<oriansj>and the 1 hour build times we used to complain about will probably seem speedy on a system with only 64MB of RAM <oriansj>and it would increase the requirements for a hosting OS to include functional swap (not sure what that would mean yet in terms of implementation complexity) <muurkha>64 MiB of RAM seems like an enormous amount to me <muurkha>like, a ridiculously excessive amount to *require* for bootstrapping, though it may be convenient to *waste* it under some circumstances <muurkha>we'll see what I can get running on them <muurkha>Wirth's Ceres workstation I mention in that thread is one of the most inspiring systems; in 2 MiB of RAM it ran not only the compiler but the GUI, networking, editor, etc. <muurkha>my Linux box had only 64 MiB of RAM from 01996 to 02001 <muurkha>it would have taken more than an hour to recompile the world, of course! but it was a 5x86-133 ***alMalsamo is now known as lumberjack123
<oriansj>muurkha: I agree that 64MB is plenty to do a lot; however MesCC needs about 2GB to compile TCC if I remember correctly <oriansj>now if I restarted work on M3; we could do the TCC compile in 64MB and then it would be enough to bootstrap linux and go from there. ***alMalsamo is now known as lumberjack123
<muurkha>oriansj: I wonder how hard it would be to get MesCC to compile TCC within a megabyte or two <unmatche1-paren>how do i actually compile M2-Planet with the produced AMD64 stage0 binaries? (I have them in ./stage0) <unmatche1-paren>when i try to compile cc_reader.c with cc_amd64 it complains about a 'missing {' <unmatche1-paren>i can't get mescc to work either... 'unhandled exception:error:(("include-from-path: not found: " "nyacc/lex.scm")) <stikonas_>unmatched-paren: mescc can't build M2-Planet if I remember correctly ***stikonas_ is now known as stikonas