<awygle> i _want_ to
<cr1901_modern> According to clifford, SV is much nastier than Verilog
<awygle> but realistically...
<awygle> implementation-wise i absolutely believe that
<whitequark> awygle: I admit that SV fixes a *lot* of issues with Verilog that Migen fixes.
<whitequark> I've had that discussion and I actually see why I might use SV
<whitequark> (not in Glasgow, of course, since there is enormous value in having everything in Python, it's not viable, but elsewhere, perhaps)
<TD-Linux> whitequark, finally someone I agree 100% with. verilog is defined by interpreter behavior and it's a disaster.
* awygle 's build finishes and he disappears for ~15m
<whitequark> azonenberg_work: poke
<awygle> oh i just remembered why we started talking about that lol
<awygle> if anyone knows of a verilog DDR* controller that works, please let me know
kuldeep has quit [Ping timeout: 240 seconds]
<TD-Linux> have you liked migen in glasgow enough that you would start with it for another project?
<whitequark> me? yeah sure, migen is my go-to HDL
<whitequark> the tooling around the HDL is excellent too, I never want to write a pcf file again
<whitequark> or a shell script to call synthesis and PAR
<whitequark> awygle: hell, in migen you can't even easily make a baud rate calculator
<whitequark> in verilog*
<whitequark> because there's no log2 builtin
<whitequark> you have to use vendor-specific functions for something as basic
<awygle> whitequark: $clog2 has been built in since... i think 2005
<TD-Linux> cool. I was also trying chisel and it was nice
<cr1901_modern> whitequark: Ever try fusesoc? It also exists to automate that crap away (but I haven't used it much)
<whitequark> awygle: I'm pretty sure I have specifically encountered the problem that making a portable log2 call was hard
<whitequark> furthermore
<awygle> i like fusesoc (and olof) but "tained with this bullshit" is a good description
<whitequark> let's say I want a #error "baud rate out of tolerance"
<whitequark> you can't express that in verilog.
<awygle> that i'll grant (and again, system verilog fixes it)
<awygle> you _can_, you just do it as "module error_baud_rate_out_of_tolerance" which is repulsive
<whitequark> well for some definition of "express" that sucks, you can
<whitequark> writing parametric verilog code is hell in general
<whitequark> I can and have done it, and I never want to do it again
<awygle> okay apparently i'm multitasking "test the build" and "argue on irc" lol. what's wrong with pcf files, or alternately, why does migen want to know about the _board_ i have?
<rqou> parameters vhdl with higher-order functions :P
<rqou> *parametric
<cr1901_modern> awygle: It wants to know the board, to generate the pcf file
<awygle> rqou: i'm not even raising the "why not vhdl" issue here lol, it seems like a distraction from the larger point
<pie_> rqou, you mean CLaSH (except its not vhdl :P)
<whitequark> there's no FOSS VHDL tooling
<whitequark> at all
<awygle> cr1901_modern: right, but like... why. what value does that bring.
<whitequark> awygle: you often have many designs targeting a single board
<whitequark> migen lets you abstract away that knowledge
<awygle> i don't get why "i have to google cr1901_modern's blog about porting a board to migen every time i have a new board" is better than "i have to tweak a pcf file every time i have a new board"
<whitequark> if you want to light up a LED, it's self.comb += platform.request("led").eq(1)
<TD-Linux> does migen have special sauce to handle e.g. ice40 tristates nicely
<cr1901_modern> awygle: By convention all I/O pins with the same purpose in migen have the same name. So in principle, it's "write PCF once, never look at it again"
<whitequark> TD-Linux: yes
<whitequark> I wrote it :p
<cr1901_modern> TD-Linux: yes, whitequark added it
<TD-Linux> sold
<whitequark> I need to add DDR too
<cr1901_modern> I think I already did that?
<cr1901_modern> at least the outputs?
<cr1901_modern> If I didn't, then that's my fault/an oversight. It was meant to be part of "icestorm improvements" PR I did last yet :P
<cr1901_modern> year*
<whitequark> I think it's there but there was some issue with it
<whitequark> and there's no inputs
<awygle> okay so that seems like _sort of_ a good idea for _specifically_ the context of dev boards
<awygle> but actively hostile (in a mild sort of way) to production hardware
<awygle> because now you have to maintain a fork of the "toolchain" to have a custom board
<whitequark> what?
<sorear> meanwhile, $elsewhere people seem quite unhappy that [rocket] is chisel and not "real verilog"
<whitequark> glasgow doesn't maintain a fork of toolchain
<awygle> unless you can stash it in the code someplace
<awygle> i'm not really sure how this works
<cr1901_modern> awygle: And why is that so much worse than a PCF
<whitequark> awygle: glasgow/platform/__init__.py
<cr1901_modern> writing a migen board file is _not_ much extra work than a PCF.
<awygle> whitequark: ah ok, that's less irritating then
<whitequark> note how it's actually structured
<awygle> cr1901_modern: i thought it had to be in migen proper, rather than being in the project
<cr1901_modern> awygle: No
<awygle> i withdraw, or at least reduce, the objection
<cr1901_modern> I just insist on adding every board I port :P
<rqou> goddammit another discussion on birbsite about fish antibiotics
<rqou> why are we so bad (good?) at capitalism
<cr1901_modern> And sb0/whitequark have been pretty okay w/ adding boards I port
<cr1901_modern> (probably b/c it's minimal maintenance)
<awygle> whitequark: okay, so nothing about that irritates me that isn't just something irritating about python, i suppose.
<whitequark> awygle: oh yeah, full ack on disliking python
<whitequark> i'd have personally used something like hardcaml
<whitequark> but i want glasgow to be useful for other people
<awygle> i have never heard of that but it is the best name
<cr1901_modern> whitequark: I didn't include input DDRs for a reason. I don't think SB_IO primitive supports DDR input?
