<cr1901_modern>
Aside, this is a cool solution for the "pc-rel addressing can't address all 4GB of memory", but it seems only riscv uses it. I wonder how other toolchains do it
<bubble_buster>
I think that might be what's going on
<sorear>
cr1901_modern: everyone else uses code models, e.g. -mcmodel=tiny
<whitequark>
a lot of architectures use linker relaxation...
<whitequark>
x86 included
<sorear>
and riscv will likely wind up with code models *too* because you need =large (>2GB executables)
<sorear>
it also uses a quadradic time algorithm in the linker. I sent a sketch of a linear-time algorithm to them >year ago but I don't think it's been implemented
<bubble_buster>
it's turning lui %hi(label); addi %lo(label) into just addi %lo(label) because high bits of label are 0
<bubble_buster>
don't want my students to get confused when they write asm test programs and ld mucks them up
<sorear>
pass -Wl,--no-relax to gcc
<cr1901_modern>
sorear: medlow, medany?
<sorear>
or put .option norelax in your .s file
<sorear>
cr1901_modern: ikr
<cr1901_modern>
I don't understand why you'd prefer one to the other. They both seem to have the same limitations
<bubble_buster>
sorear: thank you
<bubble_buster>
I thought I looked for "relax" in 'ld --help' but I think I only looked in 'as --help'
<rqou>
bubble_buster: students? where? better not be $FANCY_SCHOOL :P
<bubble_buster>
lol what's your cutoff for $FANCY_SCHOOL
<rqou>
$FANCY_SCHOOL = Berkeley because whitequark was making fun of me
<bubble_buster>
ah I wish
<bubble_buster>
next best tho
<bubble_buster>
I assume the TAs at Berkeley already have this figured out, we're just playing catch-up here at Illinois
<azonenberg_work>
cpresser: awesome, excited to see how people do
<azonenberg_work>
cpresser: in addition to feedback on the problems themselves i'd like feedback on the scoring
<azonenberg_work>
Do the point values seem fair? Are some easier/harder than the point value would suggest?
unixb0y has quit [Ping timeout: 272 seconds]
AlexDaniel-old[m has quit [*.net *.split]
AlexDaniel-old[m has joined ##openfpga
unixb0y has joined ##openfpga
<cr1901_modern>
whitequark: Didn't mean to imply that riscv was the only user of relaxation. The article made me think riscv is "special" in that (quote) "essentially no .text section symbol addresses are known until link time." >>
<cr1901_modern>
And other archs, where .text section symbols are known before link time, can optimize reg-indirect calls/jumps into pc-rel at compile time.
<cr1901_modern>
From what I;m reading tho, lots of binutils backends solve the "can this jump fit into pc-rel or not?" problem using relaxation.
<cr1901_modern>
(And, thinking about it, shouldn't "essentially no .text section symbol addresses are known until link time." apply to all ld backends that use relaxation?)
<sorear>
there are other targets where relaxation can add and remove bytes, but the best-known ones have, at most, length-preserving relaxations
<sorear>
the _other_ annoying thing is that riscv .a/.o files are twice the size of other arches, because of all the extra (intra-segment) relocations
<sorear>
x86 gcc generates .o files where relative branches are resolved, which means no reloc and no symbol
<cr1901_modern>
I mean, afaik 32-bit (sic) doesn't have a meaningful pc-rel insn, so no opportunity for optimization for calls between translation units?
<azonenberg_work>
cr1901_modern: in x86 there is pc-relative addressing if you cheat
<cr1901_modern>
(I don't know if this is length-changing. I'll have to check the v810 manual when I have a chance. But pretty sure turning a reg indirect to 26-bit pc-rel is length-changing)
<rqou>
you'll mess up the return address predictor
<sorear>
cr1901_modern: I specified "well-known" because I know it's used for some embedded targets; jimwilson knows more details
<sorear>
what you want to do is what clang does for -fPIC, call a function that does "mov [esp], ebx; ret"
<sorear>
amusingly
<sorear>
ryzen's return stack predictor is documented to have special handling of CALL $+0
<sorear>
i'm assuming special means "don't push anything", although other people on #riscv think it's fancier than that
<cr1901_modern>
>what you want to do is what clang does for -fPIC, call a function that does "mov [esp], ebx; ret" Is this in reference to return addr predictor destruction?
