kill unix

1 Name: Anonymous 2018-02-16 23:24
Why do we still put up with the absolute fucking garbage that it UNIX 1.0. It's time for UNIX 2.0. Let me explain: Processes on linux take as input env variables and an array of c strings, print text (the "universal interchange format") to stdout and stderr then they return a single byte.

Processes are the fundamental building blocks of our operating system. All the build systems, init scripts and package managers we have are made out of this. So it's important for it to be solid i.e. it should be done with well typed structured data - not piles of spagetti strings. There's heen so many problems because of it like invocations failing when inputs start with '-' becaues it get misinterpreted as a flag or programs breaking when the file path contains a space.

We should revise the concept of a process to declaratively explain its input in a well typed form, make a new kind of "terminal" to invoke them that isn't based on simulating a 1970s teletype machine (so shit like resizing and scrolling will actually work), have a scripting language that doesn't misintepret fucking everything you write and work entirely based on pasting strings together and then implement a new set of coreutils where things "ls" simply produce a *list* (an actual data type) of files in the directory instead of just some text that it's recommended you shouldn't ever parse.
2 Name: Anonymous 2018-02-16 23:56
it should be done with well typed structured data - not piles of spagetti strings.
There were operating systems like that in the 1970s but the UNIX philosophers wiped them all out. Keywords to look for are orthogonal persistence, typed object store, persistent objects.
We have persistent objects, they're called files. - Ken Thompson
3 Name: Anonymous 2018-02-17 05:09
What do you think about Plan 9 OS?
4 Name: Anonymous 2018-02-17 14:28
5 Name: Anonymous 2018-02-17 14:29
I think it's total shit and the people who bring it up are brainlets.
6 Name: Anonymous 2018-02-18 12:35
Fantastic. Opinion discarded.
7 Name: Anonymous 2018-02-24 04:11
but it's objectively correct - if it wasn't shit, wouldn't everyone be using it? for it to have had such little success, doesn't mean that there must be something fundamentally inadequate about it? how do you unify thinking that it's good with the fact that nobody, literally nobody, actually uses it for anything?
8 Name: Anonymous 2018-02-28 06:26
There is an idea in economics known as "the sunk cost". A sunken cost is a cost that has already been incurred and cannot be recovered. For the case of Plan 9, people have generally have sunk their investment in Windows, OS X or Lignux. For those people, they are not interested in changing their systems into a Plan 9 based platform.

If your negative opinion of Plan 9 is based on the idea that it has a low popularity, then your opinion has no merit on the technical abilities of Plan 9. Opinion discarded.
9 Name: Anonymous 2018-03-10 16:50
literally nobody
that's false.
Wouldnt that be great though, if everything produced and accepted arbitrary, incompatible binary bullshit. I look forward to all the exciting new code that has to be written to parse and modify the uniquely optimal output of programs like 'list' and ''.

Leave this field blank: