UmbralRaptor changed the topic of #kspacademia to: https://gist.github.com/pdn4kd/164b9b85435d87afbec0c3a7e69d3e6d | Dogs are cats. Spiders are cat interferometers. | Космизм сегодня! | Document well, for tomorrow you may get mauled by a ネコバス. | <UmbralRaptor> … one of the other grad students just compared me to nomal O_o | <ferram4> I shall beat my problems to death with an engineer.
<UmbralRaptor>
Hrm. Probably not power seriesTaylor.
* UmbralRaptor
had Griffiths in undergrad.
<UmbralRaptor>
Also certain I did things wrong. Somehow got simple harmonic motion out of a problem with friction. o_O
<UmbralRaptor>
Yes for part 1. For part 2, there was a weird coordinate transform.
<Ellied>
might be power series Taylor actually, there are rather a lot of power series problems in the book
<Ellied>
well I mean uh, a damped driven harmonic oscillator will still give you the same general x(t) as an undamped undriven system, right?
<UmbralRaptor>
If the damping and drive cancel out, anyway. otherwise…
<UmbralRaptor>
Brook Taylor FRS (18 August 1685 – 29 December 1731)
<UmbralRaptor>
Hrm
<Ellied>
I think you still get sinusoidal motion, right? Regardless of the driving frequency and resonant frequency of the system, the transients like beats and stuff will damp out and you'll end up with it just oscillating at the driving frequency with some amplitude
<Ellied>
Okay, not that Taylor. There's a color photo of him smiling on the back cover.
<UmbralRaptor>
In any case, the problem somehow involves a Lagrangian with no multipliers but friction o_O
<kmath>
<aallan> On a flight from Zurich to Singapore two @FlySWISS pilots film the reentry of #SoyuzMS04 into the atmosphere.… https://t.co/7WtfsRHldl
<SnoopJeDi>
if you feed it an iterable of ints, each int is used as the byte I think?
<egg>
(aka line feed)
<SnoopJeDi>
just looking at `help(bytes)`
<SnoopJeDi>
vs. bytes(int) which is that length of null bytes
<egg>
!u a
<Qboid>
U+0061 LATIN SMALL LETTER A (a)
<Ellied>
but like
<Ellied>
couldn't you get that with b"\x00"*10
<Ellied>
why doesn't the bytes() function convert types like other functions whose name is a type?
<egg>
Ellied: bytes([6*16+1]) should get you b"a" I guess?
<SnoopJeDi>
Ellied, I don't understand the question, what do you mean?
<SnoopJeDi>
(bytes is a class, not a function)
<egg>
Ellied: well [10] is "ints", conceptually
<egg>
Ellied: so it does, it turns that to "bytes"
<Ellied>
I mean like if you go str(10) you get "10"
<SnoopJeDi>
b"..." is a bytestring literal in the same way that you can have 3.1415 or float("3.1415")
<egg>
Ellied: but 10 is a single integer, not a list thereof
<SnoopJeDi>
hmm, I would assume str() and repr() give the same thing for bytes objects but never really used them
<egg>
SnoopJeDi: I don't think that's SnoopJeDi's question, the literal is clear enough
<egg>
um
<Ellied>
str(b'a') produces "b'a'"
<egg>
s/second SnoopJeDi/greening/ probably
<egg>
Ellied: str is a bit of an oddity
<egg>
Ellied: but bytes *does* convert something that is a bunch of integers to something that is a bunch of bytes, right?
<SnoopJeDi>
if the arg is an iterable, yes
<SnoopJeDi>
there are 5 different init cases given in the doc
<egg>
e.g. [0x61, 0x62] to b"ab"
<egg>
Ellied: now if you give it a single integer, that's a different matter, it gives you a bunch of 0 bytes of that length
<Ellied>
I want to do bitwise operations on strings (specifically I'm doing simple one-time pad encryption and want to do XOR) and Python will only do bitwise anything on ints, apparently, so I'm using ord() on each character to convert it to an int and then bytes([]) to convert the resulting int back to a bytestring
<Ellied>
I rather suspect that if I said that line anywhere on freenode I would be told "you're not intelligent enough to be using computers" and ignored, but I have no idea how you're "supposed" to do this
<SnoopJeDi>
(#python on freenode is pretty friendly IME so I'd encourage you to poke around there sometime)
<SnoopJeDi>
((uh just don't ask about threads))
Technicalfool_ has quit [Read error: Connection reset by peer]
Technicalfool_ has joined #kspacademia
<Ellied>
hheh
<soundnfury>
Ellied: map(chr, [a ^ b for (a,b) in zip(map(ord, sa), map(ord, sb))]) # or something like that?
<egg>
Ellied: so bytes(list of ints) seems like the right conversion in that direction, not sure about the other; not sure about xoring bytes either
<egg>
Ellied: you start with a bytestring, not a string, right?