<sorear>
"mov ebx, [esp]; ret"
<sorear>
sorry, mangled at&intel syntax
<sorear>
at&tel? int&t?
digshadow has quit [Quit: Leaving.]
<fseidel>
That's Motorolat to you :-P
<fseidel>
*Motorotel
<cr1901_modern>
Where can I find out more about "mov ebx, [esp]; ret" and why it works for "free PIC on 32-bit?"
digshadow has joined ##openfpga
digshadow has quit [Client Quit]
digshadow has joined ##openfpga
<sorear>
hmm, is the behavior of that code not clear to you?
<cr1901_modern>
mov dst, src; ret
<cr1901_modern>
(or did you mean at&t syntax? :P)
<cr1901_modern>
Forget it, my capacity for critical thinking is at like 5% tonight, I'm going to quit while I can still save face. Night folks :)
<azonenberg_work>
rqou: well i normally see that used in shellcode
<azonenberg_work>
where perforance isnt a priority :p
etrig has quit [*.net *.split]
unixb0y has quit [*.net *.split]
Lord_Nightmare has quit [*.net *.split]
AlexDaniel-old[m has quit [*.net *.split]
thehurley3[m] has quit [*.net *.split]
rqou has quit [*.net *.split]
azonenberg_work has quit [*.net *.split]
Adluc has quit [*.net *.split]
mietek has quit [*.net *.split]
prpplague has quit [*.net *.split]
pakesson has quit [*.net *.split]
marex-cloud has quit [*.net *.split]
stefanct has quit [*.net *.split]
rvense has quit [*.net *.split]
xdeller has quit [*.net *.split]
feuerrot has quit [*.net *.split]
tinyfpga has quit [*.net *.split]
gnufan has quit [*.net *.split]
zkms has quit [*.net *.split]
cpresser has quit [*.net *.split]
Zorix has quit [*.net *.split]
Ultrasauce has quit [*.net *.split]
dfgg has quit [*.net *.split]
mIKEjONES has quit [*.net *.split]
CoffeeFlux has quit [*.net *.split]
cyrozap has quit [*.net *.split]
kuldeep has quit [*.net *.split]
lb5tr has quit [*.net *.split]
mmicko has quit [*.net *.split]
lexano has quit [*.net *.split]
wolfspraul has quit [*.net *.split]
vup has quit [*.net *.split]
Marex has quit [*.net *.split]
felix_ has quit [*.net *.split]
noopwafel has quit [*.net *.split]
fouric has quit [*.net *.split]
knielsen has quit [*.net *.split]
ayjay_t has quit [*.net *.split]
zino has quit [*.net *.split]
nurelin has quit [*.net *.split]
clifford has quit [*.net *.split]
cr1901_modern has quit [*.net *.split]
Flux42 has quit [*.net *.split]
lrvick has quit [*.net *.split]
SpaceCoaster has quit [*.net *.split]
brizzo has quit [*.net *.split]
Ellied has quit [*.net *.split]
plaes has quit [*.net *.split]
ym has quit [*.net *.split]
wpwrak has quit [*.net *.split]
gruetzkopf has quit [*.net *.split]
asy_ has quit [*.net *.split]
iximeow has quit [*.net *.split]
hl has quit [*.net *.split]
drawkula has quit [*.net *.split]
Patater has quit [*.net *.split]
q3k has quit [*.net *.split]
indefini[m] has quit [*.net *.split]
benreynwar has quit [*.net *.split]
kem_ has quit [*.net *.split]
Mimoja has quit [*.net *.split]
specing has quit [*.net *.split]
finsternis has quit [*.net *.split]
utzig has quit [*.net *.split]
lain has quit [*.net *.split]
dingbat has quit [*.net *.split]
kc8apf has quit [*.net *.split]
esden has quit [*.net *.split]
Bike has quit [*.net *.split]
X-Scale has quit [*.net *.split]
Ekho- has quit [*.net *.split]
implr has quit [*.net *.split]
ZipCPU has quit [*.net *.split]
flaviusb has quit [*.net *.split]
bibor has quit [*.net *.split]
mwk has quit [*.net *.split]
sgstair has quit [*.net *.split]
oni has quit [*.net *.split]
awygle has quit [*.net *.split]
uovo has quit [*.net *.split]
pointfree[m] has quit [*.net *.split]
nickjohnson has quit [*.net *.split]
miek has quit [*.net *.split]
digshadow-c has quit [*.net *.split]
cblam has quit [*.net *.split]
diamondman has quit [*.net *.split]
cnomad has quit [*.net *.split]
renze has quit [*.net *.split]
