log☇︎
61 entries in 0.596s
asciilifeform: meanwhile , in the world of the c-machine, https://archive.is/e5xD9 >> 'curl contains a heap out of buffer read vulnerability... stderr... formats the output to wrap at 80 columns. The wrap logic is however flawed, so if a single word in the message is itself longer than 80 bytes the buffer arithmetic calculates the remainder wrong and will end up reading behind the end of the buffer' etc
a111: Logged on 2017-11-20 12:19 spyked: http://btcbase.org/log/2017-11-14#1737923 <-- also, spyked's adalisp is missing more fundamental things, such as closures. it's an early prototype, barely usable, but > 0. interning is of course considered, but not added yet. anyway, phf, consider the following point: built-in symbols (car, cons, etc.) still have to point *somewhere*, and that somewhere must not be addressed in a C-machine style! symbols should point to Lisp memory (via
asciilifeform: otherwise it's a purely c-machine exploitation problem
mircea_popescu: anyway, let's restate : "since you're stuck with a c library anyway as long as you're on c-machine, at the very least to do init and end ; and since musl is not that terrible and only links what you use anyway ; just use it and be done with the headache -- the alternative is bloating ada with nonsense"
phf: the notion of "static content" is a c-machine-ism
ben_vulpes: i've never been burdened with the "This Is How Things Are" of the c-machine
asciilifeform: but it ain't. because 'dun help c-machine any'
a111: Logged on 2017-12-10 14:56 phf: http://btcbase.org/log/2017-12-10#1749235 << nobody has solved the bootstrapping problem. every time you touch a system that's in any way attempts to not be a c-machine you run into that issue
phf: http://btcbase.org/log/2017-12-10#1749235 << nobody has solved the bootstrapping problem. every time you touch a system that's in any way attempts to not be a c-machine you run into that issue ☝︎☟︎
a111: Logged on 2017-11-20 12:19 spyked: http://btcbase.org/log/2017-11-14#1737923 <-- also, spyked's adalisp is missing more fundamental things, such as closures. it's an early prototype, barely usable, but > 0. interning is of course considered, but not added yet. anyway, phf, consider the following point: built-in symbols (car, cons, etc.) still have to point *somewhere*, and that somewhere must not be addressed in a C-machine style! symbols should point to Lisp memory (via
spyked: http://btcbase.org/log/2017-11-14#1737923 <-- also, spyked's adalisp is missing more fundamental things, such as closures. it's an early prototype, barely usable, but > 0. interning is of course considered, but not added yet. anyway, phf, consider the following point: built-in symbols (car, cons, etc.) still have to point *somewhere*, and that somewhere must not be addressed in a C-machine style! symbols should point to Lisp memory (via ☝︎☟︎☟︎
asciilifeform: on c-machine -- yes.
asciilifeform: we are looking at a real engineering constraint, flowing from the broken way in which c-machine implements the ring of integers.
asciilifeform: how does c-os on c-machine work as 'blob free' ?
asciilifeform: environments in the lisp sense, which resemble c-machine stack frames except not retarded.
phf: c-machine assembly level has a certain shape to it. when you write int foo { bar(); } an assembly of a certain pattern gets spit out
phf: well, c-machine assembly level is not particularly good, but it's not guaranteed to be consisten either
asciilifeform: nor does c-machine enable restarts (how many programming folk even ~CONCEIVE~ of the notion of meaningfully RESUMING a crashed proggy ?)
asciilifeform: but it suffered from the crippling limitations of the c-machine, and withered away.
trinque: openbsd's httpd doesn't look like that much src to read, while we're still in c-machine hell.
a111: Logged on 2016-07-17 17:11 asciilifeform: the c-machine is an idiot foundation, which inevitably breeds the pests we are familiar with.