<soundnfury>
alternatively, write the thing in C because this is totally the sort of low-level bit-grinding that C excels at, and call it through ctypes
<Ellied>
yeah, I'm keeping everything in bytes as much as I can since I'm dealing with a randomly-generated key which isn't a valid plain text file apparently
<soundnfury>
Ellied: yeah, if you're using py3k, it's fascistisch about "strings are unicodes, not bytes" :(
<soundnfury>
Ellied: this also reminds me of the most useless function in glibc: man 3 memfrob
<kmath>
<whitequark> #catcognition pretty sure our cat is teaching our kitten tactics for hunting moths and ambushing does this mean cats have culture
<whitequark>
yeah the kitten borrowed quite a bit of behavior from the cat
<egg>
does it also open the door?
<SnoopJeDi>
follow-up: did UmbralRaptor teach it?
<egg>
(also if is my python advice to Ellied sound? I don't actually know much python, at work I have people who do to check that I'm not submitting daft things)
<UmbralRaptor>
ping
* egg
sits next to the pil author
<egg>
!wpn UmbralRaptor
* Qboid
gives UmbralRaptor a covariant inductor
<Iskierka>
!u ?
<Qboid>
U+1F47B GHOST (?)
<egg>
!wpn Iskierka
* Qboid
gives Iskierka an identity Bekenstein rotor-like octagon
<egg>
!wpn whitequark
* Qboid
gives whitequark a Riemann extinguisher with a timer attachment
<UmbralRaptor>
Uh, I can't usefully answer, since strings of Unicode points instead of bytes actually is closer to what I want >_>;;
* UmbralRaptor
also needs to better deal with numpy.genfromtxt() creating columns with b'foobarbaz' when I just want a string. >_>;;
<whitequark>
egg: not an outdoor kitten
<whitequark>
so no desire for that
<egg>
I thought printer had found it outdoor?
<egg>
(also painted green)
<whitequark>
"if you're using py3k" if you're living in the current century you should be using py3, soundnfury
<SnoopJeDi>
RIP legacy industry
<whitequark>
egg: as a very tiny kitten who could barely navigate in space
<egg>
aha
<egg>
whitequark: is it still somewhat smol though
<soundnfury>
whitequark: umm, no. py3 is not a new version of python, it's a _different language_, and for some tasks (systems-programmingy things like a kernel guy like me does) py2 is just Better.
<soundnfury>
also I have to py2 at work because we use Twisted which doesn't support 3 yet
* Iskierka
has no idea what the just py is. Does use pypy for speed, but pypy3 doesn't work, so is on 2
<whitequark>
soundnfury: that's a wild misrepresentation of what py3 is that borders on editorializing
<soundnfury>
whitequark: sure, that was my opinion
<SnoopJeDi>
Iskierka, people use py2/py3 as shorthand for Python 2/3 (the languages, rather than implementations thereof like CPython, PyPy, etc.)
<whitequark>
it's not a different language if you can write code that works on both versions quite easily.
<whitequark>
I didn't know about Twisted though
<whitequark>
that seems.
<whitequark>
weird.
<SnoopJeDi>
i.e. PyPy and CPython both implement the Python language, so in that sense they are "the same." (In very many other ways they are not)
<soundnfury>
whitequark: you _can_; but idiomatic py2 doesn't work on py3, and vice-versa
<whitequark>
isn't Twisted like a hugely popular library
<SnoopJeDi>
It is
<soundnfury>
yeah
<SnoopJeDi>
but in the same way, so is six
<SnoopJeDi>
six alone is proof that py2 isn't going anywhere, to my liking
<SnoopJeDi>
(which is not to say I think it will become less important come 2020)
<soundnfury>
whitequark: for instance, if you want to write systems-y stuff that won't choke on malformed-UTF inputs, in py2 it's easy because a string is a bytes. In py3 you have to b'' everything, and go to great lengths to work around the "oh I'll be helpful and give you a Unicode type" stuff
<soundnfury>
heck, for the first few major versions of py3k it was Impossible, because bytes didn't support %
<SnoopJeDi>
whoa, really? Was it all str.format(), then?
<soundnfury>
if you want a memory-safe C with an object system, py2 is great for that. py3... less so.
<whitequark>
it's true that py3 has a few instances where it's hard to avoid erroring out on invalid UTF
<whitequark>
I think there was something about environment variables and stdio
<soundnfury>
SnoopJeDi: I think bytes didn't have .format() either initially
<SnoopJeDi>
It took me a few days to come around philosophically on f-strings, but I bet if I was using 3.6 it'd take a lot less
<egg>
!wpn whitequark
* Qboid
gives whitequark an isothermic raptor
<whitequark>
we're using "async def" a lot
<whitequark>
whatever python calls that thing, is it coroutines?
