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.
<bofh> LOL
* UmbralRaptor is reminded again that free oxygen is monstrously corrosive, only overshadowed by fluorine. >_>
<UmbralRaptor> Anyway, radar and 3d modeling of 16 Psyche http://outreach.naic.edu/ao/blog/radar-observations-and-shape-model-asteroid-16-psyche
icefire has quit [Read error: Connection reset by peer]
<bofh> UmbralRaptor: ooh, is that GSSR?
<UmbralRaptor> GSSR?
<bofh> Goldstone Solar System Radar
<bofh> !acr -add:GSSR Goldstone Solar System Radar
<Qboid> bofh: I added the explanation for this acronym.
<UmbralRaptor> Apparently Goldstone's 70 m dish wasn't big enough, so they used a 300 m.
<UmbralRaptor> "a 300 m" as if there's more than one. >_>
<bofh> oh, Arecibo?
<UmbralRaptor> Yep!
<bofh> https://cneos.jpl.nasa.gov/news/news199.html ahh right, GSSR was imaging Florence recently
<bofh> https://cneos.jpl.nasa.gov/images/news/florence.p5us.1Hz.s382.sep01.gif oh wow that's quite the chaotic tumbling there
<bofh> fwiw now in Goldstone it's possible to array receiving dishes, so technically while #DSS14 was the only one xmitting, #DSS24, #DSS25, #DSS26 & Canberra's #DSS36 were all receiving.
<UmbralRaptor> wait, 2 moons? o_O
<bofh> scratch that last one, apparently #DSS36 was a second transmitter: https://mobile.twitter.com/CanberraDSN/status/903550425156042752
<kmath> <CanberraDSN> Our #DSS36 dish is also pinging #AsteroidFlorence with a @CSIROnews' antenna in Narrabri @CSIRO_ATNF acting as the… https://t.co/Z0RRVTPV6S
<bofh> yep!
<Fiora> !wpn
* Qboid gives Fiora a Lambert fountain pen
<soundnfury> !wpn bofh
* Qboid gives bofh a dubious ring
<egg|zzz|egg> !wpn Fiora
* Qboid gives Fiora a standing wave ratio
<soundnfury> !wpn egg|zzz|egg
* Qboid gives egg|zzz|egg a Pauli redundant compactification
<soundnfury> (also, I don't believe your zzz is accurate)
Technicalfool is now known as TechnicallySleeping
icefire has joined #kspacademia
kmath has quit [Read error: -0x1: UNKNOWN ERROR CODE (0001)]
APlayer has joined #kspacademia
<APlayer> Hi!
kmath has joined #kspacademia
<kmath> <DanNerdCubed> Fun Fact 1: Fire Alarms that have low battery mostly go off at night because of the colder temperatures.
<bofh> Iskierka: that's a thing.
<bofh> I thought it was a bad joke at first,but nope: "Discharge of Zn electrodes also is severely curtailed at reduced temperatures by reduced dissolution current and
<bofh> the premature onset of passivation."
<bofh> and most smoke alarm batteries are Zn/MnO2 electrode alkaline batteries with Zn oxidation forming the bulk of the potential.
<Iskierka> Most batteries lose performance at low temperatures, so I figured likely true either way
<Iskierka> even if it's small it's probably larger than one day's drain
<bofh> yep, and because of the peculiarities of alkaline batteries, this results in both a voltage drop and a max possible voltage for a given current draw reduction.
<kmath> <bofh453> I thought this was a joke but nope, "discharge of Zn electrodes is severely curtailed at reduced temperatures":… https://t.co/JH0ZhDgQhZ
<egg|zzz|egg> !wpn bofh
* Qboid gives bofh a degenerate coilgun
APlayer has quit [Ping timeout: 204 seconds]
<soundnfury> !wpn egg|zzz|egg
* Qboid gives egg|zzz|egg a gatling NaN
<Ellied> NaNNaNNaNNaNNaNNaN
APlayer has joined #kspacademia
<soundnfury> (batman?)
<APlayer> I think I acquired a treasure. "Fundamentals of Astrodynamics"
<soundnfury> ooooh
<APlayer> Covers pretty much everything I ever heard of, anyway, and a good place to look things up
<soundnfury> (is there a sequel "Harmonics of Astrodynamics"? ;)
<APlayer> And the best part is, the book (new, from a bookstore) was 15 €, I think about $ 10 - 12
<APlayer> Uh, no, $19 - 22
<bofh> a gatling NaN? so like an IIR filter and a single NaN? :P
<Ellied> ugh, my hard drive has the dreaded { DRDY ERR } { UNC }
<Ellied> I think it means in this case that the SATA connection is physically dubious and occasionally interrupting reads, so it's not (yet) fatal, but tzhh
<soundnfury> Ellied: I feel your pain :S
<Ellied> there's no SATA cable in this laptop; the drive plugs straight into a socket on the motherboard. I think it might have just been exercised too many times during the crazy number of repairs I've done
<Ellied> well, not number, just the things that were replaced. (motherboard and keyboard)
<Ellied> what does DRDY mean anyway? I can't help but read it "dirty" but presumably it's an acronym?
<bofh> Data Ready
<bofh> make sure you're operating in AHCI mode in the BIOS/UEFI if that's configurable.
<bofh> also run smartctl and let me know if any of the variables are off
<bofh> 5/186/187 iirc are especially bad if nonzero
<APlayer> Uh, bofh, you seem to be a BIOS expert?
<bofh> "expert" okay that's giving me *way* too much credit
<APlayer> If you could help me with a small annoyance I have for some time... A while back I added a hard drive to my computer, and it all works well, except that I have an error message "No RAID disks installed" or something during every boot. It does not crash or anything, it's just that I briefly get this extra screen.
<bofh> my guess: you have a hardware RAID controller or a motherboard capable of hardware RAID, and it is inexplicably loading an option ROM at boottime because it thinks that there is a h/w RAID array in the machine, however the ROM knows better and sees this is lies, so it just hands off control back to the motherboard and continues boot.
