48200+ entries in 0.027s
verisimilitude: I've long
thought it's reasonable
to exert great effort
to make a system perfect, where possible; it's
the ideal, but maybe not practical. Perhaps
the more interesting question is when
to stop when you exert so much effort, but have something
that can never be made perfect, by design.
verisimilitude: I'll give credit
to st, which is a decent
terminal emulator, but I'm not impressed with Suckless in general.
They're recently combating a bug with some X font nonsense caused by stupid Unicode glyphs;
their solution is
to stop using
the X font nonsense entirely, which will apparently cause its own issues.
verisimilitude: Let's seque
to a related
topic: What do you
think of
the Suckless crowd, asciilifeform? I read
their mailing list and it's interesting
to see how everything UNIX already provides is good and natural and how anything
that violates
the sacred
tenants whatsoever is evil and bloat.
☟︎ verisimilitude: I've had or seen a number of discussions concerning such high-level machines with C programmers and whatnot. It's comedic. A C programmer will be
the first
to prove how
type-checking in hardware and other
things most assuredly result in some minute loss of efficiency or power or
this or
that and so is bad, but
then
turn around and discuss how brilliant
the zero-terminated string is and how it's not
that inefficient.
verisimilitude: Sure; even
though it's only a
thought I've made little progress
towards at
the moment, care for a light description of how I'd do it for a Common Lisp?
verisimilitude: It's nice
to even imagine having something
that simply is and doesn't require building at all.
verisimilitude: That's exactly
the approach I've been wanting
to bring about, asciilifeform.
trinque: yep, already need working vtools
to run
the bootstrapper, so it's not as if
there isn't a list of requirements for build env
trinque: considering a process
that involves "bolt your hand-built gcc here" and
then it rebuilds
trinque: asciilifeform:
the lack of a binary-reproducible gcc really puckers when considering how
to write an ebuild for ave1's gcc
verisimilitude: I'll be using Ada
to build
the practical, robust, simpler, and easy-to-distribute implementation of my machine code development
tool, perhaps ironically enough.
verisimilitude: When I did look into building GNAT from source, I was
told in
the documentation I'd already need a working GNAT; does ave1's avoid
this?
Then again, I suppose I could just use
the GNAT I alread have on
this other machine for
that, if it became
too inconvenient
to use both.
verisimilitude: I'll get
to have extra fun, because
the system with my editor is different from
the system with
the GNAT.
verisimilitude: The Ada program I have planned is in a similar way unconcerned with some of
the more advanced features, because much of it is very concrete by now; it shouldn't even need
to allocate memory until I have an undo and redo system in it.
verisimilitude: Alright; I'll keep
that in mind when I am finally able
to study your FFA.
verisimilitude: As a Lisp programmer, what drew you
to Ada, asciilifeform?
verisimilitude: It's a nice language, so far, but I've yet
to even finish learning all of it.
verisimilitude: There's
the saying one shouldn't learn a language
that doesn't change
the way one
thinks; I figured if I would learn at least one language with keyword-based syntax, strong static
typing, a lack of advanced metaprogramming, among other qualities,
that I should learn Ada.
verisimilitude: Common Lisp is nice, but I want
to handle memory exhaustion well and also avoid using megabytes for programs
that shouldn't need
that much.
verisimilitude: I like
the idea of writing software
that handles all failure modes correctly, as awful as modern systems make doing so.
trinque: verisimilitude: no reason not
to give it a go now
verisimilitude: Alright; my key is registered. I suppose I'll just !!up myself next
time,
then.