balrog has quit [*.net *.split]
marcan has quit [*.net *.split]
jhol has quit [*.net *.split]
sorear has quit [*.net *.split]
daveshah has quit [*.net *.split]
jeandet has quit [*.net *.split]
grummel has quit [*.net *.split]
Xark has quit [*.net *.split]
G33KatWork has quit [*.net *.split]
Prf_Jakob has quit [*.net *.split]
kmehall has quit [*.net *.split]
eddyb has quit [*.net *.split]
whitequark has quit [*.net *.split]
ovf has quit [*.net *.split]
elms has quit [*.net *.split]
comietek has quit [*.net *.split]
mithro has quit [*.net *.split]
pointfree has quit [*.net *.split]
bubble_buster has quit [*.net *.split]
eightdot has quit [*.net *.split]
reportingsjr has quit [*.net *.split]
florolf has quit [*.net *.split]
tnt has quit [*.net *.split]
scrts has quit [*.net *.split]
moho2 has quit [*.net *.split]
forrestv has quit [*.net *.split]
dx has quit [*.net *.split]
Morn_ has quit [*.net *.split]
hackkitten has quit [*.net *.split]
fseidel has quit [*.net *.split]
TAL has quit [*.net *.split]
kristianpaul has quit [*.net *.split]
grantsmith has quit [*.net *.split]
msgctl has quit [*.net *.split]
qu1j0t3 has quit [*.net *.split]
Kitlith_ has quit [*.net *.split]
plaes has joined ##openfpga
azonenberg_work has joined ##openfpga
CoffeeFlux has joined ##openfpga
rqou has joined ##openfpga
cpresser has joined ##openfpga
flaviusb has joined ##openfpga
daveshah has joined ##openfpga
Marex has joined ##openfpga
Lord_Nightmare has joined ##openfpga
Prf_Jakob has joined ##openfpga
Ellied has joined ##openfpga
fseidel has joined ##openfpga
Flux42 has joined ##openfpga
mmicko has joined ##openfpga
stefanct has joined ##openfpga
fouric has joined ##openfpga
Xark has joined ##openfpga
TAL has joined ##openfpga
Adluc has joined ##openfpga
_whitelogger has joined ##openfpga
CoffeeFlux is now known as Guest61919
wolfspraul is now known as Guest34117
cyrozap is now known as Guest57051
plaes has quit [Changing host]
plaes has joined ##openfpga
Guest61919 has quit [Changing host]
Guest61919 has joined ##openfpga
mmicko has joined ##openfpga
vup has joined ##openfpga
kuldeep has joined ##openfpga
lexano has joined ##openfpga
kuldeep is now known as Guest83745
<rqou>
goddammit freenode wat r u doin?
ayjay_t has joined ##openfpga
s1dev has joined ##openfpga
Ekho- is now known as Ekho
Guest61919 is now known as CoffeeFlux
azonenberg_work has quit [*.net *.split]
Adluc has quit [*.net *.split]
mietek has quit [*.net *.split]
prpplague has quit [*.net *.split]
pakesson has quit [*.net *.split]
marex-cloud has quit [*.net *.split]
stefanct has quit [*.net *.split]
rvense has quit [*.net *.split]
xdeller has quit [*.net *.split]
feuerrot has quit [*.net *.split]
tinyfpga has quit [*.net *.split]
gnufan has quit [*.net *.split]
cpresser has quit [*.net *.split]
zkms has quit [*.net *.split]
Zorix has quit [*.net *.split]
Ultrasauce has quit [*.net *.split]
dfgg has quit [*.net *.split]
mIKEjONES has quit [*.net *.split]
lrvick has joined ##openfpga
azonenberg_work has joined ##openfpga
Adluc has joined ##openfpga
mietek has joined ##openfpga
prpplague has joined ##openfpga
stefanct has joined ##openfpga
pakesson has joined ##openfpga
marex-cloud has joined ##openfpga
rvense has joined ##openfpga
gnufan has joined ##openfpga
xdeller has joined ##openfpga
feuerrot has joined ##openfpga
zkms has joined ##openfpga
tinyfpga has joined ##openfpga
cpresser has joined ##openfpga
Zorix has joined ##openfpga
dfgg has joined ##openfpga
mIKEjONES has joined ##openfpga
Ultrasauce has joined ##openfpga
<rqou>
hey i have a dumb question: freenode is centrally controlled rather than federated -- why can't they develop server software that can be seamlessly upgraded without netsplits and dropped connections?