<whitequark>
i mean, it's coroutines, but i'm not sure how python calls it
<SnoopJeDi>
I think it calls them coroutines yea
<soundnfury>
anyway, my hatred of py3k is because this <http://www.catb.org/esr/faqs/practical-python-porting/> is far longer than it should need to be, and requires doing crazy things like lying to Python and telling it "all of my data is Latin-1"
<whitequark>
that's kind of absurd
<whitequark>
all of your stdio is *not* in latin-1
<whitequark>
unless you ignore anything that's outside the borders of the US, your tools *will* come in contact with utf-8 input
<whitequark>
what this porting guide *should* have done is three things
<soundnfury>
whitequark: readability of code matters, b" prefixes are bad for that.
<whitequark>
yeah, python is already unreadable, you can't make it worse
<soundnfury>
and it was written before that PEP went in, and not having b''% is a real problem
<whitequark>
2. you can just monkey-patch % in
<whitequark>
3. you should re-open stdio streams with a binary encoding
<whitequark>
not latin-1
<soundnfury>
whitequark: or you could just use py2 and not deal with this shit, because py2 is already the language you want to be using (i.e. strings are bytes without you having to say so)
<soundnfury>
and maybe consider throwing money at someone if they offer to support py2 after the PSF EOLs it for realz
<whitequark>
and make your stuff incompatible with everyone else, who has moved on years ago, sure
<whitequark>
have you considered m4
icefire has quit [Ping timeout: 183 seconds]
<whitequark>
I heard it has good formatting capabilities
<soundnfury>
I think comparing a language to m4 should be the equivalent of Godwin's Law
<whitequark>
hey, most of FOSSland still actively uses that.
<soundnfury>
well, for values of "run a canned autobarf recipe thrice plagiarised once removed"
<soundnfury>
and the _good_ projects keep themselves small enough that Makefiles (plus maybe a hand-written configure script for a handful — no more than that — of platform-dependent feature #defines) will do
<whitequark>
lol good luck with that if you do any real work
* whitequark
glares in the general direction of ANGLE
<soundnfury>
whitequark: the frigging *Linux kernel* can be built with make. If your project can't, don't go claiming the problem's on make's side.
<whitequark>
what
<whitequark>
how is this a sensible argument?
<whitequark>
the linux kernel has an extremely conservative environment it builds in.
<soundnfury>
ummm, in that Linux is quite possibly one of the biggest and most complex software projects in existence?
<soundnfury>
and is _hugely_ configurable (through its own 'Kconfig' system)
<whitequark>
it's not particularly complex, but that's not my point
<soundnfury>
whitequark: ahahahahahaha.
<whitequark>
i know how linux works, thanks. i did a linux port or two.
<whitequark>
anyway, my point is that linux only concerns itself with building on a not-too-old gcc and binutils
<whitequark>
i'm not even really counting clang because most of the clang effort was making clang more like gcc
<whitequark>
and the rest was eradicating especially esoteric code patterns from linux
<soundnfury>
uhh, the "not-too-old" is only there because gcc keeps _breaking_ things that a kernel _needs_ to do
<soundnfury>
and the gcc requirement is because, hmm, gcc-style inline asms.
<soundnfury>
(which, again, good luck writing a mdoern kernel without)
<whitequark>
a) gcc doesn't break things, things were already broken in kernel and gcc just exposed that
<whitequark>
(maybe don't write a kernel in C too)
<soundnfury>
whitequark: lolno.
* soundnfury
searches for relevant linus-rant-collection…
<whitequark>
i've read all linus rants and i don't give a fuck
<whitequark>
he's barely competent to develop a kernel
<soundnfury>
money quote: "There have been situations where documented gcc semantics changed, and instead of saying "sorry", the gcc people changed the documentation."
<soundnfury>
whitequark: ok, let's see your better kernel :P
<soundnfury>
anyone claiming Linus is "barely competent" has a _massive_ burden of proof
<whitequark>
fuzz it for a hour and you have a local privesc
<whitequark>
all the shit that should never be inside the kernel
<whitequark>
fucking *x509 parsing*
<whitequark>
who signed off on that? an idiot, that's who
<whitequark>
the disaster that is seccomp
<whitequark>
jit in kernelspace? a great idea
<whitequark>
kaslr is incredibly broken
<whitequark>
this is just off the top of my head
<soundnfury>
whitequark: wait, you're speaking ill of the ebpf+jit bits?
<whitequark>
yes
<soundnfury>
now I'm _really_ angry :P
<soundnfury>
and, again, let's see your better kernel
<whitequark>
that's not an argument
<soundnfury>
oh and better make it popular enough that people write drivers for it, because drivers are the bulk of the work in an OS kernel in practice
<whitequark>
i don't have to be a chef to point out that your pizza is burnt
<whitequark>
but i dunno, xnu is okay at times
<soundnfury>
Sure it is. You're claiming he did a bad job. That implies you think a better job could be done. So why hasn't anyone?
<whitequark>
you're *still* missing my point about buildsystems
<whitequark>
linux builds in 1 (one) build system configuration
<soundnfury>
whitequark: how many arches?