<whitequark> cr1901_modern: of course it does
<whitequark> it would be really dumb if it didn't
<whitequark> take a look at the circuit
<cr1901_modern> There's no setting for it in the I/O parameter that determines the mode
<awygle> aargh okay this is unrelated but this C# code is reaching a function call site and then just fucking off to god knows where and it's infuriating
<whitequark> cr1901_modern: yes
<whitequark> it's always active
<whitequark> because it's just another input register
<cr1901_modern> whitequark: Oh... well, then fire away and add DDR inputs :P. Idk what's wrong w/ the outputs, I couldn't test them
<rqou> whitequark: how long until we get octopart for pharma?
<whitequark> rqou: regulators hate him!
<rqou> loool
<whitequark> but yes, a "silkroad" for regular ass medications and not just weed and shit would be super useful
<whitequark> unfortunately there aren't millions to be made on it
<sorear> it's easy to get people worked up moralizing about orphan drug prices, is there any way we can leverage that
<rqou> do you have to make money? i mean, erowid exists
<whitequark> fix the FDA?
<whitequark> like, FDA more or less creates this problem
<whitequark> even though it's supposed to do the opposite
emeb has quit [Ping timeout: 264 seconds]
<whitequark> it's not useless like most libertarians would tell you, but it's definitely complicit
<rqou> but what about avoiding another thalidomide?
<whitequark> rqou: did I suggest abolishing FDA
<sorear> US culture around risk is fucked in an extremely cross-cutting way
<whitequark> yep
<awygle> very yes
<rqou> how do you fix it without potentially causing a new problem like that?
<sorear> security theatre, thalidomide, NYC cutting subway speeds by 25% system-wide because one person got killed a few years ago,
<awygle> it sort of feels like we got to, idk, 1955 and just decided no new risks were allowed
<whitequark> rqou: i could answer that
<gruetzkopf> here they're optimising up all the time
<whitequark> but let me answer a different question
<awygle> we kept all the _old_ risks, but stopped allowing new ones
<awygle> that's not really true but that's how it feels sometimes
<whitequark> even if you get another thalidomide, if people stop dying from preventable causes like "diabetes" and "lack of basic antibiotics" it's still better than the current situation
<gruetzkopf> level crossings are gonna die out in europe
<awygle> serious question that i'm aware sounds dumb/naive - how close are home pill printers to being a thing?
<whitequark> not gonna happen
<rqou> heavily regulated afaik
<awygle> i just mean technically
<sorear> the closest modern equivalent to thalidomide is the coxibs, and i'm far from convinced we made the right call
<TD-Linux> drug synthesis is super complex and each one is totally different
<whitequark> awygle: you need a cleanroom at a pharma plant
<sorear> gruetzkopf: can't tell if you see that as good or bad
<whitequark> among other things
<whitequark> like what TD-Linux said
<whitequark> it's never going to work out at home
<awygle> yeah, that's what i figured
<gruetzkopf> it's grad
<gruetzkopf> great
<gruetzkopf> instant Vmax upgrade from 160 to 200
<whitequark> like, I have considered drug synthesis for things I can't obtain that I need
<whitequark> it's absurdly hard to do it right
<whitequark> even with all the resources I have at my disposal in RU
<awygle> that's why i asked you, i figured you'd know if anybody did
<sorear> drug synth requires thousands of bulk chemicals, I could see make-to-order being a thing in a well-stocked facility but scaling the equipment for home use is … problematic
<gruetzkopf> <whitequark> awygle: you need a cleanroom at a pharma plant can get that
<whitequark> like an NMR is just a phone call away
emeb has joined ##openfpga
<whitequark> gruetzkopf: yes, but not at the kind of home awygle means
<whitequark> *I* could probably do it, with an enormous amount of effort and many years
<awygle> awygle is open to all levels of home
<rqou> nurdrage didn't need a cleanroom :P
<awygle> for purposes of this discussion
<rqou> although it did take like two years
<whitequark> easily packaged solutions that any bored person can use? nah
<whitequark> rqou: is it pure enough to actually use tho
<rqou> idk, but he had an nmr spectrum
<gruetzkopf> (worked on the clean-room monitoring equipment and the certification)
<sorear> gruetzkopf: so, not a massive waste of money and a ball and chain around future rail construction? my US bias is to see basically everything as an attempt to block future rail construction
<prpplague> awygle: i was thinking this morning... (i know... very dangerous), maybe starting off with the replication of a 16v8 GAL
<azonenberg_work> whitequark: so $clog2 has been around for years but ISE doesnt support it
<rqou> it had ether contamination for sure
<whitequark> awygle: ^
<whitequark> rqou: ether is fine
<gruetzkopf> big rail projects still get initiated here
<whitequark> it's harmless
<azonenberg_work> it works fine in vivado
<gruetzkopf> we've had VDE8.2 complete in the last year
<awygle> whitequark: okay, different question (this is interesting). how hard would it be for you or someone with your level of knowledge to get into the business of manufacturing pills for generic drugs?
<gruetzkopf> ex-gdr to munich at >250
<awygle> how much capital would you need, etc etc
<awygle> azonenberg_work: ise is dead to me
<whitequark> awygle: not the right person to ask
<whitequark> i'm interested in technical side, not regulatory andmanagement side
<awygle> whitequark: fair enough
<whitequark> and whatyou want is the latter two
<rqou> afaik most o-chem students can pull of paracetamol, but that's not particularly useful lol
<whitequark> paracetamol is trivial
<TD-Linux> do vitamin C pills count? :^)
<whitequark> there's tons of drugs i could trivially manufacture at good purity and scale
<whitequark> but none of them have overpriced versions
<whitequark> that need replacement
<sorear> the question is always "what from what"
<whitequark> for obvious reasons
<pie_> man now you have me thinking about how active ingredients is one thing but afaik the delivery is also important?