<bofh> there's prolly an option somewhere in BIOS that's set incorrectly.
<APlayer> How would this option look like? I tried disabling RAID or something like that, but IIRC it didn't boot at all then. I really need to see that menu again, it was a while ago when I tried to solve that
<bofh> So I honestly wouldn't bother touching it, it's harmless (much like my laptop which blinks a "network cable not connected!" message every boot since I prioritize PXE over physical ethernet over local HDD...)
<bofh> But chances are there's an option in the ROM itsself to not display it or just bypass checking for RAID arrays. Can't say anymore without knowing first both what the BIOS firmware revision is *and* the make/model of disk/RAID controller on the motherboard.
<APlayer> Okay then, I guess I will just leave it at that. I've had no other problems with it, it was just a curious little thingy that started to appear.
<APlayer> Thanks for the clarification, though!
<Ellied> bofh: 5 is zero, but I don't see 187 or 187. You don't mean 196/197, right?
<Ellied> is there a way to tell in userland whether it's in AHCI mode? I forget what it was in the setup util.
<Ellied> nvm I'm just gonna reboot and check
<bofh> I actually meant 184/187/188/196/197/198/201, yeah
<bofh> also 10
<bofh> (some of those may not exist)
<bofh> (depends on the drive model)
<bofh> 5 being zero is a very good thing.
<Iskierka> warning: warning may not warn
<bofh> LOL
<Ellied> 10 is also 0
<Ellied> 196 is not zero, it was 70 before reboot and now it's 77. 197 is 712.
<egg|zzz|egg> <bofh> 5 being zero is a very good thing. <<< what is so good with characteristic 0
<egg|zzz|egg> s/0/5/
* egg|zzz|egg not awake
<Qboid> egg|zzz|egg meant to say: <bofh> 5 being zero is a very good thing. <<< what is so good with characteristic 5
<Iskierka> uuuuh those sound bad
<Iskierka> 197 is number of "unstable" sectors waiting to be remapped due to fatal read errors
<Ellied> I remember getting DRDY ERR issues on a perfectly good drive that completely stopped it from booting, but it turned out to be a problem with the then-bleeding kernel version 4.5
<Iskierka> but I guess that could be caused by faulty connector
<Ellied> there were a lot of infuriating archlinux threads that were like "uh that's definitely 100% a hardware problem" "no dude I KNOW it's not" "well then you're stupid"
<Ellied> 199 - is that related to connector issues? I remember reading about CRC Errors being that. It's zero.
<Iskierka> new hardware can be faulty but you can always cross-reference with other OSes
<Ellied> This laptop has been through some kind of Unspecified Severe Thrashing Event (it needed a new mobo and keyboard) so it would not surprise me if the HDD is also damaged
<Iskierka> yes, but determined by another system that it could possibly fail itself when disconnecting?
<Iskierka> but zero sounds unlikely for intermittent problems
<Ellied> previous owner claims it died during windows 10 update, dunno if I buy that that was the only thing
<bofh> 14:44 <@Ellied> there were a lot of infuriating archlinux threads that were like "uh that's definitely 100% a hardware problem" "no dude I KNOW it's not" "well then you're stupid" <- never change, arch forums. actually please fucking do.
<bofh> Ellied: all of those should be zero iirc
<Ellied> gr8
<bofh> egg|zzz|egg: thank you for reminding me about all those times in my linalg class we had to add a clause of "except over base fields of characteristic 2"
<bofh> (there were many)
<Ellied> the error message I keep seeing is READ FPDMA QUEUED
<Iskierka> is it HDD or SSD?
<Ellied> HDD
<Iskierka> then I think 198 should exist so that's that at?
<Ellied> I still don't own any SSDs
<Ellied> 198 is zero but it also says Offline in the 'updated' column
<bofh> Iskierka: yeah I did say "some of those may not exist, depends on the drive model" for a reason :P
<Iskierka> so whether the drive is getting uncorrectable errors writing is unknown. I'm not sure if 77 reallocations is bad for a HDD - would be unremarkable (but the start of the end) for an SSD but might be a lot in HDD
<Ellied> iunno what it is on my other computers, don't have them handy bc I'm on a vacation for the weekend
<Ellied> (regretting not bringing the toughbook already; I hope this new thing lasts the weekend without dying)
<Iskierka> my 2 GB drive is at 0/0/0/0 for 196/197/198/201
<Ellied> I have to wonder if this is just some arcane bug in btrfs
<Iskierka> ... my SSD is 0 on all of them also but it's claiming the parsed value for temperature is currently 66 (??), with 54 as the highest (?????), and a raw value of 3600140022
<Iskierka> so my SSD is, in order: absurdly hot for one, going backwards in time, and hotter than the core of the sun
<bofh> so those temp values are often miscalibrated and insanity
<bofh> also 66degC is not at all hot for an SSD imo
<Ellied> brb, writing proposal for SSD-fusion reactor
<Iskierka> SSDs are low power and it's not being baked by another piece of hardware so 66 C shouldn't be the case
<bofh> on a tangent I still love how after I cleaned out a metric fuckton of dust from my heatsink/fan/heatpipe combo my normal CPU temperature went from ~100degC idle to ~60degC idle
<Iskierka> I probably need to clear my filters
<bofh> (in case you are wondering how hot it got under load, the answer is "I disabled most of the thermal shutdown mechanisms because they were kicking in too often")
<bofh> (but ~140degC wasn't uncommon)
<Iskierka> ... temperature is definitely nonsense as the HDD thinks temperature currently 117, worst 106, raw 23
<Iskierka> charting the actual temperature sensor system says HDD is at 35 and SSD at 34
<Iskierka> 35 and 34 are definitely in C and you can't convert any of the smart values from F to C to get those numbers either
<Iskierka> so yeah, they're miscalibrated insanity
<Ellied> it's probably something absurd like °C/2^n
<Ellied> or I guess it could just be a thermistor and an ADC, no? So the raw value is just ADC bits and doesn't mean anything without calibration
<Iskierka> that would make sense of the SSD raw value but not the nonsense parsed values that don't correlate in C or F
<Ellied> although that's a weirdly large number of bits for an ADC
<APlayer> 0 * °C + randomInt()
<bofh> LOL
<Ellied> say, bofh, are you aware of any FPGAs that come in remotely hand-solderable packages? DIP, SOIC, even TSSOP, just not QFP/QFN/BGA
<Ellied> ideally with open-source programming tools but I'll take whatever I can get
<Ellied> my prof really wants to get into small-scale FPGA projects but we can't find anything we have the ability to solder and development boards are generally overkill for what we're looking for
<Ellied> we've got some of the crazy giant dev boards that look like gaming PC motherboards, but it would be really cool to have something small and inexpensive for little projects, like what you'd use a cheap MCU for
<Ellied> my cursory googles have just found threads saying "yeah no, just use an MCU"
<egg|zzz|egg> bofh: in my linalg class I was constantly telling the prof "that doesn't work in infinite dimension" :-p
<bofh> egg|zzz|egg: dude so we literally begain each statement with "let V be a FDVS and..."