<whitequark>
that doesn't matter for its makefiles. it always cross-compiles itself, which is at least one thing it gets right
<whitequark>
compare this to, say, solvespace
<soundnfury>
for userland code, the solution is to _write to standards_. When autobarf was invented, that wasn't an option because no-one followed the standards and it wasn't even clear which standard to choose
<whitequark>
which has to build under macos->macos, linux->linux, linux->windows via mingw and windows->windows under both mingw and msvc
<whitequark>
have i mentioned the vendored geometry libraries we track? and ANGLE, because OpenGL is one of the worst disasters in the history of APIs?
<whitequark>
I literally wish I could use mesa's DirectX interface but no, that leaves NVidia out
<whitequark>
and solvespace... is a tiny project
<whitequark>
i don't remember if it's 20k or 80k lines of our own code but either is so insignificant
<egg>
whitequark: that's about the size of principia, which is kinda scary about principia
<whitequark>
keeping it all together and building and testing green on all of the platforms is a massive task though
<soundnfury>
I'm afraid to me that sounds isomorphic to someone wrestling with npm and react.js and angular and similar crap
<soundnfury>
i.e. your ecosystem is fucked
<soundnfury>
and it fucks your code and build and deployment for you
<whitequark>
40k lines, actually
<whitequark>
soundnfury: i don't give a fuck
<whitequark>
i write for people, not for machines
<soundnfury>
and that's not a problem that wrapping autobarf around it will really solve
<whitequark>
if people are on windows, i will deploy to windows
<whitequark>
if people are on chinese inside out fork of android i'll deploy to that
<whitequark>
btw, cmake lets me manage it in at least somewhat sane way
<soundnfury>
and you're free to suffer the pain that results from that.
<soundnfury>
personally, I prefer to write for _sane_ targets and tell people whose systems aren't sane to fix their broken insane systems and not come crying to me about it
<whitequark>
sane target, n.: whichever target is the most convenient for the wanker using the expression
<whitequark>
feel free to do that, and preface all your opinions with "I know I don't do anything useful, but"
* soundnfury
reminds the honourable gentlemen that I maintain the driver for the network card that runs the world's financial exchanges
<soundnfury>
the honourable gentleman is of course entitled to his own opinions on how useful that is
<whitequark>
i'm not a gentleman (or a man), and providing infrastructure for capitalism sounds rather counterproductive, yes
<soundnfury>
oooookay, I'll just file you under "crypto-marxist" (or possibly without the crypto- bit) and move on
<whitequark>
marxist? not by a long shot
<whitequark>
not in a place that was steamrolled over by marxist-leninists, uh, around 100 years ago, will it be, and will likely never recover
<soundnfury>
so I basically file people into two buckets: "believes mammon is moloch" and "believes mammon is the solution to moloch"
<soundnfury>
and "providing infrastructure for capitalism sounds rather counterproductive" sounds a lot more like the former than the latter
<whitequark>
about as much social insight as linus has. no wonder you like him so much
<soundnfury>
or maybe it's because we both have insight that everyone else buries under a ton of monkey-politics bullshit
<soundnfury>
(which is to say, I like linus so much because he's an aspie, and aspies are the only people whose mental workings I even slightly understand and can empathise with. Y'all neurotypicals are just _inexplicably weird_.)
<whitequark>
I'm autistic. I even have a formal diagnosis!
<whitequark>
not that it wouldn't be obvious from spending more than one minute with me
<whitequark>
interestingly, that doesn't prevent me from, hmmm, considering any other point of view other than the one that my ass comfortably sits on right now
<whitequark>
and i'm quite tired of people using "aspie" as an excuse for being an asshole
<whitequark>
go make some effort and then return
<soundnfury>
whitequark: then don't criticise me for lacking social insight, because you haven't got any either
<soundnfury>
(btw, I have a formal diagnosis too. If that matters.)
<whitequark>
that's not how it works
<whitequark>
you don't need to be able to imagine shapes to be great at geometry
<soundnfury>
one of the main defining traits of the autistic spectrum is the systematising-empathising axis.
<whitequark>
you don't need *any* empathy to recognize how others perceive the world
<soundnfury>
and you're basically criticising me for being overly systematising and insufficiently empathising
<whitequark>
nope
<whitequark>
i'm criticizing you for using an oversimplified model, if you will.
<soundnfury>
in that I consider the distinction I named above, between, as I see it, marxism and freedom, as the only *relevant* distinction in political philosophy
* whitequark
snorts
<whitequark>
"marxism and freedom"
<whitequark>
this is my new favorite dumbest political idea
<soundnfury>
whitequark: hey, I'm an an-cap. I'm not going to apologise for that.
<whitequark>
this is a dumb idea *even coming from an ancap*
<whitequark>
there are ancaps I respect!
<soundnfury>
well I'm afraid I find it hard to respect someone who says "providing infrastructure for capitalism sounds rather counterproductive"
<soundnfury>
I can't conceive of any non-dumb interpretation of that statement.