<whitequark> sorear: aniline? :p
<awygle> sure, yes
<gruetzkopf> i did both acetylsalicylic acid and paracetamol in high school chemistry
<azonenberg_work> yeah all of the excipients are nontrivial
<azonenberg_work> to improve bioavailability
<whitequark> pie_: depends on the drug
<rqou> gruetzkopf: wat
<whitequark> heavily
<whitequark> some just work with whatever
<whitequark> some need to be micronized
<sorear> whitequark: L-ascorbate from aniline? lol have fun with the chiral centers
<azonenberg_work> awygle: spartan6 and CPLDs are still interesting for some applications
<whitequark> or prepared for IM
<whitequark> sorear: paracetamol
<awygle> azonenberg_work: 7-series is barely interesting for my applications :p
<azonenberg_work> awygle: lol
<rqou> gruetzkopf: how was your schooling such a not-clusterfuck?
<TD-Linux> I would kinda like icecpld to be a thing just to have 5v tolerant parts for my retroputing needs
<azonenberg_work> the bigger issue for me is yosys support, thats one of the things holding me off from going full systemverilog
<gruetzkopf> my school isn't even considered that good :P
kuldeep has joined ##openfpga
<awygle> azonenberg_work: +30
<gruetzkopf> (but the chem teacher was great. and kind of a pyro)
<pie_> gruetzkopf, one would think pyros are the best for some reason
<whitequark> lool
<whitequark> we also had one of chem teachers like that
<azonenberg_work> TD-Linux: why would you need 5V tolerance?
<rqou> my chem teacher was "just" a civil engineer or something like that
<TD-Linux> >retroputing needs
<cr1901_modern> azonenberg_work: Retrocomputing
<whitequark> i remember him mixing perchlorate with red phosphorus with a massive grin
<whitequark> and then hitting it
<cr1901_modern> b/c level shifters take space
<azonenberg_work> i mean i'm putting it in for starshipraider as kind of like a "no way i will EVER need to go higher than this"
<whitequark> with a stick or something
<whitequark> the explosion was great
<azonenberg_work> whitequark: armstrong's mixture?
<whitequark> azonenberg_work: never heard that name but yes
<TD-Linux> well, it also just takes a bazillion level shifters for e.g. a 32 bit bus
<whitequark> azonenberg_work: he mixed like a gram of the damn stuff
<pie_> no i want to do chemistry again i hate you guys :D
<pie_> *now
* azonenberg_work never understood the whole draw of retro computing stuff
<whitequark> was really insistent we don't get anywhere close
<azonenberg_work> i want to work with the most advanced tech i can get my hands on
<TD-Linux> but it turns out I'd much rather spam 245s than actually work on icecpld
<azonenberg_work> whitequark: lol a GRAM
<awygle> whitequark: my chem teacher did something similar (possibly that) and blew a test tube shard through the drop ceiling
<azonenberg_work> how did the windows take it? :p
<whitequark> awygle: so
<rqou> btw whitequark just from a technical perspective what do you think about nurdrage's pyrimethamine procedure?
<azonenberg_work> that stuff is pretty potent
<whitequark> awygle: he has also synthesized a carbide (aluminium carbide maybe? or calcium) for us
<awygle> i don't think it was _on purpose_ though lol
<whitequark> awygle: he put it in a small crucible and covered with a cap and heated
<TD-Linux> azonenberg_work, who says retroputing isn't advanced? my magneto-optical drives have six degrees of freedom. can your silly ssds top that
<awygle> show yourself coward, i will never solid state
* gruetzkopf hides TALs SGI Onyx2
<whitequark> awygle: it did nothing for a while and then the cap made an impression on the concrete ceiling
<pie_> advanced age
<pie_> jkjk
<azonenberg_work> gruetzkopf: lol
<whitequark> awygle: he did not tell us that this will happen and was very pleased with himself i think
<azonenberg_work> i finally got rid of my Origin 2000 rack
<gruetzkopf> :(
<whitequark> i was filming it on my cellphone, sadly the record is now lost
<gruetzkopf> i'd have taken the routers
<azonenberg_work> dont worry, it went to a good home
<whitequark> azonenberg_work: the fume hood glass all survived
<gruetzkopf> ah
<azonenberg_work> also it was just the rack
<whitequark> maybe not a gram
<gruetzkopf> ah
<azonenberg_work> not the nodes
<whitequark> it seemed like a shitton
<cr1901_modern> azonenberg_work: It's complicated why I like retrocomputing. You don't have to understand it :).
<gruetzkopf> modules;)
<awygle> there's an amazing picture of the moment of ignition of a KNO3+sugar mixture in our chem lab
<gruetzkopf> nodes are the cards in them
<awygle> a perfect sphere of flame with a center made of smoke
<whitequark> nice
<pie_> awygle, post pls
<awygle> a friend of mine overheated it on a bunsen burner
<awygle> uhh lemme see if i can find it
<azonenberg_work> gruetzkopf: years ago it was known as momentum.cs.rpi.edu
<azonenberg_work> well, one of eight racks
<azonenberg_work> i found it on the loading dock next to the dumpster
<azonenberg_work> grabbed a few CPUs out of the sockets to decap
<TD-Linux> >not using a double boiler to make your rocket fuel like a propery candy maker
<azonenberg_work> then pulled everything off the rack and rolled it a quarter mile downtown to my apartment
<gruetzkopf> my chem teacher had one of the farmers kids bring NH4NO3
<azonenberg_work> used it as my main equipment rack for years
<azonenberg_work> gruetzkopf: lol
<whitequark> we also made NI3
<whitequark> but it didn't explode i think
<azonenberg_work> whitequark: o_O
<whitequark> azonenberg_work: what
<azonenberg_work> That stuff would probably go boom if you hit it with a laser pointer
<whitequark> it's not actually dangerous
<gruetzkopf> and then practical demonstration of initiation energy ensued
<azonenberg_work> > didnt explode
<whitequark> it's too explosive for any significant amount to accumulate
<azonenberg_work> no i meant, you made NI3 and it did not explode
<whitequark> oh
<whitequark> probably didn't get dry
<whitequark> enough
<azonenberg_work> i've heard stories of it going off during the drying process just from it shifting and contracting
<whitequark> dunno why, it was weird
<gruetzkopf> on a tree stump.