<rqou>
i.e. build a "modern" chat app that just still speaks irc on the edge but is something custom internally
<sorear>
they can't even get open proxy scanning on connect working with outside help, and other irc networks were doing that when I started using IRC in 2003
<sorear>
s/with/without/
<rqou>
why do they even have custom server software if it isn't good?
<rqou>
also i just realized you can totally do a seamless upgrade even on a traditional federated irc network
<rqou>
you just need to develop the software to plan for this from the start
<sorear>
there's at least one ircd that does this
<sorear>
i know this because they've whined about musl's dlclose implementation
<rqou>
wat
<rqou>
why is dlclose involved?
<sorear>
they do updates by putting all of the logic in a shared library, then reloading it to update
<rqou>
um... i guess you can do that
<sorear>
as opposed to the systemd approach of serializing everything and then execve
<rqou>
i would have done it that way
<sorear>
i mostly agree
azonenberg_work has quit [Ping timeout: 252 seconds]
<rqou>
this approach only uses the boringest of features that basically every OS supports
<rqou>
unlike using dlclose
_whitelogger has joined ##openfpga
Bike has quit [Quit: Lost terminal]
<awygle>
that is a weird way to do that, yeah
<awygle>
As much as I (currently) love shared libraries
SpaceCoaster_ has quit [Quit: ZNC 1.6.5+deb1+deb9u1 - http://znc.in]
SpaceCoaster has joined ##openfpga
rohitksingh_work has joined ##openfpga
Guest83745 has quit [Remote host closed the connection]
kuldeep has joined ##openfpga
<sorear>
otoh, I think it makes TLS “interesting”
_whitelogger has joined ##openfpga
<rqou>
which tls lol?
<rqou>
thread local storage?
<rqou>
afaik all current implementations leak it
<sorear>
the other one :p
<rqou>
oh huh i didn't think about how most TLS (the crypto) libraries don't have a "serialize all state" API
<whitequark>
i thought they do
<whitequark>
tls session resumption or sth
<rqou>
session resumption isn't "all the state" though
<rqou>
it's not enough for transparently picking up a connection that's still open
<rqou>
iirc it's just enough information to skip the expensive asymmetric cryptography
<rqou>
also this TLA collision sucks
<rqou>
especially if you're weird like me and care about both
azonenberg_work has joined ##openfpga
<sorear>
especially connections halfway through the handshake
futarisIRCcloud has joined ##openfpga
<rqou>
wow github is starting to finally catch on: "If your school issues email addresses prior to paid student enrollment, or alumni and retired faculty have lifetime access, then we require further proof of your affiliation."
<azonenberg_work>
rqou: lol
<rqou>
can't be a student forever :P
<azonenberg_work>
what so joe.schmoe@alumni.rpi.edu isn't a valid email for educational access?
<rqou>
heh Berkeley (on the current system) doesn't even mark alumni accounts in any way
<rqou>
unfortunately people like awygle that didn't want to deal with the CAA and/or roundcube lost access
digshadow has joined ##openfpga
<rqou>
in general I'm actually quite impressed how good Berkeley IT is at doing migrations
<rqou>
although they were in general pretty overdue for upgrades since Berkeley apparently had some of the oldest infrastructure among universities
<rqou>
*among US universities
_whitelogger has joined ##openfpga
oeuf has joined ##openfpga
uovo has quit [Ping timeout: 255 seconds]
xdeller has quit [Read error: Connection reset by peer]
xdeller has joined ##openfpga
rohitksingh_work has quit [Read error: Connection reset by peer]
s1dev has quit [Ping timeout: 268 seconds]
Flux42 has quit [Quit: Patching.]
