240100+ entries in 0.134s
mircea_popescu: instead,
those rapidly diminishing returns drive a diminishing of quality.
mircea_popescu: there is absoloutely no reason
there should be more
than maybe a dozen or so fonts. but
there are, because "easy
to make", and in spite of
the rapidly diminishing returns of "making".
mircea_popescu: a) you can not meaningfully say if you've found a "good" font, seeing how you can't say what portion of fonts youv'e actually seen ; b)
the very notion of good font is impossible because in principle all possible fonts were made already in
triplicate.
mircea_popescu: consider a simpler problem - you wish for whatever reason
to "pick a good font". suppose your "good" is specified. you proceed
to... look at fonts, selected by some criteria. eventually you are
tired
to
the point your eyes blur, and haven't seen 5% of
the available offerigs in "gothic medieval" from one place.
there's more places. when are you done ?
phf: now
the question of whether or not designing a font language
that simulates movement of calligraphic brushstrokes is a good idea..
mircea_popescu: phf even limiting
the matter
to
ttfs (seriously, postcript is a whole extra layer of crazy) one quickly discovers some very serious conceptual problems. i don't even mean from a
technological pov.
a111: Logged on 2016-11-30 22:42 asciilifeform: obesity is not per se damning but i'd really love
to know wtf
there is justifiably 52 zipped MB of , in
there.
mircea_popescu: the undigested corn nibblets in used
toilet are no sort of prize whatsoever ; except for bacteria.
a111: Logged on 2016-11-30 22:35 asciilifeform: soooo lucent bought by nokia, and
the latter now by microshit, which now owns
the bones of bell labs.
trinque: in reading
the source,
the bsd folks crapped a bunch of
things
that should be in a sed-like into cat.
a111: Logged on 2016-11-30 22:31 phf: i've no idea what
that cook_buf is supposed
to do
phf: i
think
that's a separate set for
their own postscript implementation
phf: oh
that makes sense,
they use bitmap fonts
a111: Logged on 2016-11-30 22:03 phf: well, we've already established
that running everything on 2005
thinkpads is one viable option, but
that's until supplies last
phf: probably mercurial
tree
phf: lol you find flaws with ~everything~. it's utf-8 heavy, no reason
to even look :p
trinque: they got it a .io for
the better
to hackernews you with
phf: (it also has a nifty set of astro
tools,
that i use
to chart
the skys from
time
to
time, because it's really easy
to use and i don't know any better)
phf: trinque: i very rarely explicitly run it, it's just
there for when i want
to read some c code. i sometimes gauge
tool sanity by
trying
to do
the same
thing with plan9
tools. anyway it's plan9port running on mac 10.9.5
phf: i've no idea what
that cook_buf is supposed
to do
☟︎ phf: well, i only have plan9 cat on my box and it's
trully
tiny, but
then it's plan9 (just does read/write
to buf[8192])
phf: cat ought
to be 8-bit clean, i suspect most of
the bloat will come from gnu attempting all kinds of optimization. it probably optionally mmaps cat foo, etc.
phf: i don't know. it has a whitepaper
though
trinque: and something quite stupid and forth-y could handle parsing a rudimentary set of possible statements
to a db which eats from one serial port, shits
to another
trinque: I've
thought having a hardware parser/filter on a serial line where only certain bits can possibly flow over, and only in certain order, would compliment it.
trinque: though hardware ones, and not
the kind normally meant
trinque: though a
topic I've been
thinking upon a lot is,
trigger warnings
to asciilifeform, mitigations
trinque: phf: dunno
that I'm making an argument for anything *else* either
shinohai is still
trying
to source a nice 2005
thinkpad
phf: well, we've already established
that running everything on 2005
thinkpads is one viable option, but
that's until supplies last
☟︎ trinque: seems
this runs
the risk of continuously being a refugee. "they broke my emacs, so I left for hemlock/climacs/etc,
then
they broke my X,
then
they broke my ..."
trinque: I'll be clinging
to a frozen emacs version and frozen pile of elisp
til valhalla.
a111: Logged on 2016-11-30 21:02 asciilifeform: it is, laugh or cry, an actual problem, and imho
the reason climacs was stillborn
phf:
http://btcbase.org/log/2016-11-30#1574893 << fwiw mit-scheme's editor edwin has an elisp emulation layer,
that reportedly can run gnus of some vintage. when i heard about climacs from beach i actually
though he was going
to do an elisp
translator
too. one option might be
to pickup edwin, but
that's an exercise for someone else entirely
☝︎ phf: i've been low-key reviving cmucl's hemlock, can't use portable hemlock nor climacs for
that matter, because neither have
terminal versions. needless
to say
the process is slow and painful. like right now i'm
trying
to figure out why scrolling
the buffer is slow slow you can see each individual line redrawing. also arrow keys don't work
☟︎ phf: breaks my heart all
that's been done
to emacs. it used
to have a strong culture of backwards compatibility. large packages would have compat files compat-19.el, compat-20.el, etc. where's now only latest and greatest work. used
to be very non-dwim, now every single package insist on some "smart input" dwimy interaction mode. but all
this complexity is broken,
things clash and interact with each other in all kinds of funny ways
trinque: at any rate, if I were
to invest
time in climacs,
that is
time stolen from learning osdev
to put a lisp interpreter on iron, which is
time stolen from learning
to fab silicon in a garage, which is
time stolen from ...
☟︎ trinque: there was a "10 cans of soda"
thread
trinque: I keep hoping for a g_l writeup on
the subj, but do not see a blog anywhere
trinque: is climacs usable
to
this degree?
☟︎ ben_vulpes: betcha
that's exactly what gabriel_laddel did
☟︎ a111: Logged on 2016-11-30 21:12 asciilifeform: re elisp, imho
the 'apocalyptic' scale of
the problem is overblown, if every d00d were
to rewrite
the few 100 of elisp
that he actually ~uses~, in climacs, or whereever, job would be done in a week.
mircea_popescu: reducing
the problem
to "if every dude ran
this script and
then picked multiple-choice options 5
times"
mircea_popescu: in principle
this should be automatable
tro a huge degree
mircea_popescu: muscle memory is
the
thinking man's worst enemy ; not least of all because it
tends
to win.