<whitequark> yeah I was expecting that
<whitequark> he has also shown us liquid bromine
<azonenberg_work> in a 3kg ampule? :p
<whitequark> ... without a respirator, in an open flask
<whitequark> it was fuming
<whitequark> in retrospect his disregard to safety was pretty callous
<awygle> yeah people like to tell crazy chem lab stories but uh... that can't be good for your education lol
<gruetzkopf> were our chem teachers the same person, leading a secret double life
<whitequark> awygle: why not
<whitequark> it taught me that i can work with bromine
<whitequark> and i worked with bromine recently
<whitequark> it's smelly af and annoying but you can work with it
<awygle> whitequark: well you managed not to learn "respirators aren't important", but someone else might not have
<awygle> .... the negatives in that sentence didn't really work but you get it
<whitequark> awygle: i think the school just didn't have money for those
<rqou> i actually didn't work with that many "fun" chemicals probably due to living in the People's Republic of California
<whitequark> awygle: we had to buy our own lab coats
<azonenberg_work> +1 to rqou for using the correct name
<awygle> rqou: i went to school in ventura, you were just unlucky
<rqou> eh maybe
<awygle> also "peoples republic of california" is even more irritating than "micro$oft" et al
<rqou> also my chem teacher really was serious about safety
<whitequark> awygle: yeah but know what's more irritating? prop 65
<whitequark> i don't even live in california and i hate it
<awygle> whitequark: full ack
<rqou> since injuring your students isn't a very good look
<azonenberg_work> whitequark: my galvanized screws have a prop 65 warning on them
<awygle> whitequark: see also, 8 and 13
<gruetzkopf> i always brought my own safety glasses
<whitequark> awygle: astounding
<azonenberg_work> maybe for traces of chromium in the plating or something?
<whitequark> azonenberg_work I mean
<gruetzkopf> because i like them being actually transparent
<rqou> azonenberg_work: outside (iirc at SFO) has a prop 65 warning
<azonenberg_work> gruetzkopf: yeah i learned that veeery young
<azonenberg_work> rqou: lol
<TD-Linux> brass usually has ~2% lead and so needs prop 65
<rqou> wtf
<whitequark> awygle: what thefuck is prop 3
<whitequark> *13
<TD-Linux> rqou, disneyland also has prop 65 warnings
<awygle> ugh now i'm just mad about california
<azonenberg_work> "this trumpet may be carcinogenic in california"
<rqou> does that really affect anything if you don't live in Flint?
<gruetzkopf> my chem teacher would have accidents with IPA plus EthOH to make the scratched ones go totally intransparent
<gruetzkopf> -> replace
<TD-Linux> I moved to california intentionally
<Bike> prop 13 is the one that messed up taxes, right?
<rqou> lol
<whitequark> yes
<awygle> whitequark: it amended the state constitution to limit property taxes to 1% and increases in assessed value to 2% annually
<whitequark> yes
<whitequark> exactly
<awygle> it's fucking horseshit
<whitequark> what the fuck
<azonenberg_work> awygle: some rich senator wanted to make sure his investment properties would be more profitable?
<awygle> A large contributor to Proposition 13 was the sentiment that older Californians should not be priced out of their homes through high taxes.[4] The proposition has been called the "third rail" (meaning "untouchable subject") of California politics, and it is not popular politically for lawmakers to attempt to change it.[5]
<TD-Linux> azonenberg_work, more like tons of people ended up with real estate that was suddenly super valuable
emeb has quit [Quit: Leaving.]
<awygle> the thing about california is that you can just buy consitutional amendments there
<awygle> goddammit i'm too mad to be on the internet now. i'm signing off for the night, so long everybody.
<azonenberg_work> lol
<whitequark> lol
<azonenberg_work> the bigger thing about california is the invasive species
<azonenberg_work> That's the main reason i'd never move there... you can't go a block withut seeing nests
<TD-Linux> like me?
<azonenberg_work> they're overrunning the whole state
<azonenberg_work> H. Sapiens
<Bike> are you a species
<sorear> prop 8 isn’t that bad, it’s one of very few we actually got rid of in a finite period of time
<TD-Linux> a vc and coffee-fueled 20something? I'd say so
<Bike> that makes you a member of a species
<sorear> the type specimen is carolus linneaus
<whitequark> lol
<Bike> i mean he actually is, right
<pie_> > italics in ird, wtf
<pie_> irc
<Bike> actually does that mean his body is preserved somewhere in case we need it
<Bike> no it's in some tomb. bullshit
<whitequark> lol
<whitequark> can we make lenin a type specimen
<Bike> i think that's the responsible course of action
<whitequark> >His successor, Vladimir Putin, opposed this, pointing out that a reburial of Lenin would imply that generations of citizens had observed false values during seventy years of Soviet rule.
<sorear> while I have seen unironic references to _Homo sovieticus_,
<whitequark> hmmmm
<Bike> thanks putin
<TD-Linux> thanks putin
<whitequark> thutin
<whitequark> should i write an ONFI applet for Glasgow
<rqou> why?
<whitequark> why not?
<whitequark> there's an SPI flash applet
<rqou> oh this is the old "normal" nand with a 8 bit parallel bus
<rqou> not the weird stuff awygle was working on
<sorear> Flash programmer or simulator?
<whitequark> programmer
<sorear> Hmm, is onfi obsolete now?
<sorear> I guess old interfaces never die *cough SDRAM*
<comietek> how do you deal with isopropyl streaks like on the photos here? https://www.bigmessowires.com/2017/05/25/how-to-clean-a-pcb/
<whitequark> it's not isopropyl
<whitequark> it's flux residue
<whitequark> i use acetone personally, it works way better
<whitequark> doubles as a moisture removal
<comietek> I’m cleaning dirt off old motherboards with isopropyl and it leaves streaks
<comietek> there is no flux involved
<whitequark> there is flux on the motherboard
<whitequark> pure IPA does not leave any residue
kuldeep has quit [Ping timeout: 252 seconds]
<whitequark> that's simply impossible
<zkms> the residue's from something that was on the motherboard or the components soldered to it
<whitequark> either your IPA isn't pure, there's flux on the motherboard, or the dirt isn't well soluble in IPA
<comietek> I have 99.9% anhydrous IPA from MG Chemicals
<comietek> I'm using a Johnson's cotton bud
<sorear> ipa leaves a solid residue below *checks wikipedia* -89 C
<comietek> I can re-clean a given square inch of empty PCB and it will still leave a faint streak
<qu1j0t3> the anhydrous aspect of that probably isn't helping you clean .