rohitksingh has joined ##openfpga
Guest81843 is now known as felix_
futarisIRCcloud has quit [Quit: Connection closed for inactivity]
rohitksingh has quit [Ping timeout: 252 seconds]
Sinclair2 has joined ##openfpga
Flux42 has joined ##openfpga
zino has quit [Quit: Leaving]
zino has joined ##openfpga
ym has quit [Ping timeout: 240 seconds]
zino has quit [Remote host closed the connection]
zino has joined ##openfpga
massi has joined ##openfpga
rohitksingh has joined ##openfpga
rohitksingh has quit [Quit: Leaving.]
pie_ has joined ##openfpga
<gruetzkopf>
my uni migrated from sun convergence to exchange in 2014 (and i'm still not impressed)
pie_ has quit [Ping timeout: 252 seconds]
<Zorix>
i used to manage sun iplanet servers
pie_ has joined ##openfpga
pie_ has quit [Ping timeout: 252 seconds]
pie_ has joined ##openfpga
pie__ has joined ##openfpga
pie_ has quit [Read error: Connection reset by peer]
genii has joined ##openfpga
m_t has joined ##openfpga
Guest55255 is now known as Ellied
moho2 is now known as moho1
Sinclair2 has quit [Quit: Bye Bye]
<pie__>
shapr, apparently i missed some messages from you
m_t has quit [Quit: Leaving]
cr1901_modern has quit [Quit: Leaving.]
cr1901_modern has joined ##openfpga
<awygle>
hopefully telebears is no longer a modem in a closet pretending to be a touch tone phone
massi has quit [Remote host closed the connection]
<daveshah>
Somewhat interesting (and seems rather ambitious)
<tnt>
for 22k$, yeah, that seems ambitious.
<daveshah>
afaict they have no rtl for their i2c boot yet, and they plan to tape out in nov, which is a bit disconcerting
<daveshah>
I suppose 22k$ would do the mpw, but adding packaging, pcbs, fulfilment and the risk of needing a second tapeout makes that seem hard to do
<tnt>
"what could possibly go wrong" :)
rohitksingh has joined ##openfpga
rohitksingh has quit [Quit: Leaving.]
rohitksingh has joined ##openfpga
<rqou>
awygle: they've finally managed to slay the telebear :P
<rqou>
the course registration system whose shittiness has lasted literally for generations :P
<rqou>
as in, there are nonzero cases of students complaining about it and the learning that their parents who also went to Berkeley had to use the same system
<whitequark>
amazing
<rqou>
hey, it was actually amazing when they first rolled it out
<rqou>
Berkeley was one of the earliest universities that had a "modern" computerized course registration system that you could access with a TouchTone(TM) phone
<rqou>
aaaand then the backend didn't significantly change for three decades
Miyu has joined ##openfpga
<rqou>
although it did gain a website eventually
<rqou>
where you had a bunch of form fields that you had to input the same numbers that you would have input into the phone system
<rqou>
if you input anything that wasn't a number like a space the system would break
<qu1j0t3>
rqou: ever since helping my wife register for her courses in 2010 i've wondered why they are SO BAD
<gruetzkopf>
i2c boot AAAAH
<gruetzkopf>
i still need to do class registration for the next semester, let's see if new system is worse than vZPA
<rqou>
apparently Berkeley actually did pretty well because they jumped from a "consistently dumb" essentially-phone system to a new thing
<rqou>
unlike e.g. Cornell that got stuck on a Java abomination
<jn__>
gruetzkopf: when's the deadline for class registrations?
<gruetzkopf>
should be sometime in october
<gruetzkopf>
(then again, EE deadlines are usually after the CS deadlines)
digshadow has quit [Ping timeout: 252 seconds]
<awygle>
rqou: that tweet, while somewhat distressing, implies a level of sophistication that i didn't realize c++ possessed
<whitequark>
c++ has fractal amounts of sophistication
<rqou>
yeah right now c++ has basically everything
<whitequark>
every time you think you're mostly done with it you discover a new way various its features unintentionally interact
<whitequark>
and someone gave it a name like "curiously recurring template pattern"
<awygle>
lol
<whitequark>
or "Koenig lookup"
<whitequark>
(not invented by Koenig, obviously.)
<whitequark>
like ADL is how std::cout << x works, worse, ADL *exists* to make std::cout << x work, and it's a feature so hideous it took like decades before anyone has implemented it correctly
<rqou>
i haven't read the spec myself but my housemate was claiming that the c++ spec at this point actually includes ~all the concepts of the Rust borrow/lifetime system. it's just that where Rust would give you an error C++ just gives you a UB ouroboros that deletes your program
<whitequark>
christ. c++ has so much sophistication, you could tie it into a noose and hang yourself on it
<rqou>
and/or gives you an exploit of course
<whitequark>
and you probably will want to.