<whitequark>
at best, capitalism is a necessary evil. being proud of your work specifically because it supports infra of capitalism is an extremely weird position.
<whitequark>
for example, i provide infrastructure for quantum physics research.
<whitequark>
i heard my tcp/ip stack is pretty good
icefire has joined #kspacademia
<soundnfury>
whitequark: capitalism is _precisely_ the system which enables humans to pursue their (diverse) ends
<soundnfury>
there's nothing evil about it
<soundnfury>
without capitalism, there'd be no quantum physics research, because you'd never manage enough capital formation for the equipment needed
<whitequark>
ever heard of Lev Landau?
<soundnfury>
also, because farmers don't have much need of quantum physics papers
<whitequark>
I'll be the last person to *defend* USSR, but your statement is factually completely wrong
<whitequark>
even under feudalism, research proceeded apace, funded or done by local nobility
* egg
tosses soundnfury into a бкл singularity
* SnoopJeDi
glances at scrollback
<soundnfury>
whitequark: feudalism never formed enough capital for a particle accelerator (so research yes, quantum physics research no)
<Iskierka>
Most of the best research now is done by govt labs. Capitalism tends to go for the easy improvements, like simple scale-up
<Iskierka>
revolutions don't tend to come from the free market
<whitequark>
that's also an absurd statement, a cyclotron is extremely cheap
<soundnfury>
whitequark: only because you have access to a heck of a lot of materials that are only that cheap because of wide-ranging trade and specialisation
<whitequark>
have you ever looked into building a cyclotron?
<whitequark>
or early physics research in particular
<soundnfury>
how cheap do you think you could make a cyclotron in 1800?
<Iskierka>
... I'm thinking the most expensive part would be the steam engine & generator powering it
<soundnfury>
maybe you could do any individual bit of the labour yourself and it wouldn't be _too_ hard. But to do _all_ of it would be a herculæan task.
<Iskierka>
you could build early cyclotrons by hand
<whitequark>
^
<whitequark>
a cyclotron is a desktop device. you need some magnets, some copper wire, a vacuum vessel, a pump, a lot of glasswork
<whitequark>
this is not something that alessandro volta would be fundamentally unable to build
<soundnfury>
Iskierka: ... with modern metallurgy
<whitequark>
what fucking metallurgy
<soundnfury>
even the vacuum would be difficult
<whitequark>
difficult?!
<whitequark>
have you seen a vacuum pump?
* Iskierka
checks. First (partial) vacuum chamber: 1671
<whitequark>
start with a hand-cranked rotary vane pump, then remove the rest of the gas using a mercury pump
<whitequark>
needed: a glass tube (1), mercury (a few grams)
<whitequark>
run this shit for several weeks and you have a vacuum good enough you can do fucking triodes
<soundnfury>
the glass-blowing would be more expensive and less precise, the mercury would be more expensive…
<whitequark>
...
<whitequark>
what
<whitequark>
you can refine mercury yourself if you make a trip to the nearest cinnabar deposit and then light a fire with some dead trees
* egg
bashes head against laptop
<whitequark>
glassblowing?
<Iskierka>
If you explained it to them I'm not sure a turbomolecular pump would be beyond them. Just difficult of how do you apply power to it (crazy gearbox direct on steam engine?)
<whitequark>
ever heard of "alchemists"?
<egg>
or Venice
<whitequark>
those guys were fucking *rad* with glass
<soundnfury>
just think in terms of opportunity cost. When you start measuring everything in (say) median wages, all this shit is _insanely_ expensive in 1800
<whitequark>
in fact modern glass is *shit* compared to what was done in 1800s and 1900s
<whitequark>
no it isn't
<whitequark>
the hard part you needed for a cyclotron is electromagnetism
<whitequark>
and some luck
<soundnfury>
I'm not saying you _can't_ build a cyclotron. I'm saying you can't build it _cheaply_
<whitequark>
actually building the thing is extremely easy
<whitequark>
do I *have* to do a "primitive technology" style video building a cyclotron?
<whitequark>
a cyclotron is *way* easier to build than the Antikythera mechanism
<whitequark>
and the latter predates feudalism.
<soundnfury>
whitequark: but the skill base for building the Antikythera mechanism already exists in clockworks. The same isn't true of the kind of engineering you need for a cyclotron
<whitequark>
it historically is
<soundnfury>
which is why you can build a cyclotron in 1800, but you can't build it *cheaply*.
<whitequark>
drop James Clerk Maxwell into 1800, and he'll get it done in a few years
<Iskierka>
... I'm not aware of mechanical clocks existing at the time of the Antikythera mechanism?
<soundnfury>
(you're still not thinking economically, Marty!)