<bofh> (Finite-Dimensional Vector Space)
<bofh> I really wish I still had my notes for that course, it was great. Featuring things like deriving Cayley-Hamilton on an assignment and going thru Jordan Canonical Forms and touching on the exterior algebra to properly explain what a determinant is.
<egg|zzz|egg> bofh: we didn't have this abbreviation, but eventually the prof got the habit of saying "and let V further be finite-dimensional" or somesuch (in german anyway) while nodding in my direction :-p
<bofh> LOL
<bofh> 15:13 <@Ellied> say, bofh, are you aware of any FPGAs that come in remotely hand-solderable packages? DIP, SOIC, even TSSOP, just not QFP/QFN/BGA
<bofh> ask whitequark when she returns, not my domain really.
<bofh> (she'll be back once digitalocean unfucks things)
<Ellied> alright
<Ellied> also, she? did I miss a memo? Last I heard it was they
<egg|zzz|egg> Ellied: use the synapsid website https://mastodon.social/@whitequark/16038885
<Ellied> ahh I have not into mastodon yet
<Ellied> perhaps I should change that.
APlayer has quit [Ping timeout: 204 seconds]
<Iskierka> that definitely happened on old facebook account. I think it might still be floating around with deadname but labelled she
<Ellied> ok there I'm now on the mastodon website
<egg|zzz|egg> Ellied: fwiw I'm eggrobin there, but I hardly actually write anything (tbh I hardly post on the diapsid website either)
* Iskierka neither
<Ellied> egg|zzz|egg: I'll follow you anyway
<Iskierka> ... wait how can a chemical be 4 flammability and an oxidizer?
<bofh> organic peroxides are fun like that
<bofh> they combine fuel *and* oxidizer into one convenient unstable molecule
<Iskierka> so it's basically a warning "it's highly flammable and you can't stop it"
<bofh> yep have fun!
<Ellied> god
<bofh> I should follow people on synapsid website
<Ellied> I'm also @diodelass over there, looks like iskie already found me
* Iskierka had to figure out what it was on twitter because couldn't remember current name switch
<bofh> oh is Iskierka on twitter?
<Ellied> I'm not a huge fan of anything whose storage requires constant, uninterrupted power or it blows up.
* Iskierka wonders what the NFPA for antimatter would be
<Iskierka> bofh: I think @Kate_ZJ. If avatar is a space unicorn it's probably correct
APlayer has joined #kspacademia
<Ellied> Health:0, Flammability:0, Instability:5
<bofh> oh that's you
<bofh> followed
* Iskierka has poked around twitter threads occasionally but is uncertain where she might've been noticed
* APlayer thinks that there must be some internet standard to indicate gender
<Iskierka> Reading of typing style
<APlayer> Iskierka: How often have I mentioned you with "he"?
* Iskierka shrugs. Lots of logs disappeared so couldn't easily find out either
<soundnfury> APlayer: it's fine, "he" is a gender-neutral pronoun
* e_14159 hands APlayer a "they".
<Iskierka> subconscious enough that I'll bet lots of trans people pre-coming-out probably have a measurable bias in typing style
<Iskierka> ... actually I'm just gonna sanity check because there's lots of things I can't find grepping logs but I didn't check if I've zipped older ones or such
<Iskierka> hm, nope, they were lost, nevermind
<bofh> Iskierka: *is* there a measureable difference in typing style between genders?
<APlayer> Why would one zip logs? When they reach the need-to-size ratio where it makes sense to zip them, I'd usually delete them :P
<e_14159> What did you want to grep for?
<soundnfury> bofh: you could probably find that out by trying to train a neural net to classify them
<Iskierka> I'm pretty sure I've seen multiple studies before finding differences
<Iskierka> I mean, people found out JK Rowling was the author of a thing by training a neural net on harry potter
<APlayer> LOL
<APlayer> But author comparison is a different thing
<Iskierka> it is but it worked in an entirely different genre using only HP as reference data, IIRC
<APlayer> You usually look for frequently used phrases, sets (tuples/triples/quadruples) of words and stuff
<APlayer> Don't even need a neural netwrk for that
<Iskierka> e_14159, it was the other day I was first wanting to find imgur links from maltesh in kspo, to look at his spacebus designs. then wanted to find previous discussion with UmbralRaptor in unofficial on the energy output of free neutronium
<e_14159> Iskierka: Ah, sorry, I don't have logs for these.
<bofh> Iskierka: so that sounds way easier than classification based on some categorical population difference imo.
<Iskierka> hence wanting to look in my logs :p
<bofh> (authorship determination)
<soundnfury> APlayer: yeah, you could just use a markov model and Bayes it
<bofh> APlayer: I zip logs b/c logs are handy as shit to have and even DEFLATE does a good job on tet
* UmbralRaptor is summoned.