<qu1j0t3> comietek: less water doesn't imply better cleaning
<comietek> ok
<whitequark> ^ that too
<whitequark> try this:
<whitequark> clean a part of PCB with acetone
<whitequark> any old acetone that's reasonably pure works
<qu1j0t3> (i wouldn't even buy 99% let alone 99.9% lol)
<whitequark> then clean again with IPA, and observe no residue
<comietek> acetone sounds kinda harsh
<qu1j0t3> indeeed :)
<whitequark> nah, PCBs don't care for acetone
<comietek> qu1j0t3: should I mix it with deionised water?
<whitequark> connectors do
<zkms> also if IPA makes residue appear then just wash with more IPA
<whitequark> so don't clean connectors with acetone
<qu1j0t3> and you know, take reasonable precautions with acetone
<zkms> since you've conclusively proven that the residue components are soluble to some extent in the IPA
<qu1j0t3> don't breathe the shit, etc
<sorear> why is acetone better or worse than ipa. i know they're interconverted in vivo
<comietek> not going to use acetone
<zkms> if they werent they'd be sessile
<comietek> I hate how it smells
<whitequark> comietek: just use a respirator
<whitequark> sorear: it's a much better solvent
<whitequark> well, "better" is subjective. it dissolves more commonly encountered things
<whitequark> and dissolves easier
<rqou> whitequark: not going to recommend something chlorinated? :P
<whitequark> like flux is very soluble in acetone but only marginally in IPA
<whitequark> rqou: not necessary
<balrog> https://www.coursera.org/learn/vlsi-cad-logic looks like this is being offered again, though at a fee
<whitequark> I mean R141b would work great but you don't need it
<sorear> isn't "flux", like, not one thing
<balrog> (and free to audit)
<whitequark> sorear: I mean RMA basically
<whitequark> but all fluxes used in microelectronics are soluble in acetone
<gruetzkopf> i'm mostly using 70% ipa because i happened upon it at zero cost
<TD-Linux> awww yeah my first pcbway boards arrived
kuldeep has joined ##openfpga
egg|egg has quit [Read error: Connection reset by peer]
<whitequark> azonenberg_work: any luck with SWD?
unixb0y has quit [Ping timeout: 240 seconds]
unixb0y has joined ##openfpga
soylentyellow_ has joined ##openfpga
soylentyellow has quit [Ping timeout: 252 seconds]
<azonenberg_work> whitequark: was doing construciton
<azonenberg_work> i did do some work on the network protocol side for improved swd support though
<azonenberg_work> havent pushed that yet
kuldeep has quit [Ping timeout: 244 seconds]
kuldeep has joined ##openfpga
_whitelogger has joined ##openfpga
kuldeep has quit [Ping timeout: 252 seconds]
kuldeep has joined ##openfpga
grantsmith has quit [Quit: ZNC - http://znc.in]
grantsmith has joined ##openfpga
kuldeep has quit [Ping timeout: 240 seconds]
rohitksingh_work has joined ##openfpga
<TD-Linux> if you're super bored you can watch me assemble some ice40 boards https://www.twitch.tv/tdlinux
kuldeep has joined ##openfpga
kuldeep has quit [Ping timeout: 252 seconds]
rohitksingh_wor1 has joined ##openfpga
rohitksingh_work has quit [Ping timeout: 246 seconds]
kuldeep has joined ##openfpga
rohitksingh_wor1 has quit [Quit: Leaving.]
rohitksingh_work has joined ##openfpga
<sensille> mithro: is it a bug of icetime or arachne-pnr? what info is needed for the bug report?
<mithro> icetime
<sensille> done
<mithro> Thanks!
ym has joined ##openfpga
kuldeep has quit [Ping timeout: 272 seconds]
kuldeep has joined ##openfpga
kuldeep has quit [Ping timeout: 264 seconds]
kuldeep has joined ##openfpga
<openfpga-bot> [jtaghal] azonenberg pushed 1 new commit to master: https://git.io/fAolo
<openfpga-bot> jtaghal/master 8693f7e Andrew Zonenberg: Added transport field to protobuf schema
<openfpga-bot> [jtaghal-apps] azonenberg pushed 1 new commit to master: https://git.io/fAol6
<openfpga-bot> jtaghal-apps/master c6ee4ba Andrew Zonenberg: jtagd now sets and checks transport-layer protocol per connection
<rqou> whitequark have you seen this? https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6111272/
<rqou> apparently viral tweets can lead to actual serious studies
<whitequark> rqou: this is kinda obvious
<whitequark> all of it
<whitequark> to anyone who was paying attention
<whitequark> i dont normally make fun of academia but this is some ivory tower shit. "we finally noticed how people actually live, wow"
<rqou> lol
m4ssi has joined ##openfpga
<azonenberg_work> whitequark: this is actually interesting to me
<azonenberg_work> I knew that fish meds were commonly recommended among survivalists for apocalyptic scenarios
<azonenberg_work> But i didnt realize they were consumed by people in first-world countries who didn't want to deal with / couldn't afford doctors
<rqou> uh... this has been making the rounds on birbsite for some time now
<azonenberg_work> rqou: Given how cheap antibiotics are, and how docs hand them out like candy
<rqou> if you have insurance
<azonenberg_work> I thought even the non-insured prices were super low
<azonenberg_work> like $20 or so for a bottle
<azonenberg_work> for most of the common ones
<rqou> maybe it also depends on exactly where you live?
