p@FailurePersonified@ins0mniak@rher GBA is nice because it's the exact sweet spot where the hardware is powerful enough to do really interesting things and use "normal" toolchains, but it's still dead simple enough that you can really understand what's going on.
@allison@FailurePersonified@ins0mniak@rher Yeah, and enough real 2D hardware in it, so you could do all the fun stuff, move stuff around just by writing the new address or tricks with palettes. You can't do tricks with palettes any more.
I use sshfs to fuse mount the remote directories to the computer in my living room. Those mounts don't include my work files. Those are on a separate system with separate accounts and ssh keys. They can only be accessed from my home office cpus.
@Humpleupagus@n3f_X@FailurePersonified@ins0mniak@p@rher I’m mulling this over now and I was hoping to find a way to do that with one computer, but TV/Playstion media server access software sucks. Apple‘s attempt isn’t much better (and requires the TV to do something it doesn’t want to: be an Apple device). I have all the stuff I just…don’t want work in my living room, but if that’s your setup thats probably the way to go 😑
@p@FailurePersonified@Humpleupagus@ins0mniak@rher Making files is easy under the UNIX operating system. Therefore, users tend to create numerous files using large amounts of file space. It has been said that the only standard thing about all UNIX systems is the message-of-the-day telling users to clean up their files. -- System V.2 administrator's guide
I also tried this with a pi. Biggest problem were the VMs and cross compiling the packages. Also made me appreciate plan9's network first architecture a lot more. Really, this type of distributed computing would be a lot better if we were using plan9.
It would be nice to have system wide accounts and devices.
At this point though, I have a script that tests whether a linux system is arch based or debian based and then sets nearly everything up based on a series of options, so I can set up a system from scratch and get it connected to the network pretty quickly.
Most of my computers, with a few exceptions, are just glorified thin clients at this point.
>but it's easier to get the history of computers themselves rather than the businesses. It is easier, but you need to understand the participants in order to understand the industry. From those autodesk files you get information about the types of applications being developed, the number of clients, the precise availability of programmers, whether these programmers are organized into a community, general timetables for development, need for marketing, need for user education. A lot of these information are very difficult to get from general writeups that focus on the larger firms and present an aggregated image. Oh and John Walker's technical and industrial insights are a lot better than those of dedicated industry historians.
> From those autodesk files you get information about the types of applications being developed, the number of clients, the precise availability of programmers, whether these programmers are organized into a community, general timetables for development, need for marketing, need for user education. A lot of these information are very difficult to get from general writeups that focus on the larger firms and present an aggregated image.
Indeed.
> Oh and John Walker's technical and industrial insights are a lot better than those of dedicated industry historians.
He's pretty great overall, and very prolific writer.
You know, he's alive and responds to email, he's probably got some good pointers, what to read.