62900+ entries in 0.033s
a111: Logged on 2018-09-14 16:27 asciilifeform:
to round out
thread,
http://btcbase.org/log/2018-09-14#1850330 << in my orig udp attempt, i dispensed with
the
traditional 'gotta be in network byte order (tm)' doctrine, in favour of 'if packet doesn't pass muster,
THEN flip
the endianism and
try again 1ce'
a111: Logged on 2018-09-14 17:09 mircea_popescu: iirc we even had a discussion re standardizing a byte order ;
though from
the fact
that i don't recall
the results i
take it i got shown broken cats.
a111: Logged on 2018-11-10 16:56 asciilifeform: also how come you need procedure Cast_LE( LE: in out Raw_Types.Octets ) ? i
thought
the algo was 'if message invalid, flip endianism in whole
thing and re-eval before discard'
mircea_popescu: Mocky_ incidentally, now
that you've rested, spending a few weeks publishing fleshouts / completing
things / etc will be invaluable.
mircea_popescu: "do it like
this because smarter people
than you sat down once and figured it ALL out, and go read logs if you have issues"
mircea_popescu: moreover,
THIS is what a language.io stdlib should even fucking be in
the first place.
a111: Logged on 2018-01-04 17:38 mircea_popescu: anyway, absent
this logical rule, we'll do what, add "wordwrap(magicnumber" in
the unloved v style manual ?
diana_coman: mircea_popescu, myeah, whether
there IS a problem really or not quite
mircea_popescu: only question is if
this gains anything or just pushes some garbage downstream
☟︎ diana_coman: yes,
that's what I had in mind; and I
think it does solve
the problem because it effectively fixes
the size, yes
mircea_popescu: because whatever is
there extra is padding. so instead of returning 6 correct records and discarding 2 records' worth of padding, you return 8 records, and let
the application asking for records figure out
that "6" means it needn't process 7 and 8
diana_coman: the only other option is
to say it's always nmax*...
diana_coman: if it is described as a variable length i.e. n* ...
then it's variable, what can I do
diana_coman: in implementation it gets defined as such although it still gets represented on 8 bits because
thus specified; logically speaking it shouldn't be on 8 bits
to represent ...40 max so yes, it makes sense
to say, pack
mircea_popescu: the idea was
to define n as a ada-style var, "from 1
to 19"
diana_coman: if you have however n and
then max and
then whatever
then sure, yes
mircea_popescu: otoh i don't really want all
the structure at one side of message for crypto-shamanistic considerations.
diana_coman: if you have n and
then n*x and
then something else
diana_coman: so
that
the "not meaningful" is essentially padding
diana_coman: i.e. it should be
the last
thing in
the message
diana_coman: this is what I had in mind; it helps but: diana_coman> in which case fine, it can perhaps even work like
that for most messages except stuff
that has max>message size and/or stuff
that has meaningful data *after*
this variable part
mircea_popescu: diana_coman
the reason n is
there is
to
tell you how much data is useful. what if you fixed n
to max at depack size, and
then delivered n=max records
to application, and it is ITS job
to discard
the extras ?
diana_coman: in which case fine, it can perhaps even work like
that for most messages except stuff
that has max>message size and/or stuff
that has meaningful data *after*
this variable part
diana_coman: basically
the only way is
that: put always max and just write somewhere how much of
that is actually meaningful
diana_coman: for
the other I really don't see how exactly
to have at
the same
time flexible size (specifying n and
then n* IS flexible size) AND record'write
diana_coman: for one
thing
the solution as implemented is perfectly fine; sure, not a one-line record'write or something like
that but not a
terrible
thing either
diana_coman: mircea_popescu,
the problem (if
there is one) is not space in itself for sure
a111: Logged on 2016-07-19 14:07 mircea_popescu:
the basic, and really only, rule of hermeneutics is :
that
then you've understood a
text when, far from its shortcomings appearing inexplicable errors,
they become
the actual pillars upon which
the damned
thing is constructed, and what originally seemed
to you sensible and structural
takes its
true place as accidental.
mircea_popescu: then a) ada should allocate
this as a 4 bit value and b) ada allocating space for 15 records is entirely fine anyway. with a being a very minor point but b being a major point because we've reoriented ourselves
to where features of
the language are indistinct rather
than inconvenient.
this is very much
the
http://btcbase.org/log/2016-07-19#1505418 thing.
☝︎ diana_coman: at implementation
time
that is how I define
the counter for keyset for instance
mircea_popescu: then
the "it allocates whyole counter" becomes nonproblem.
mircea_popescu: point in case, defining variable in "bitsize". why am i not simply saying "this is a counter from 1
to 19" or w/e ?
diana_coman: I'm not sure what would
the bitwise
thing buy exactly?
mircea_popescu is not at all adverse
to permitting ada metasyntactic considerations leak into
the protocol.
there's no rule specwork gotta happen in meta-c.
a111: Logged on 2018-11-10 16:43 mircea_popescu: diana_coman is
the structure actually problematic ?
mircea_popescu: "no fucker,
this isn't an "int",
this is a 15 bit counter"
mircea_popescu: diana_coman would it be smart if i defined
the count
types narrowly ? ie, bitwise ?
a111: Logged on 2018-11-10 16:52 asciilifeform: diana_coman: and while we're nitpicking, Serpent message
types can be an enumeration (see barnes)
diana_coman:
http://btcbase.org/log/2018-11-10#1870620 -nitpick away! I actually DID consider
that and I decided it's not really fitting because
think of it: if I make it enumeration
that *also* means you can actually go succ and prev which however isn't something
that makes sense for message
types; it's not like
they are an actual ordered
thing
☝︎ BingoBoingo: Could be
two years and
their new governor Gavin decides no water for anyone
BingoBoingo: Sure it'll
take
that long.
This is KKKalifornia
mircea_popescu: in another decade, "find out which celebrity hasn't had running water
this week"
mircea_popescu: because it's
totally "normal" for your fucking house
to burn, dumb piggy, and it's "an event" which "happens", and
there's no such
thing as agency in zeklands and so on.
mircea_popescu: asciilifeform you can define hooks for replacement. i suspect
this might be a
theme
thing. mine replaces gpg code with fixed format... apparently her
theme replaces :)
mircea_popescu: %
Total % Received % Xferd Average Speed
Time
Time
Time Current
deedbot: mircea_popescu updated rating of Mocky from 3
to 4 << ἀποκρισιάριος دولة قطر
Mocky: took
the bike out for a spin. had
to bundle up, bit of cold shock
to my system
Mocky: another 9 hours of sleep. now I feel like I'm back
to life
phf: asciilifeform: will look into it
this weekend
mircea_popescu: ill
tell you
this
tho -- only AFTER
this is made as
thin as it gets are we significantly more advanced on
the path
towards replacing php
mircea_popescu: actually,
this is
the correct direction -- should be collapsed into as few files as necessary (prolly
three, seeing how it uses
the apache access model
to limit admin/users, so you'd have an admin and a public dir), and
then cut from
that.
BingoBoingo: Can only ada
things up in steps, and
there's other moving parts underneath www-blogthing
mircea_popescu: this, amusingly, was on my
to-do (really,
to-have-done, but let's not split camel
toes) list for ... 12 years now!!!
mircea_popescu: in
the present situation i'd settle for a
three-file 5k loc mp-wp reduction from
the current imensity.