<rqou> i know i've never had this problem
<azonenberg_work> me neither
<azonenberg_work> I also remember on several occasions while traveling, having moved recently, etc
<azonenberg_work> being prescribed some medication for an illness at a pharmacy i didn't usually go to
<azonenberg_work> and, upon being told the price without insurance, not even bothering to dig out my card
<rqou> in quite a few comments people were complaining about the cost of the doctor visit needed to get a prescription rather than the drugs themselves
<azonenberg_work> Ah, ok that might be more of an issue?
<azonenberg_work> i mean i know like weird chemotherapy that 50 people worldwide take is super expensive
<azonenberg_work> But antibiotics are so common i figured they'd be cheap
<rqou> so is insulin
<azonenberg_work> Especially ones like amoxicillin vs something a little more unusual like clindamicin or gentamicin
<rqou> and yet you get periodic "died at exactly 26" stories
<whitequark> azonenberg_work: it's not even "making rounds on the birdsite", i've known this for years
<whitequark> and i'm not american
<whitequark> this has been going on for a very long time; rqou is right, copays and deductibles can be prohibitively expensive to poor people
<whitequark> also the part where you're working "39 hours" and can't afford to miss any for a doctor visit
<gruetzkopf> the kind of 39 hours where 39 is your sleep per week in minutes?
<rqou> no, it's yet another american loophole
<rqou> 40 hours per week is "full time" which activates a bunch of regulations
<whitequark> ^
<rqou> 39 is less than 40
<whitequark> some even go "39.5".
<azonenberg_work> whitequark: That reminds me of back when $WIFE used to work at walmart
<azonenberg_work> overnight when the store was closed, unloading stuff from trucks and putting it on the shelves
<azonenberg_work> Each assigned task had a certain number of hours allocated to it, according to some model of how long an average stocker would take to unload that much merchandise
<azonenberg_work> It was routine for her to get like "18 hours" of work and be expected to complete it on one evening
<azonenberg_work> Because, although corporate had calculated you needed 18 hours to do the work, the store managers were too cheap to hire the necessary personnel to do it :p
<azonenberg_work> Why pay 3 people to work for 6 hours each, when you can pay one person and work them to death for 8?
<whitequark> how unsurprising
<whitequark> I'm reminded of the amount of people who work full time at walmart and still require food stamps to not starve
<azonenberg_work> They wouldn't let her be full time, that meant giving them health benefits etc
<azonenberg_work> idk if she was ever 39 hours, i think it was closer to 30 a week
<azonenberg_work> Just to make sure she didn't accidentally go over the cutoff :p
<whitequark> "full time", as in "spending most of their useful life working or commuting"
<azonenberg_work> Yeah
<azonenberg_work> She lived with her aunt rent-free which is the only reason she wasn't totally screwed financially
<azonenberg_work> And heaven forbid any of the employees try to celebrate Thanksgiving with family... that was the kickoff for the holiday shopping extravaganza and they had Black Friday sales starting as early as like 1600 Thanksgiving day
<azonenberg_work> Anyone who failed to show up for work that day was summarily fired
<whitequark> obviously.
<azonenberg_work> The next year they tried to pull the same stunt but by this point she had met me and we were already planning to move out west
<azonenberg_work> So she told them where to shove it, then went off to the family dinner :p
indy has quit [Quit: ZNC - http://znc.sourceforge.net]
indy has joined ##openfpga
cr1901_modern has quit [Read error: Connection reset by peer]
<whitequark> azonenberg_work: ha, I was wondering what're you doing with the tunnel, until I realized that scanners are pounding the port 2222 continuously
GuzTech has joined ##openfpga
<G33KatWork> any reason why a load_bitstream() in litex for an icestick wouldn't work but a flash works?
<G33KatWork> iceprog -S on the commandline shows the same behaviour
<G33KatWork> the IOs just stay in high-z
<whitequark> if a flash works it seems like the jumpers on the board are in wrong position
<G33KatWork> there's a jumper?
<whitequark> oh wait
<whitequark> does icestick even allow SRAM programming at all?
<whitequark> yeah I think it doesn't
<G33KatWork> I always have to go through the flash?
<whitequark> yeah
<G33KatWork> why can't I just load a bitstream directly into the fpga?
<G33KatWork> sounds weird
<whitequark> the electrical connections aren't right for that on the board
<whitequark> let me double-check
<G33KatWork> ah okay
<G33KatWork> yeah, in the manual are screenshots
<G33KatWork> and they only show SPI programming
<G33KatWork> with the diamond programmer
<G33KatWork> SPI flash it is then, thx
cr1901_modern has joined ##openfpga
<G33KatWork> ah, now I get why the sram loading doesn't work
<G33KatWork> they are using the FTDI to talk directly to the SPI flash
soylentyellow__ has joined ##openfpga
soylentyellow_ has quit [Ping timeout: 252 seconds]
soylentyellow__ has quit [Quit: Leaving]
soylentyellow has joined ##openfpga
Bike has joined ##openfpga
rohitksingh_work has quit [Read error: Connection reset by peer]
emeb has joined ##openfpga
openfpga-github has joined ##openfpga
openfpga-github has left ##openfpga [##openfpga]
<openfpga-github> [openfpga] cr1901 opened pull request #125: Update logtools to bring in Windows ERROR fix. (master...win-log) https://git.io/fAKJB
<cr1901_modern> rqou: Is cmake supposed to install the xc2 tools or do I have to manually run cargo for it to work?
Miyu has joined ##openfpga
<Prf_Jakob> awygle: Hey saw your tweet about a EPC5 SO-DIMM module. Which SO-DIMM connect will you use?
rohitksingh has joined ##openfpga
<awygle> Prf_Jakob: not sure yet. Leaning towards 260-pin
<awygle> There are roughly 200 i/o in the package, if DDR and other on board stuff eat enough pins might do 204 instead
<awygle> Prf_Jakob: why, do you have a preference?
<sorear> Is SO-DIMM just a cheap connector or will it let you connect to existing systems/boards? Which?