<Iskierka>
like I'm pretty sure that's close to a one-off
<whitequark>
Iskierka: Antikythera? I think modern consensus is that it was relatively mass produced
<whitequark>
soundnfury: glasswork: just the usual stuff chemists did casually
<Iskierka>
Hm, that's a surprise, though I meant more as the design. I can't think of many (any?) other mechanisms of such a type that old
<whitequark>
it is actually really not a hard skill to acquire, mostly requires patience
<Iskierka>
certainly none nearing that complexity making it quite an anomaly
<whitequark>
wires: lol
<Iskierka>
meaning we could easily have a cyclotron anomaly in 1800
<soundnfury>
(you're still not thinking economically, Marty!)
<whitequark>
magnets: probably would need to use electromagnets
<whitequark>
you said, direct quote, feudalism never formed enough capital for a particle accelerator
<whitequark>
but you don't actually need a lot of capital for a particle accelerator.
<whitequark>
a single person with meager supplies can build it over a few years starting from *ore*.
<SnoopJeDi>
"Can you build a capacitor? Bam, particle accelerator."
<whitequark>
and you wouldn't need that, you could acquire a lot of what you needed elsewhere
<whitequark>
SnoopJeDi: an X-ray tube powered by static electricity, maybe
<Iskierka>
Copper wiring was exceedingly common since the 1820s (from formal creation of electromagnet and telegraph). We could obtain it reasonably easily in 1800 with a purpose. Mercury displacement pump was 1855 but we could probably push that anachronism ourselves also
<whitequark>
wimshurst machine, I mean.
<SnoopJeDi>
whitequark, yep, sure
<whitequark>
this is like... you could do it in a few *days*
<egg>
!wpn -add:wpn cyclotron
<Qboid>
egg: Weapon added!
<whitequark>
anyway, regardless of that, almost all basic research today is funded by federal and supernational labs of various levels
<whitequark>
pretty marxist. zero freedom for sure.
<whitequark>
we're taking money from the USG DoE, which is like an epitome of central planning
<whitequark>
they mostly do it because the army wants to break ciphers or something like that
<whitequark>
NIH is... well... what's the word most opposite from "anarchy"?
<whitequark>
rhetorical question.
<soundnfury>
you know, shortly before I got into arguments here, I got into a different argument in a channel that's _not_ full of academics. They were _much_ more willing to entertain the notion that economics has its own laws that they don't yet understand
<soundnfury>
I'm beginning to understand why academia leans _so fucking far_ left.
<whitequark>
i'm not an academic nor do i claim to understand every law of economics
<whitequark>
i don't have a degree.
<soundnfury>
"I don't need to understand economics, because I know how to build a cyclotron, therefore capitalism is unnecessary to support my research"
<Iskierka>
I understand that the free market will stifle any competition, because that's easier than improving yourself
<soundnfury>
Iskierka: s/the free market/a government
<Qboid>
soundnfury thinks Iskierka meant to say: I understand that a government will stifle any competition, because that's easier than improving yourself
<FluffyFoxeh>
hm doesn't the government actually enforce competition
<whitequark>
I demonstrated that one of your factual claims is false. you responded with a complete misrepresentation of everything I said, probably because you never actually tried to understand it
<Iskierka>
It does
<whitequark>
FluffyFoxeh: in theory
<soundnfury>
(hint: corporations don't have the police power. Even private protection agencies ('private police') don't have the police power.
<Iskierka>
and a government wants improvement so it gets voted for
<whitequark>
soundnfury: they do in today's UK. serco and g4s.
<FluffyFoxeh>
a free market doesn't work without competition. it's also in a firm's best interest not to have any competition. something has to reconcile those things
<Iskierka>
Hint: the US is very far from a free market anyway
<Iskierka>
see: somalia for better example
<whitequark>
soundnfury: they did historically. british east indian company or w/e
<whitequark>
FluffyFoxeh: in practice no one actually enforces antimonopoly laws
<whitequark>
or we'd have google and arm broken up
<soundnfury>
whitequark: agents of the gov't, acting in the gov't's name. That is not what I meant and you know it
<whitequark>
soundnfury: no, actually, I don't know what you meant
<whitequark>
they're not accountable to people whose taxes pay their salaries
<soundnfury>
whitequark: no corporation is able to stifle competition / exert a police power / violate people's liberties _except_ by co-opting or acting for the government
<whitequark>
of course not
<FluffyFoxeh>
I believe they are enforced, it just hasn't happened (yet?) in those cases. Intel and Microsoft had the hammer dropped on them, IIRC
<Iskierka>
also why talk about the police anyway when the market offers much better ways to destroy startups. Like just temporarily undercutting them on price, bribing staff, or outright buying them or the company
<soundnfury>
Iskierka: now learn some economics
<whitequark>
a corporation can use its domination of one market, perhaps a market where it was a player when it emerged, to strongarm another one
<Iskierka>
antimonopoly laws tend to be saved until the monopoly is being more actively evil as it's expensive to fight monopolies
<Iskierka>
soundnfury, maybe you learn some history
<soundnfury>
Iskierka: actually, antimonopoly laws are mostly used to prop up cartels
<whitequark>
for example: a [certain IP core maker] bundling [certain CPU cores] with [certain GPU cores]
<whitequark>
as a result, it essentially killed all other GPU core vendors
<Iskierka>
Feudalism was in fact pretty much the definition of free market. Or at least what happens to it, because power consolidates
<whitequark>
it's simply unprofitable anyway
<soundnfury>
Iskierka: no.