<bofh> Iskierka: thanks
<APlayer> For compressing text, I'd use BWT, LZW and ANS algorithms, not DEFLATE
<e_14159> But both the markov model and NN would probably be really sensitive to genre-changes.
* Iskierka lost large chunks of logs to an IDE nuking the home directory and started doing backups since then
<Iskierka> hence mine being lost and not in a zip where I'd prefer if they grew too much
<bofh> APlayer: DEFLATE and LZW are pretty similar in terms of compressibility in my experience
<bofh> I mean the former is just LZ77
<e_14159> Hm. I wonder whether you couldn't train it to produce (author, genre)-pair embeddings. Assuming you can find a nice representation for genres.
* Iskierka would probably just use whatever default zip can easily be run through zip tools to grep within
<APlayer> DEFLATE is LZ77 followed by Huffman
<soundnfury> from the abstract: "content-free text, such as found in many Internet applications" (previously) "such as Twitter [and] Facebook"
<soundnfury> (meow!)
<Iskierka> "Experimental results show that
<Iskierka> SVM outperforms the Bayesian-based logistic regression and
<Iskierka> AdaBoost decision tree for identifying author’s gender from
<Iskierka> a given text document."
<Iskierka> ... okay thanks pdf
<APlayer> If you use BWT before running DEFLATE on a text, though, you'll most likely get much better results
<Iskierka> "We find that there are significant differences between men and women in personal writings such as e-mails, and gender differences also exist between authors of news articles even though neutral language is dominant there."
<bofh> yeah but it doesn't say what those features *are* unless I'm mis-reading things
<Iskierka> towards the end it is saying what feature sets get most accuracy independently
<bofh> APlayer: yes but at that point why not just skip from BWT to Huffman/arithcoding (after some simple transform first to make it more amenable to the entropy coder)
<Iskierka> but with only a clance
<bofh> yeah, but it doesn't say which features in the set are important
<Iskierka> *glance I don't see features stated
<bofh> unless I'm not reading it carefully enough
<e_14159> Just use an RNN-based autoencoder. Who needs lossless encoding?
<bofh> okay Feature-Set Selection has some promising citations
<Iskierka> meanwhile whitequark and fiora have an interesting twitter convo on air being narcotic
<Iskierka> (apparently reaction times improve in a helium-oxygen atmosphere versus nitrogen-oxygen)
* Iskierka assumes that was tested in a way where air resistance couldn't make tiny differences
<APlayer> The point of BWT is to make algorithms such as the LZ family or RLE or stuff more efficient, by moving text around to create more equal characters in sequence. It enhances the aforementioned algorithms
<Iskierka> does it enhance beyond just increasing the setting to equal total processing time?
<APlayer> And then you run an entropy coder over it in addition to that, to compress as much as possible.
<APlayer> Iskierka: It basically exploits that there are recurring patterns in every language and exposes them in such a way that, say, LZW can pick them up.
<egg|zzz|egg> !wpn UmbralRaptor
* Qboid gives UmbralRaptor a reversed frequentist otter
<Iskierka> this sounds like you're throwing a lot of processing on top making it hard to search without decompression to drive when it's already very compressible data
<egg|zzz|egg> !wpn Iskierka
* Qboid gives Iskierka a Soviet exception
<egg|zzz|egg> !wpn e_14159
* Qboid gives e_14159 a silver gazebo which vaguely resembles an insulator
<APlayer> bofh: Anyway, Huffman is not the best entropy coder, though a relatively fast one. Arithmetic coding is better in terms of compression, but pretty slow. ANS is almost as fast as Huffman and almost as good as Arithmetic coding, and my personal favorite.
<Fiora> "switch from BWT to huffman" seems like a weird phrasing
<Fiora> since one is an entropy coder and one is not
<APlayer> ^
<Iskierka> in soviet russia, exception is fatal to you
<APlayer> Nobody expects the soviet inquisition. :P
<Fiora> also, "arithmetic coding" is a very general category, imo i'd include ANS/etc in that general category but even then there's an infinite number of context schemes
* UmbralRaptor measures the p-value of the otter.
<egg|zzz|egg> !wpn Fiora
* Qboid gives Fiora a separable bolter
<APlayer> Arithmetic coding is a very specific algorithm. I think you confused it with assymetric numeral systems
<Fiora> eh? not really...
<Fiora> there's tons of ways to implement arithmetic coding
<Fiora> like, literally anything that represents the data as a continued decimal (or a moral equivalent)
<Fiora> i mean, even personally i've worked with multiple distinct arithmetic coders
<APlayer> But arithmetic coding does a very specific thing, regardless of implementation?
<Fiora> er, not really
<Fiora> i don't think arithmetic coding is even a well defined term
<Fiora> like, i mean, to start with, you have binary vs non-binary coding
<Fiora> adaptive vs non-adaptive coding
<Fiora> an infinite number of possible context schemes
<Fiora> different schemes for context transition tables
<Iskierka> is it supposed to be all in russian and not get a translate option?
<Fiora> different schemes for computing the new range_lps values
<APlayer> Are we talking about the same thing? Which can also be called Range Coding?
<Fiora> yes
<bofh> Fiora: I think by "arithmetic coding" APlayer is specifically thinking of the MQ/QM coder
<Fiora> i have no idea what that is
<bofh> at least that's my suspicion
<bofh> the arithcoder used by JPEG
<Fiora> which is just like... one of a gazillion different arithcoders
<APlayer> I didn't even know JPEG used lossless compression algorithms
<Fiora> even just -personally- i've looked at.... lets see
<APlayer> :D
<Fiora> uh... of course it does
<Fiora> literally every compressed format uses lossless compression algorithms
<Fiora> that's what an entropy coder is
<Fiora> you can't compress data without a lossless compression algorithm
<bofh> also it seems a bit odd to consider an arithcoder that differs only in the transition tables a separate algorithm to me
<APlayer> No idea about the JPEG algorithm, really
<bofh> but I guess that depends on how you categorize things I s'pose
<Fiora> bofh: i mean, i consider them all features of how you choose to define your particular arithcoder
<Fiora> insomuch as they make it incompatible
<Fiora> APlayer: dct -> quant -> delta DCs (iirc) -> custom huffman tables
<Fiora> optional arithcoding nobody uses
<bofh> Fiora: so I guess it's more I consider some of the properties you listed above "arithcoder properties" and some of them "context model/scheme properties", since I just realized I sort of separate out the underlying coder and the context model in my mind pretty heavily. apologies.