<whitequark>
rqou: yep
<whitequark>
Box<> is unique_ptr<>, and Arc<> is shared_ptr<>
<whitequark>
and borrow/lifetime system is just how stack-allocated variables behave in C and C++
<awygle>
yeah i mean i thought that was pretty much known. rust is just a lot more bondage-and-discipline about it.
<whitequark>
i'd say c++ is
<whitequark>
because c++ requires discipline
<awygle>
whitequark: rust has discipline so you don't have to
<fseidel>
people used to complain that the templates were turing complete, but at this point I'd wager the typesystem is too :-P
<awygle>
(TM TM TM)
<whitequark>
fseidel: I think Rust's is, too
<awygle>
.... okay i don't know a _lot_ about BDSM, but i feel like i just implied that Rust puts programmers into sub space?
<fseidel>
Rust seems interesting, but I've also heard it's borderline impossible to write a doubly linked list with decent performance in it, among other thing
<fseidel>
s
<whitequark>
fseidel: 1) that's bullshit 2) and it's also irrelevant
<rqou>
the answer i usually hear for that is "why would you want that anyways"
<awygle>
fseidel: i feel that this impression comes from people treating "unsafe" as "never use". it's hard (impossible?) to write a doubly linked list in safe rust, but it's easy in unsafe rust, which is fine.
<whitequark>
it's bullshit because you can write anything in Rust that you can write in C++ and about as easily
<awygle>
it's a marketing thing
<whitequark>
you can write a DLL in safe rust in a variety of ways, which quickly expose the fact that a DLL is a useful data structure approximately never
<fseidel>
okay, the only people I know who write Rust work at a company doing things where they have to write proofs about their code
<fseidel>
because it's for flight safety
<fseidel>
and they can't use it
<fseidel>
maybe I'll give Rust a shot then
<whitequark>
there's a DLL in the standard library anyway
<awygle>
DO-178C doesn't require proofs, but ok
<whitequark>
... which is probably going to be deprecated because it turns out that a DLL isn't a fucking useful data structure
<whitequark>
well i guess it's useful at interviews
<fseidel>
it's not required, but the company makes them do it
<fseidel>
at least, I think so, anyway, one of them was telling me this
<whitequark>
it's not a bad idea
<fseidel>
some sort of RTOS
<whitequark>
but... if you're writing proofs anyway, why wouldn't you use unsafe?
<fseidel>
you know, that's a really good question...
<fseidel>
I should prod that dude and find out
<whitequark>
you just need to prove a stronger statement about your code than what the rust type system can express
<awygle>
i guess you only have to prove the unsafe bits. except that's not true, because behavior is _somewhat important_ to flight safety
<rqou>
whitequark: what about the "mumble mumble intrusive list" part? :P
<whitequark>
rqou: intrusive lists are definitely not one of Rust's strong points
<awygle>
intrusive list is a great name
<whitequark>
the current best solution is basically "use macros" and I'm not exactly happy about that
<whitequark>
but I've never actually written any code that needs a single intrusive list, much less so many of them that I'd think about abstracting them out
<whitequark>
so I have no idea how bad Rust would be for that sort of code anyway
<awygle>
i used to use them on the msp430 a fair amount
<awygle>
iirc chibios/nil uses an intrusive list of threads for its scheduler?
<whitequark>
yeah but did you *need* to use them
<awygle>
well, sorta, because i was porting an existing rtos
<awygle>
i get your point though, yes
<rqou>
btw my housemate claimed that they directly inspired the "mumble mumble kernel embedded" line of the "Learning Rust With Entirely Too Many Linked Lists" :P
<fseidel>
hah
digshadow has joined ##openfpga
m_t has joined ##openfpga
bibor has quit [Read error: Connection reset by peer]
gruetzkopf has quit [Read error: Connection reset by peer]
gruetzkopf has joined ##openfpga
bibor has joined ##openfpga
[X-Scale] has joined ##openfpga
X-Scale has quit [Ping timeout: 260 seconds]
[X-Scale] is now known as X-Scale
pie__ has quit [Remote host closed the connection]
pie__ has joined ##openfpga
prpplague has quit [Read error: Connection reset by peer]
<fseidel>
holy shit, did someone seriously do that?