<Iskierka>
History disagrees with you.
<Iskierka>
Literally all of it.
<whitequark>
do you expect ancaps to learn history or what
<soundnfury>
Power only consolidates if most people think a consolidated coercive power is _just fine and dandy_ because mumble divine right of kings (or mumble will of the democratic majority)
<Iskierka>
Power consolidates because how are you going to fight the people with twenty times more guns/swords/bows/slings/sticks than you?
<Iskierka>
who stick together because they're being given a good enough deal and if they try break off they're suddenly the one with twenty times more aimed at them
<soundnfury>
Iskierka: all that proves is that a near-hegemony becomes a hegemony
<Iskierka>
It only takes a small advantage to bully a few more to join you and snowball from almost nothing, in a total power vacuum, to being the only power
<soundnfury>
if people start by thinking that coercion is _not okay_, where're you gonna find enough people to carry those guns?
<Iskierka>
outside of a power vacuum, this snowball effect fails
<whitequark>
it's honestly quite amazing how posting links to esr, linux zealotry, ancap adherence and lack of reading comprehension tends to come together in a nice package, over and over again
<Iskierka>
By paying them
awang has quit [Ping timeout: 200 seconds]
<Iskierka>
Which is part of the fundamental problem: monetary value is the only thing that matters and the one in power has all of it
<soundnfury>
whitequark: maybe because I have a coherent philosophy and just because it's different to yours in several places (which naturally are the places we end up arguing about) doesn't mean that all of those places are definitionally insane
<soundnfury>
Iskierka: no, because you can only buy things at a price someone is willing to sell
<Iskierka>
People are very willing to sell themselves at the price of "living comfortably"
<Iskierka>
which is easy to give if you have all the money
<soundnfury>
and if someone starts acting all imperial, people who are the an-cap-society equivalent of civic-minded will _not_ interview to be hired in that someone's army.
<Iskierka>
and then you can bully everyone else to reduce costs of living so you can pay minimally
* whitequark
giggles
<soundnfury>
because there's money _enough_ for comfort outside the tyrant's sphere, and if you go and join the tyrant's army soon there won't be
<whitequark>
i just realized something i can do about this discussion of power
soundnfury was banned on #kspacademia by whitequark [*!*edward@*.5-4.cable.virginm.net]
soundnfury was kicked from #kspacademia by whitequark [soundnfury]
<FluffyFoxeh>
oh yeah, kick people that disagree with you
<FluffyFoxeh>
great
* whitequark
shrugs
<whitequark>
consolidation!
<egg>
heh
<Iskierka>
because clearly such a large proportion of people in the modern world care about more than the next election cycle, anyway
<UmbralRaptor>
To be fair, soundfury apparently doesn't realize how much money goes into rent, food, etc.
<Iskierka>
and those who do will organise coherently enough to fight those off
<whitequark>
FluffyFoxeh: I was going to /ignore them, but this seemed more amusing
* Iskierka
doesn't actually know what channels soundnfury is on
<egg>
yeah this seemed eggstremely appropriate given the discussion :-p
soundnfury has joined #kspacademia
awang has joined #kspacademia
<egg>
whitequark: also, I join Iskierka in saying that I'd be entertained in a video of cyclotron-building :-p
<egg>
um, interested, or entertained by, /me cannot into words at this time apparently
<Greys>
the grammar works if you are both present in the video, and entertainded while being so
<whitequark>
egg: I soooooorta want to do it but I'm like five months behind on my backlog
<whitequark>
because of (now-solved) mental health problems
<SnoopJeDi>
...also wow I just realized I know one of them
<whitequark>
oh NICE
<whitequark>
that's even easier than I expected
<SnoopJeDi>
yea and it's a "real" cyclotron
<whitequark>
like this is not really the way I'd build it in 1800 but
<SnoopJeDi>
something a la Lawrence wouldn't be super useful today, but 2 MeV isn't a pea-shooter
<SnoopJeDi>
discussed Lawrence's cyclotron with some local-cyclotron people over the weekend while camping, took a stab in the dark about the accelerating voltage for Lawrence's and happily was about right: ~ keV gaps
<SnoopJeDi>
err, kV*
<whitequark>
"The first cyclotron was a pie-shaped concoction of glass, sealing wax, and bronze. A kitchen chair and a wire-coiled clothes tree were also enlisted to make the device work."