<APlayer> Why is it even there? JPEG is supposed to be very small and very fast, arithcoding defeats the latter
<bofh> and yep, I still need to get off my arse and implement support for arithcoded JPEG in libavcodec
<bofh> but the JPEG spec is a nightmare, the reference lib unreadable, and nobody seems to care about its lack so ¯\_(ツ)_/¯
<Fiora> APlayer: whoever said it was supposed to be "small and fast"?
<Fiora> neither of them were really design goals
<Fiora> after all, it doesn't use a fast DCT
<Fiora> nor does it use a fast entropy coding scheme (just stock huffman, not something optimized for fast decoding at the expense of simplicity)
<Fiora> i mean, jpeg is too old to really *have* such things as design goals; it predates most modern image compression schemes
<Fiora> i don't think fast DCTs had even been invented yet, or at least not used in practice
<bofh> like putting it in perspective, JPEG is (approximately, roughly, with a standardized DCT) equivalent to an MPEG-1 iframe.
<APlayer> Well, as I said, I have no clue about JPEG
<Fiora> then don't assume things about it ;-)
<bofh> nobody uses MPEG-1 for video because it's *that old* (and people only use MPEG-2 still b/c broadcast TV/DVDs, as far as I can tell).
<bofh> ^
<Fiora> I mean, MPEG-2 is hardly different from MPEG-1, it just adds interlacing and more resolution modes
<Fiora> though I don't remember if MPEG-1 had DC prediction (I think it does?)
<Fiora> not sure what MPEG-1 did for DC coding either.
<APlayer> Well, I've got to go. Be back soon.
<Iskierka> MPEG-1 has limited resolution choice?
<bofh> does JPEG have DC prediction? I actually don't remember
<Fiora> i think it codes DCs as deltas
<Fiora> i might be wrong
<Fiora> "The previous quantized DC coefficient is used to predict the current quantized DC coefficient. The difference between the two is encoded rather than the actual value. "
<Fiora> okay yes
<Fiora> in coded order
<bofh> ahh. okay.
<bofh> "Due in part to the similarity between the two codecs, the MPEG-2 standard includes full backwards compatibility with MPEG-1 video, so any MPEG-2 decoder can play MPEG-1 videos." ahh, okay
<bofh> Iskierka: so it supports up to 4095x4095 no problem, but there's a sub-profile that only requires 352x240, 352x288, or 320x240 resolution support.
<Iskierka> ... old video standards are weird
<bofh> I wonder if MPEG-1 has any GOP limit.
<Fiora> i don't think so
<Fiora> such limits are usually a function of implementations like DVB
<bofh> right, that makes sense.
APlayer has quit [Ping timeout: 383 seconds]
<bofh> well, DVB-MPEG, since "DVB" itsself just defines the physical and link layers
<Iskierka> okay going back through that thread which I only saw in feed at postulating anasthesia operating at something above consciousness, that thread took some crazy turns
<Fiora> DVB-MPEG is a subset of DVB ;-)
<Iskierka> (and xenon is fucking weird)
<bofh> touché
<bofh> been reading too much DVB-S specs lately, that's my excuse. :P
<bofh> (GOES rebroadcast is just DVB-S2 except the payload is, like, magnetometer and EUV imager data instead of TV)
<Iskierka> as well as the anasthesia and everything else mentioned, it also interrupts the mechanisms for apoptosis, and is used to significantly increase survival rate of newborns who don't immediately start breathing as healthy cells don't mistakenly do it from oxygen deprivation
<Fiora> huh, wait, what??
<Ellied> wait, xenon?
<Fiora> i though anesthesia primarily affected the brain
<Fiora> or are you saying it interrupts apoptosis -everywhere-?
<Iskierka> It was described as xenon when I found out about it but it could be a compound (but that would be even scarier)
<Fiora> xenon is an anesthetic, yes
<Fiora> but "anesthetic" almost (90% of the definition, at least) just means "lipid soluble gas, passes blood/brain barrier"
<Iskierka> The apoptosis mechanism is in the mitochondria so I assume everywhere, yes. the effect observed of increased survival rate after oxygen deprivation is a real one
<Iskierka> no idea if it's shared by other anasthetics
<Fiora> i believe all anesthetics have essentially the same effect
<Fiora> and, the same mechanism, most likely
<Fiora> apoptosis: are there -other- ways of interrupting it? i'd assume if we know how it works, one could concoct chemicals to block it that aren't anesthetics
<Iskierka> the way I saw it described is the apoptosis triggers on oxygen reintroduction, but assumes it's a failure of the cell (which is actually healthy), rather than of the supply, so by interrupting it you can prevent that and have a healthy recovery as cells are largely intact
<Iskierka> I don't know more details but it could be something to search
<Fiora> another interesting thing to note, actually, related to that
<Fiora> whitequark explained this one to me
<Fiora> so do you know why methanol is so toxic, and why 10mL can blind you?