<whitequark>
if i was to believe the person who told it to me, yeah
<rqou>
"oh some guy who died" <-- how are you so nonchalant about this?
<whitequark>
rqou: why wouldn't i be?
<whitequark>
i don't even know his name. all i know was the mistake he made and the character of death (swift and certain)
<azonenberg_work>
whitequark: do you know what was in the bottle? :p
mumptai has quit [Read error: Connection reset by peer]
<azonenberg_work>
First guess: incompletely purified meth
<azonenberg_work>
(this sounds like the kind of thing that would happen in a drug lab)
<whitequark>
azonenberg_work: don't think it was a drug lab
<azonenberg_work>
Judging by how good drug cooks are at lab safety etc
<whitequark>
some factory iirc
mumptai has joined ##openfpga
* tnt
recently had the issue that sodium hydroxide attacks PET :/
<rqou>
oh yeah, how do meth labs all end up (being described as) massively contaminated?
<rqou>
the procedure doesn't sound _that_ complicated
<gruetzkopf>
doesn't help if you're on meth while doing it
<whitequark>
tnt: PE for all the things
<whitequark>
or PP
<whitequark>
not PET
pie__ has quit [Remote host closed the connection]
<azonenberg_work>
rqou: i think they just get residue all over the place
<rqou>
or glass? (not so good for sodium hydroxide thought)
pie__ has joined ##openfpga
<azonenberg_work>
dont wear gloves, or do and dont wash their hands after
<azonenberg_work>
etc
<azonenberg_work>
spilling things on the ground
<tnt>
whitequark: yeah, I just grabbed drink bottles I had on hand (and properly labelled them !). I did check it was OK for diluted Sulphuric acid and somehow didn't bother to check for NaOH.
<whitequark>
rqou: i vaguely recall looking it up and apparently it was pretty bad
<whitequark>
but i dont really feell like dredging it up again
<whitequark>
i think it's something about the crude processes people have to employ in the us
<whitequark>
like generate tons of bog standard low level chemical waste with no means to dispose
<rqou>
but somehow US law enforcement really likes to play up how dangerous the waste is
<azonenberg_work>
And the boobytraps
<azonenberg_work>
:p
<azonenberg_work>
Drug labs are notorious for that
<awygle>
yeah one near my elementary school was trapped with nitroglycerin or something
<sorear>
I feel like relatively few substances will provide a “swift and certain” death
<awygle>
at any rate it exploded
<azonenberg_work>
One technique i've heard of is to get a syringe and narrow gauge needle and squirt gasoline through the base of a standard incandescent light bulb
<azonenberg_work>
When power is turned on the filament heats up, boils some of the gas
<azonenberg_work>
the pressure ruptures the bulb, then boom fireball
pie__ has quit [Ping timeout: 268 seconds]
<gruetzkopf>
brilliant
<rqou>
heh that's pretty cute
<gruetzkopf>
i have some NOS compensation caps that'll reliably explode ~3min after powerup :D
<whitequark>
lol
<sorear>
so _that’s_ why those things are being phased out everywhere
<qu1j0t3>
haha
pie_ has joined ##openfpga
<rqou>
hmm looking into the illicit synthesis procedures none of them look that dangerous?
<rqou>
are meth cooks just that bad at chemistry? :P
<florolf>
rqou: re "having the irc network talk some different protocoll internally and only interface using the irc protocol": https://robustirc.net/
<sorear>
gruetzkopf: some what nows?
pie_ has quit [Ping timeout: 244 seconds]
Guest57051 has quit [Quit: Client quit]
cyrozap has joined ##openfpga
cyrozap has quit [Quit: Client quit]
cyrozap has joined ##openfpga
gnufan1 has joined ##openfpga
gnufan1 has quit [Client Quit]
gnufan_home has joined ##openfpga
Bike has joined ##openfpga
genii has quit [Remote host closed the connection]
Guest34117 has quit [Quit: leaving]
wolfspraul has joined ##openfpga
digshadow has quit [Quit: Leaving.]
digshadow has joined ##openfpga
digshadow has quit [Client Quit]
digshadow has joined ##openfpga
pie_ has joined ##openfpga
azonenberg_work has quit [Ping timeout: 252 seconds]