4 ms·
Kind of on the contrary I feel like this is a great reminder of how possible doing very complicated things with much more rudimentary tools than we have availab
by stormbrew 13y ago
Kind of on the contrary I feel like this is a great reminder of how possible doing very complicated things with much more rudimentary tools than we have available now is, and that we're really not so far from the 'ancient' ways than we like to think.
And though the product would be quite limited, if you're willing to rely on the BIOS doing this from a raw kernel implemented in itself wouldn't be that bad (and having a VM to do so faster would help, but booting off a floppy would still be possible). If anything it might make things a little simpler since you wouldn't have to even use a syscall to get memory, if you set up the pagetables to map to physical memory 1:1 you've got free reign.
And that would definitely be an interesting process to try to take this through.
- rst 13y agoThe funny thing is that the EDSAC, one of the first functioning computers (perhaps the first, depending on how you define "functioning" and "computer") had a boot loader (the "initial orders") which was capable of manipulating single-letter address labels in a manner at least reminiscent of HEX2. (Or perhaps the reverse?) And in fact, this "annotated machine code" was how programs were generally written for the machine. There are a lot of fine paywalled descriptions of the system. One that's generally available is in the documentation for the EDSAC simulator here (Tutorial Guide, chapter 3): http://www.dcs.warwick.ac.uk/~edsac/ http://www.dcs.warwick.ac.uk/~edsac/ http://www.dcs.warwick.ac.uk/~edsac/Software/EdsacTG.pdf http://www.dcs.warwick.ac.uk/~edsac/Software/EdsacTG.pdf