<bofh> so I could imagine something that affects lipid membranes gumming up this mechanism specifically
<bofh> Fiora: go on
<bofh> "opening of the mitochondrial permeability transition pore can greatly reduce ATP production, and can cause ATP synthase to begin hydrolysing, rather than producing, ATP." LOL OOPS
<Fiora> so if I recall correctly
<Iskierka> http://www.sciencedirect.com/science/article/pii/S0940960212001367 this seems to be comparing it to another GA in a more specific circumstance
<Fiora> basically, mitochondria need a continual system to stay alive. the cell continually recycles the mitochondria outwards
<Fiora> and moves ATP the other direction
<Fiora> it needs energy to run the process
<Fiora> so if you interrupt it, the cycle breaks
<Fiora> i.e. there's no ATP to start it up again
<Fiora> the longer the cell, i.e. the longer the path between nucleus and mitochondria, the more susceptible it is, because the longer the transport path is
<Fiora> the optic nerve has the longest cells
<Fiora> so it dies first
<Fiora> this is effectively a case where a *temporary* interruption of something causes *permanent* cell death because a continual operating cycle is needed merely to continue operation
<Iskierka> but then this one says isoflurane causes apoptosis that xenon can prevent
<Ellied> Fiora: huh, that's interesting
<Fiora> i.e. "it needs ATP to make more ATP"
<Iskierka> Clearly a system that someone would indicate an intelligent designer
<Fiora> nah, a remotely intelligent design would be able to turn off without permanent death ;-)
<Fiora> imagine a computer that could never be turned off or else it would never go on again
<Ellied> yeah, I'd use that as a counterexample
<Fiora> though, one thing I wonder
<Iskierka> And clearly /s needs to be used? :p
<Fiora> there must be some types of cells out there, even bacteria etc
<Fiora> that do have the ability to temporarily turn off
<Fiora> i mean, like, for example, there's uhhh what are they called. tardigrades
<Iskierka> tbf tardigrade survival rate is often significantly overstated. It's just also significantly non-zero
<Fiora> clearly though it -is- nonzero
<Fiora> i.e. you CAN turn off and turn back on
<Fiora> whereas i don't know if any human cells can do that
<Ellied> wait, so how does methanol break that cycle? I got lost up there somewhere
<Iskierka> I do recall a case of a woman frozen by accident for like a few hours who mostly or completely recovered. She might have mutations towards making that more viable?
<Iskierka> and yes methanol was never mentioned :p
<Fiora> Methanol is converted to formaldehyde via alcohol dehydrogenase (ADH) and formaldehyde is converted to formic acid (formate) via aldehyde dehydrogenase
<Fiora> Formate is toxic because it inhibits mitochondrial cytochrome c oxidase
<Ellied> ahhhh right
<Ellied> thank you
<Fiora> the specific mechanism is relevant to why it causes blindness at lower doses than it causes death
<Fiora> whereas most toxins don't do that.
<Ellied> makes sense.
<Iskierka> so as well as not ingesting it, avoid getting it in the eyes when using as cleaning agent
<egg|zzz|egg> bofh: RealSolarSystem is neat http://i.imgur.com/Ii9zheX.png
<Iskierka> ... is this also relevant to having that warning on alcohol cleaning agents?
<Fiora> no
<Fiora> most cleaning agents do not use methanol.
<Fiora> too toxic
<Fiora> and needles, regular ethanol or isopropyl work fine
<Fiora> *needless
<Iskierka> hence alcohol cleaning agents; I learned that they do have a quantity of methanol to make them slightly toxic so they can be sold without alcohol tax, as they're not ingestible
<Fiora> that's called denaturing, yeah
<bofh> egg|zzz|egg: :D
<Fiora> but methanol is less used nowadays BECAUSE of toxicity
<Fiora> some countries have even banned it
<Iskierka> egg: that is a big screen
<Fiora> i.e. there's much better options that aren't as deadly
<egg|zzz|egg> bofh: it has trouble with the fancy coastline though
<Fiora> methanol-denatured alcohol has killed tons and tons of people, it's just outright evil
<Fiora> better nowadays is using something that merely *tastes* awful (maybe combined with something mildly but not terribly toxic)
<Fiora> ex: sometimes they add denatonium
<Iskierka> Was about to ask if this was related to the nintendo switch cartridges and wikipedia immediately says yes
<Iskierka> and I am amused that something they know children will lick has NFPA 110
<Fiora> 110 isn't a big deal, i mean, even many things in food are worse
<Fiora> citric acid is 210!
<Fiora> and the concentration of denatonium is likely extraordinarily low
<Iskierka> presumably referring to when completely pure, however
<Fiora> like, 1 ppm low
<Iskierka> "However, some cats have been known to be tempted by them" uh-oh incoming "my cat ate my switch cartridge"
<bofh> I've used denatonium as a flavouring agent before. You basically need to volumetrically reduce it down to a solution at about 100ppb for it to be overpoweringly strong, but not too much.
<bofh> it's *very* bitter
<Iskierka> How much solution do you get from a drop doing that?
<bofh> like, significantly moreso than aspartame is sweet
<Ellied> wow
APlayer has joined #kspacademia
<Ellied> wiki says EU guidelines for denatured alcohol: "Per hectolitre (100 L) of absolute ethanol: 3 litres of isopropyl alcohol, 3 litres of methyl ethyl ketone and 1 gram denatonium benzoate."
<bofh> Iskierka: I usually do 1mg (approx, weighing that little is always a bit inaccurate) in 250mL water for a ~5ppm solution
<bofh> then add a few drops of that to instant noodles
<Ellied> wait why
<bofh> it's tasty?