a111: Logged on 2016-07-17 17:23 asciilifeform: the c-machine pointer mechanism, and its non-boundchecked array access in general, is quite similar to mains wiring prior to the invention of the fuse, where mains socket has nfi that it oughtn't source 300 amps and melt itself, igniting the entire wall
a111: Logged on 2016-07-17 17:20 asciilifeform: anywhere other than ~equator, a block of flats without a boiler, plumbing, and radiators, is quite analogous to the c-machine
asciilifeform: the standard argument in favour of c-machine, 'transistors have a cost', holds 0 water: the modern cpu is nothing like pdp11, and has plenty of transistors wasted on idiot epicycles to paper over the retardation, e.g., page table, tlb cache, etc
asciilifeform: the c-machine pointer mechanism, and its non-boundchecked array access in general, is quite similar to mains wiring prior to the invention of the fuse, where mains socket has nfi that it oughtn't source 300 amps and melt itself, igniting the entire wall ☟︎
asciilifeform: anywhere other than ~equator, a block of flats without a boiler, plumbing, and radiators, is quite analogous to the c-machine ☟︎
Framedragger: c-machine as in von neumann machine
a111: 31 results for "c-machine", http://btcbase.org/log-search?q=c-machine
Framedragger: $s c-machine
asciilifeform: the c-machine is an idiot foundation, which inevitably breeds the pests we are familiar with. ☟︎
asciilifeform: the crapolade is attributable to individual вредители in all cases, but exists ~structurally~ because of c-machine and idiot unix.
asciilifeform: it was ~conceptually~ The Right Thing in 1,001 ways that c-machine is not, but not something that really deserves to be rebuilt verbatim (other than for study)
ascii_deadfiber: this is what c-machine buys you.
PeterL: aha, so if you ditch c-machine then you could do better?
asciilifeform: (the other choice, if trying to escape from ada, is to ditch the c-machine silicon, but i just checked my sofa cushion and still did not find a spare $10m in it with which to do this.)
asciilifeform: or whatever else on c-machine.
asciilifeform quite deeply uninterested in c-os for c-machine.
asciilifeform: mats: again with the 'let's seeeekk00000r3 the c-machine with magic pixie dust'
ascii_field: 'let's run same old idiot c-machine but with sprinkle of haskell on top for flavour'
ascii_field: likewise, 'mitigations' which ultimately prolong the agony of c/c++/c-machine - are works of evil.
asciilifeform: scheme with continuations (call-with-current-continuation aka call/cc) doesn't cleanly compile to the c-machine.
asciilifeform: 'targeting' them as if they were c-machine arch would result in a terrific blob of unreadable shit
asciilifeform: has to interface with megatonnes of c-machine os crud
decimation: but that would be a nightmare of c-machine trickery
decimation: in fact, to even seriously consider writing software for which you are liable is to instantly consider dumping the existing c-machine
decimation: dynamic libraries are the depressing result of the c-machine's sad attempt to have a global namespace
ascii_field: (c-machine - sucks)
ascii_field: selling turd where c-machine abstraction stack continuously leaks into the supposed 'lisp' - is pure sc4mz4tr0nix
decimation: this is probably the most rational approach toward the modern c-machine
decimation: c-machine implementation of characters is enraging
decimation: I'm not sure how any c-machine can really be 'formally verified'
decimation: the c-machine is a mistake, but ganging multiple c-machines was an even bigger mistake
decimation: derping on c-machine is essentially free
gabriel_laddel: you're claim is that the problem with lisp is that it's not built for the c-machine. my point is that even on c-machines it's still better than anything else out there.
gabriel_laddel: decimation: re, the problem with lisp... this isn't CL's problem, and not a real problem anyway. Compare to writing ~ALGOL on c-machine.
decimation: the problem with lisp is that it is alient to the modern c-machine, especially the multi-core variety
mircea_popescu: in other words, back to the topic : i think the people arguing that c is just the way to talk to the c-machine, aka von neumann machine have it wrong - c is just the way average people abstract. fortunately that fit into an already existing machine, but if it hadn't it'd have been forcibly fit anyway.
decimation: inevitably there's going to be some usg 'dept of c-machine hygiene'
decimation: asciilifeform: what kind of bell must be rung to convince people that 'reviewing' the monstrous pile of C code (on the C-machine) modern software uses is not just difficult - it's impossible
decimation: asciilifeform: do mix Mr Yarvin's metaphor - what would "Haakon the Fairhaired" do about computers today? would he pointlessly toil with c-machine kazoo or would he bide his time and work on something unique?
decimation: http://fish2.com/ipmi/river.pdf << problem: c-machines are buggy and crash all the time ... solution: glue another c-machine on top of the c-machine ... problem: hardware vendors have crap security on new c-machine ... and you can't examine the firmware