<awygle> sorear: cheap and small and high (enough) speed
<sorear> What board is used on the other side?
GuzTech has quit [Quit: Leaving]
<awygle> The idea is to use it as a component of various projects, similar to azonenberg's INTEGRALSTICK
<awygle> It doesn't plug directly into any existing dev board that I know of, or anything
cr1901_modern has quit [Ping timeout: 245 seconds]
wpwrak has quit [Read error: Connection reset by peer]
wpwrak has joined ##openfpga
<Prf_Jakob> awygle: Ah cool, so 1.2mm and 4 or 6 layers?
<awygle> Prf_Jakob: that's the idea yeah
<mithro> awygle: Any chance you could make your SO-DIMM module compatible with the RPi compute module form factor?
<Prf_Jakob> awygle: No real preference as long as I can run 3.3V over the pins. The cost difference of the connectors are negligible.
<mithro> awygle: It would give you a pretty decent existing ecosystem and the ECP5 is more then big enough to run a Linux system....
<Prf_Jakob> mithro: At least for my needs you don't get enough IO if you want to be compatible.
<mithro> Prf_Jakob: In what way?
<Prf_Jakob> mithro: I need at least 60 GPIO.
<mithro> Prf_Jakob: How many pins are on the RPi compute module?
<awygle> mithro: i considered it, i'll probably go that way if i go for a lower pin count connector. i'd basically redefine almost everything "special purpose" as GPIO
<awygle> 200 pins, of which 46 are marked "gpio"
<Prf_Jakob> Preferedly somewhere around 70-90.
<mithro> Prf_Jakob: BTW Have you looked at Ken Hubbard's new minispartan 6?
<awygle> of course lots of that are GND
<Prf_Jakob> mithro: Not enough IO, by just enough.
<mithro> awygle: Another option would be TOFE :-P
<awygle> mithro: too big
<mithro> awygle: I assume the N2 standard thingy doesn't have enough pins, right?
<awygle> mithro: not sure i know what that is
<Prf_Jakob> mithro: Also doing 204 pins means 1.0mm depth, and I think you are stuck with 4 layers then.
<mithro> daveshah: ping?
<mithro> awygle: M.2, formerly known as the Next Generation Form Factor (NGFF),
<daveshah> mithro: pong
<mithro> awygle: Mangling two things together it seems :-P
<awygle> mithro: oh *M*.2. yeah no, not nearly enough. i thought about doing an M.2/mPCIe board a la PicoEVB but decided this would be more generally useful.
<mithro> awygle: Otherwise the DDR3 SODIMM form factor is probably the cheapest connector these days...
<awygle> like i said i'm leaning 260-pin. if there's a 260-pin SO-DIMM compute module pinout to follow though i'd be glad to explore compatibility with that
<awygle> but further development will tell
<awygle> Prf_Jakob: you can definitely do 6L at 1.0mm, it's just more expensive
<Prf_Jakob> awygle: Thanks, my research involved going to various cheap china pcb manfacturing pages and seeing what I could make.
<Prf_Jakob> awygle: I got a KiCad 260 SO-DIMM template if you want to use it.
<awygle> Prf_Jakob: oo yes please
<awygle> tyvm :D
<Prf_Jakob> Template is the wrong word, but board and footprint.
<Prf_Jakob> Np :)
* awygle luxuriates in laziness
* awygle checks the clock, frantically gets ready for work
<Prf_Jakob> <starfox>Good luck</starfox>
cr1901_modern has joined ##openfpga
m4ssi has quit [Quit: Leaving]
emeb has quit [Ping timeout: 272 seconds]
emeb has joined ##openfpga
azonenberg_work has quit [Ping timeout: 272 seconds]
azonenberg_work has joined ##openfpga
azonenberg_work has quit [Read error: Connection reset by peer]
azonenberg_work has joined ##openfpga
GuzTech has joined ##openfpga
rohitksingh has quit [Quit: Leaving.]
oeuf has joined ##openfpga
azonenberg_work has quit [Ping timeout: 246 seconds]
* pointfree Is going to mvre tonight.
<pointfree> First time in a long time.
azonenberg_work has joined ##openfpga
<rqou> so whitequark: i just noticed today that $WORK has a tiny prop65 warning on the front door of the building I'm in
<rqou> does this mean that I'm going to get cancer now? :P
<fseidel> you're required to by CA law
<prpplague> rqou: hehe
<prpplague> rqou: i've had to start putting prop65 notices on all my little breakout boards and such
<prpplague> rqou: real waste of time
<fseidel> by law, or for giggles?
<fseidel> oh
<prpplague> fseidel: i don't have to put it on there for all states, just for CA, so i could simply restrict the sales to exclude CA
<pie_> oyu're required to get cancer by law
<prpplague> hehe
<prpplague> i selected ovarian cancer as my option
<prpplague> they didn't say it had to be an applicable cancer
<fseidel> I really hope that prop65 BS gets pulled in at some point, this is generating so much noise with very little signal
<prpplague> fseidel: not likely
<zkms> i wish it'd actually *incentivise* not using stuff like Pb in potable water systems/faucets (which still happens, you can put lead in brass that touches potable water and if it's under a certain amount you can market it as "lead-free") and avgas, rather than just making people put up those useless fucking signs
<prpplague> fseidel: CA for decades has basically been able to set the standard for a lot of products and warnings, since companies don't want to have one product for CA and another product from the rest of the country
<prpplague> fseidel: they simply comply with the CA standard and ship that everywhere
<prpplague> fseidel: but now with more detailed record keeping and the "net" it's a lot easier to just say "we don't ship this product to CA"
<qu1j0t3> "we only ship to the badlands"
<prpplague> hehe
<rqou> zkms: meanwhile i saw a reddit post (so take it with a grain of salt) where some guy was talking about owning a "classic" sports car that was designed for leaded gasoline and just putting 100LL into the car
<rqou> high octane, still contains lead, and you can "easily" get it
<zkms> D:
<sorear> need to amend the CA constitution to allow legislative modification of initiatives more than 30 years old
<zkms> (oh and there's also Pb in car/home keys, god knows why)
<rqou> leaded brass machines well?