<Ellied> I thought it was supposed to taste the Worst
<bofh> so I *like* overpoweringly bitter things
<Iskierka> you're only 2.5x under what the EU regulates would make alcohol unbearable
<Ellied> aight
<Iskierka> you could totally drink paint stripper
<bofh> ...except for that, like, toxicity aspect :P
<Iskierka> I didn't say *and survive*
<bofh> touché
<APlayer> You can at least try anything at least once in your life
<APlayer> :P
<Iskierka> anyway, if xenon prevents apoptosis by the same mechanism, there's a different solubility or barrier-passing requirement, as it's noted to reduce *apoptosis* caused by other GAs (and therefore presumably has this effect at sub-GA concentrations, which is convenient for its cost considerations)
<Iskierka> s/\*apoptosis\* caused/apoptosis \*caused\
<Qboid> Iskierka meant to say: anyway, if xenon prevents apoptosis by the same mechanism, there's a different solubility or barrier-passing requirement, as it's noted to reduce apoptosis \*caused\ by other GAs (and therefore presumably has this effect at sub-GA concentrations, which is convenient for its cost considerations)
<Iskierka> close enough
<Fiora> bofh: interesting thing related to that
<Fiora> so i -cannot- handle bitter things
<Fiora> my taste buds are hypersensitive, such that e.g. a tiny amount of horseradish on something means i cannot taste anything else on that thing
<bofh> Fiora: somewhat same, but in my case this is often desired
<bofh> (I have a phenomenally awful sense of smell that is basically not there unless you're wafting thiols my way, most things taste like various forms of cardboard/very bland to me)
* Iskierka wonders if bofh, in handling denatonium, has ever experienced it at full purity by accident
<APlayer> Detonatium, more like
<APlayer> Where did you even buy it?
<Iskierka> I suspect 110s you can buy fairly trivially from chemical suppliers
<Iskierka> I mean I know people who made a bulk order of nitrous oxide by claiming they were totally ordering for a catering company they couldn't name. Pretty sure most people with limited-harm and not-actually-illegal stuff don't care
<bofh> so 1g was like $45 or something from sigma eldritch and prolly will last me several decades
<bofh> Iskierka: LOL that's hilarious
<Iskierka> It was like 600 cartridges for the dispenser things. Only just fit in a giant camping backpack and was heavy enough to considerably move the wearer's cg
<bofh> damn. I'm mildly curious what they were used for ultimately.
<Iskierka> in fact IIRC it wasn't all of it in the camping backpack
<Iskierka> filling balloons that didn't stay filled for long
<bofh> Ahh. Of course. Gotcha.
<kmath> <thebecwar> @a2_4am I'm convinced that there's a page in the bluetooth standard that just says "LOL, sucker"
<bofh> yep, accurate
<UmbralRaptor> Iskierka: "hippie crack"?
<UmbralRaptor> (that's the term I saw thrown around locally)
<Iskierka> never heard that term but it entered people more rapidly than via cake and into a different organ
<UmbralRaptor> Basically the same, then.
<Ellied> do people use desktop email clients/aggregators anymore
* UmbralRaptor uses Thunderbird
<Ellied> I keep trying to use thunderbird but keep running into stupid problems and falling back on kmail
<Ellied> but kmail depends on akonadi, and akonadi on mysql, and it's a huge teetering stack that I hate
<Ellied> and as of just now, part of it is broken and kmail won't work, so...
<bofh> https://mobile.twitter.com/bofh453/status/904047835317972993 so I just took a look at the Uranian magnetic field and what the hell
<kmath> <bofh453> & then it turns out it does & it's insanely weird (tilted 59° w.r.t rotational axis, not originating from the cente… https://t.co/nqVKmXedYq
<bofh> not only is it extremely tilted against the rotational axis, *it doesn't originate from the geometric center*
<Ellied> I've read about that. Uranus is borked.
<bofh> like can we get a goddamn Uranus orbiter already? Neptune too.
<Ellied> a book I had as a kid had a great little illustration of a bar magnet inside uranus, off-axis and off-center
<Ellied> yes please, I want a landing on Triton while we're at it
<bofh> Landers are complicated, but I suppose without a dense atmosphere a Triton lander would basically be an atmospheric/soil probe.
<bofh> Mind you, I'm pretty sure you could do most Triton atmospheric science via just UV occultation tbh.
<Ellied> yeah true but the surface looks weird and cool from the Voyager images
<Iskierka> I'd say maybe with nasa looking at NTRs again but actually we've done jupiter and saturn and both require less dV to do useful rendezvous
<Iskierka> guess they're just not patient enough for neptune/uranus
<UmbralRaptor> There was a report released a month or 2 ago on ice giant missions. I'd guess that they'll be a priority on the next decadal survey.
<bofh> the trick is timing it right since we need Jupiter:Uranus (and Jupiter:Uranus?:Neptune) to be in the right spot for grav assists.
<Iskierka> if you can get to jupiter with only an earth assist you can get to uranus the same way, on the capture savings
<bofh> well getting directly to Uranus from Earth is incredibly slow without a Jupiter assist.
<kmath> <pnc> the setup on this wikipedia article is fully into wile e. coyote territory https://t.co/BqqcTjv3uH
<UmbralRaptor> The report considered solar electric boost stages. Not sure if they looked at nuclear electric.
<bofh> Better save that space/electricity for RTGs for instrumentation.
<bofh> (other things I'd like to see someday: an interstellar probe: basically Voyager 1 in its current instrument state (more sensitive Magnetometer, more wideband Plasma Wave Spectrometer, various particle counters, UV spectrometer), but with a much higher power transmitter, larger HGA and Americium-241 RTGs. But that's low-priority, especially since there's a very high likelyhood both Voyagers will still be
<bofh> working mostly in 2027)
<egg|cell|egg> !Wpn bofh
* Qboid gives bofh a sinusoidal interferometer with a Toblerone attachment
<bofh> convenient, tho as I said I prefer Tobleramide
<bofh> Ellied: kmail hard-depends on akonadi? ffs
<bofh> akonadi is a goddamn travesty of a tire fire
<Iskierka> Rolos are now smaller as well as toblerones ):
* UmbralRaptor pokes Aetna with a stick.
* UmbralRaptor wonders if they'll be able to effectively enroll in the student plan and transfer the subsidy from the Humana bronze one.
<bofh> "We also prove the existence of strongly regular hyperbolic automorphisms of the hovel, obtaining thus good dynamical properties on the building at infinity ∂Δ."
<bofh> The Building at Infinity is my fav prequel to The Restaurant at the End of the Universe.