* whitequark
stabs soundnfury in the capital
<whitequark>
that's actually even more ghetto than I remember from first reading about it
* soundnfury
stabs whitequark in the still-doesn't-understand-economics
<SnoopJeDi>
it really is a cute device
<soundnfury>
in fact, I'll just go back to my previous elucidation:
<soundnfury>
you're still not thinking economically, Marty!
<whitequark>
you don't even need metal-to-vacuum seals (we use kovar for that today, but I think kovar is recent)
<Iskierka>
That seems pretty economical to me
<whitequark>
fucking *wax* works
<whitequark>
metal-to-glass even
<SnoopJeDi>
well, not all vacuums are created equal, of course
<whitequark>
SnoopJeDi: sure but a cyclotron needs high vacuum
<SnoopJeDi>
but yea, as a "make particles go" goal, you can get there well enough
<whitequark>
mean free path larger than the device
<whitequark>
(at least)
<SnoopJeDi>
for generous values of "high" I guess? my internal ruler for vacuums is ruined by thinking about hadron colliders
<SnoopJeDi>
well, any recirculating machine*
<SnoopJeDi>
not having any horizontal aperture makes it a lot more forgiving
<whitequark>
10^-3 torr
<SnoopJeDi>
(unless you're PSI and literally overlap 3 of your turns >_>)
<whitequark>
is what I'd call high
<SnoopJeDi>
oh, hrm, I didn't realize "high" spanned 6 orders of magnitude, I stand corrected
<whitequark>
yeah I'm not talking about UHV
<SnoopJeDi>
I thought it was ~ 1e-6 to 1e-9 and there was something else in there
<whitequark>
you don't need a turbopump for a cyclotron though
<SnoopJeDi>
yea nothing very exotic
* egg
is confused by pressure units
<egg>
[1 mTorr = 133 mPa = 1.3 μbar for those similarly confused]
<Iskierka>
...
<whitequark>
soundnfury: a lawrence cyclotron is generally only slightly more advanced than contemporarily existing equipment
<whitequark>
so no
<whitequark>
I am thinking economically
<SnoopJeDi>
egg, the Pascal is less than practical :P
<whitequark>
I've already explained above how you don't need any complicated supply chain
<egg>
SnoopJeDi: I know, I'm just confused
<egg>
SnoopJeDi: also don't you say phl is impractical :-p
<SnoopJeDi>
(determining whether I mean the SI unit or the programming language is left as an exercise to the reader)
<whitequark>
in fact, after looking at it more in depth, it's *even cheaper than I thought*
<egg>
SnoopJeDi: or my father
<SnoopJeDi>
is his name Pascal or something egg I'm missing an inferential link here
<soundnfury>
contemporarily existing equipment has uses applicable to contemporarily existing problems, so "tooling" (for want of a better word) exists
* Iskierka
was not expecting the allcaps for some reason
<whitequark>
tooling for making a lawrence cyclotron has existed in 1800s
<whitequark>
most of it existed in *1600s*
<SnoopJeDi>
but not exactly the E/M to make it go-go
<whitequark>
to be specifc, a lawrence cyclotron could be made by a clockmaker significantly easier than a reasonably precise clock
<whitequark>
!u
<Qboid>
What code point do you want me to look up?
<whitequark>
!u BIRD?
<Qboid>
U+0042 LATIN CAPITAL LETTER B (B)
<Qboid>
U+0049 LATIN CAPITAL LETTER I (I)
<Qboid>
U+0052 LATIN CAPITAL LETTER R (R)
<Qboid>
U+0044 LATIN CAPITAL LETTER D (D)
<Qboid>
U+003F QUESTION MARK (?)
<whitequark>
oh
<egg>
yeah it has no query by name afaik
<egg>
it has query by codepoint,
<Iskierka>
!help
<Qboid>
Iskierka: I sent you a private message with information about all my commands!
<egg>
!u u+a0
<Qboid>
U+00A0 NO-BREAK SPACE ( )
<SnoopJeDi>
Interesting aside: Maxwell wrote a paper that used Vlasov sorts of dynamics to show Saturn's rings were neither solid nor a continuous fluid
<SnoopJeDi>
which is more than sophisticated-enough dynamics to make a strong focusing machine, even
<Iskierka>
That is curious as aerospace teaches that the continuum approximation should mean that you can't tell the difference between continuous fluid and *lots* of particles
<Iskierka>
or at least that each can be modelled as the other, possibly
<SnoopJeDi>
Iskierka, indeed this is what Maxwell basically does: he shows that in the continuum case there are instabilities.
<Iskierka>
Plasma I presume would still also be different equations than turning collision-only particles into a continuum model
<Iskierka>
(low-velocity-collision-only)
<SnoopJeDi>
oh, yes, interaction is very different
<SnoopJeDi>
although electron cooling is kind of a funny duck in that following the work of Derbenev & Skrinsky looks an awful lot like scattering problems you work in any grad mechanics course, and you Play Games™ with choosing the impact parameter to use in a continuum model