<balrog> zkms: probably to make them easier to machine, because lead is cheaper than the alternatives
<balrog> just like why cadmium was used to coat hand tools for a long time instead of chromium
<gruetzkopf> i still have a bunch of Cd-plated MIL-spec round connectors
<zkms> and TEL was used to make petrol engines run, yeah, i'm unfortunately aware of the phenomenon
<sorear> I don’t see why I should be concerned about Pb in *durable goods*
<zkms> well it poisons people who lick their hands or touch food after touching keys and it poisons locksmiths https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3814711/
rohitksingh has joined ##openfpga
rohitksingh has quit [Client Quit]
<zkms> i certainly wouldn't be averse to paying an order of magnitude more per key if it means no risk of accumulating Pb from keys, i spend enough time around 100LL-burning aircraft as it is and my IQ's low enough, I don't need physical interaction with *more* sources of Pb in my life :|
* qu1j0t3 googles recent federal regulatory changes
<qu1j0t3> Oh, asbestos is back, maybe that is an antidote for lead?
* prpplague wonders if there are prop65 warnings on silicone breast implants
<gruetzkopf> the railway industy likes leaded solder very much
<rqou> tbh for most hobbyists you probably pick up negligible lead from soldering
<awygle> space <3s leaded solder
<TAL> asbestos is really a problem, a lot of specialized gaskets which had that stuff in it can't be produced/sold anymore
<TAL> but tbh, how dispose that stuff properly when you even't realize that stuff was really a commodity at some point
<rqou> I'm sure azonenberg_work can give you a lot of "thoughts" about asbestos abatement :P
Miyu has quit [Ping timeout: 272 seconds]
<zkms> awygle: i saw something from Sandia that talked about lead-free solders and nuclear stockpile components >_>; https://prod.sandia.gov/techlib-noauth/access-control.cgi/2010/106586.pdf
<prpplague> zkms: oh that is interesting
* prpplague puts that in his reading queue
<awygle> yeah cool
<prpplague> zkms: if you are interested in such things, it's really fun to search the area around sandia with google maps and bing maps
<prpplague> zkms: lots of interesting things
<zkms> it's good to know that Sandia might come out with an RoHS-compliant nuclear warhead some day tbh
<prpplague> hehe
<prpplague> zkms: because you want your nuclear warheads to be environmentally friendly
<prpplague> This warhead is *Green* approved!
<awygle> on the plus side if SAC305 turns into wet sand after five years the military industrial complex will fix it
<prpplague> zkms: sandia labs... roads to nowhere
<kc8apf> TEL in aircraft fuel is completely unnecessary. Just stop using terrible engine designs
<prpplague> zkms: this one is my favorite - https://www.google.com/maps/@35.0051614,-106.4885893,122m/data=!3m1!1e3
<prpplague> zkms: car park on the side of a mountain with concrete entry way
<Bike> M855A1 projectiles for 5.56×45mm NATO rifles replace traditional lead alloy cores with an environmentally friendly copper core with a 19-grain (1.2 g) steel "stacked-cone" penetrating tip.
<zkms> tbh getting lead out of small-arms ammo isn't a bad idea either, if the projectile isn't fully enclosed then it makes vaporised lead appear around the shooter and that shit is no good
<Bike> i'm not complaining. nor about the nukes. but it's funny
<sorear> Is a RoHS nuke like a LEED platinum parking garage
<Bike> the parking garage will probably get more use than the nuke, which exists mostly to be scary before being badly disassembled in fifty years
azonenberg_work has quit [Remote host closed the connection]
azonenberg_work has joined ##openfpga
<qu1j0t3> I question how environmentally friendly it is to waste copper (or any resource) on projectiles, but *shrug*
<azonenberg_work> qu1j0t3: i think this is mostly the army tired of getting in trouble with the EPA for contaminating practice range areas
<azonenberg_work> For every round fired in combat they fie tens of thousands in training
<prpplague> azonenberg_work: exactly
<qu1j0t3> azonenberg_work: *nod*, but thinking bigger picture...
<qu1j0t3> azonenberg_work: yeah most ranges are effectively superfund sites aiui
<azonenberg_work> qu1j0t3: Only the poorly managed ones
<azonenberg_work> The club i'm a member of has a sand backstop shielded from rain
<prpplague> 25 years ago most of the firing ranges at Ft.Knox had moved to a custom config where the backdrop was sand, and they a steel-mesh net under it
<azonenberg_work> that's spiked with lime periodically to keep the pH highly alkaline
<prpplague> every 3 months they'd life the mesh up and strain out all the rounds
<qu1j0t3> azonenberg_work: There's a suggestion they're all poorly managed :)
<azonenberg_work> So the lead won't dissolve in groundwater and leach out
<azonenberg_work> We dont shoot as much as fort knox, but they do excavate every year or two and filter out the rounds
* prpplague procrastinates doing real work and reads about GAL logic structures
<prpplague> thanks azonenberg_work and awygle , now i have a year long project to occupy my free time
* prpplague glares
<sensille> GAL20V8?
<awygle> what did *i* do lol
<prpplague> sensille: 16v8
<prpplague> awygle: you, azonenberg_work and tinyfpga got me thinking about a small disintegrated FPGA build
<prpplague> awygle: some LUTs and CLBs
<awygle> ah
<awygle> the tinyfpga ex has hyperram, right?
<TD-Linux> a GAL dump to Verilog compiler would be cool
Bike has quit [Ping timeout: 252 seconds]
GuzTech has quit [Ping timeout: 272 seconds]
Bike has joined ##openfpga
azonenberg_work has quit [Quit: Leaving.]
azonenberg_work has joined ##openfpga
azonenberg_work has quit [Ping timeout: 240 seconds]
azonenberg_work has joined ##openfpga
<prpplague> i think it's time for a beer
* prpplague 's head is exploding after reading datasheets all afternoon