<egg|cell|egg> :D
<egg|cell|egg> Why are you reading about hovels Bofh
<UmbralRaptor> δΔ?
<UmbralRaptor> This raises notation questions.
<UmbralRaptor> Well, healthcare.gov accepted my new address.
<bofh> UmbralRaptor: no, (partial)(Delta). as in boundary.
<Iskierka> https://i.imgur.com/YomV94m.mp4 sheep are sometimes supersonic and sometimes subsonic
APlayer has quit [Ping timeout: 383 seconds]
<UmbralRaptor> "However, common aluminium alloys such as the 7075-T series are difficult to make, because certain additives such as selenium and zinc are very rare on the Moon."
<e_14159> What are sheeps' Isp?
<Iskierka> ;wa how fast can a sheep run
<kmath> Iskierka: sheep->maximum speed on land: (data not available)
<UmbralRaptor> Selenium being rare on the moon is far too amusing.
<UmbralRaptor> Iskierka: wait, are the planets named after gods?
<Iskierka> apparently sheep can run at 25 km/h, so about 0.7 seconds
<Iskierka> UmbralRaptor, or the gods *are* planets
<UmbralRaptor> Or that.
<UmbralRaptor> I thought Akatosh was the time dragon?
<e_14159> Hm. That's not efficient enough.
<Iskierka> also mentioned on page: multiple endings of games are canon and consistent
<bofh> UmbralRaptor: LOL
<UmbralRaptor> Iskierka: ¯\_(ツ)_/¯
* UmbralRaptor didn't know that Voyager 1's Titan had a fuel leak.
<UmbralRaptor> (and then the Centaur saves the day!)
icefire has quit [Read error: Connection reset by peer]
icefire has joined #kspacademia
<bofh> UmbralRaptor: so I actually have very little knowledge of the details of that, got a link?
<Iskierka> ... voyager 1 launched after voyager 2?
<UmbralRaptor> bofh: Watching The Farthest
<UmbralRaptor> Iskierka: yep!
<Iskierka> also as I thought, checking the type of titan, fuel leaks are nasty
<bofh> Iskierka: yep. the numbering is b/c Voyager 1 was on a shorter trajectory that flew closer to Jupiter
<bofh> so it would be the first of the two to actually reach it
<Iskierka> why are the titan launch numbers swapped also?
<bofh> Iskierka: no idea
<Ellied> bofh: yeah given how well I've seen it work (very not), I'd sooner depend on a roof made of particle board
<bofh> Ellied: so akonadi is literally the worst dogshit and I have a fake package installed whose sole purpose is to fake being an installed copy of it.
<Ellied> lel
<bofh> (I also have one for pulseaudio, because pulseaudio is literally cancer)
<Ellied> hheh
<Ellied> KDE software has a remarkable range of quality. Some of it's fantastic, some of it's shit.
<Iskierka> sounds like windows and mac
<egg|zzz|egg> !wpn UmbralRaptor
* Qboid gives UmbralRaptor a contravariant teledildonics
<bofh> !wpn egg
* Qboid gives egg a 5-choosable bicyclops which vaguely resembles an analemma
<UmbralRaptor> uh
<egg|zzz|egg> !wpn bofh
* Qboid gives bofh a yttrium pointer
<UmbralRaptor> !wpn egg|zzz|egg
* Qboid gives egg|zzz|egg a thorium edge
<bofh> ...I kind of feel like milling one of those now.
<Iskierka> Qboid liking the nuclear materials
<Ellied> I have yet to see any way of handling bluetooth audio better than KDE does as a frontend for pulseaudio. Tbh, I don't think I've figured out *any* way of getting it to work besides that.
<bofh> I mean I'm pretty sure my room contains like 1% of the Yttrium in this city at this point :P
<Iskierka> so confirmed bofh does not live in Ytterby
<Ellied> it's really nifty how you can just grab an application's audio stream and drop it onto the bluetooth device's output from a taskbar widget
<bofh> (and yes, some of it is in metal form, but most of it is in ceramic form)
<bofh> Iskierka: :P
<Iskierka> ... I don't appear to have such widgets. More recent functionality than 16.04?
<Qboid> 3d 0h 0m 0s left to event #8: Intelsat 37e & BSAT 4a/Ariane 5 ECA [at 2017-09-05 21:51:00]. Say '!kountdown 8' for details
<Ellied> no, it should have that. If you connect a device that has a bluetooth audio sink, you should be able to go to the audio volume control widget, switch to the Applications tab, and grab a playing audio stream, and then it will switch you back to the Devices tab and you can drop it on the device output.
<Ellied> should also work the same if you have multiple sound peripherals installed, bluetooth or not
<Iskierka> BBC says no cloud for that launch. Unusual for ESA!
<Iskierka> although whatever google uses says scattered thunderstorms
<bofh> Iskierka: huh, dang. nice.
<bofh> usually it's cloudy as fuck with ESA launches
<Iskierka> with the amount of disagreement between weather services it could be anything
<Iskierka> but with one of them saying clear skies that'd be nice
<Ellied> I think that's been in since whatever update to Plasma added per-application controls, which was a while ago, like 5.4 or 5.5 I think?
<Ellied> Debian has 5.8, I'm on 5.9
<Ellied> on this machine with Manjaro that is
<Ellied> I have Neon on my HP, so that's whatever the newest is.
<bofh> egg|zzz|egg: I see partly cloudy for Tues
<egg|zzz|egg> yeah
<Iskierka> Ellied, oh. Had no idea. That actually works nicely
<Ellied> :D
<Iskierka> though further in the backend I found "enable play to all outputs" and "switch new output to default" which both work very nicely
<Iskierka> and outputs can be muted individually
<Ellied> that sort of thing has kept me on KDE for like 3 years now. It handles a few things like that really quite well compared to anything else.
<Iskierka> when bluetooth device connects, everything gets pushed over, disconnected, back to the previous output.
<Ellied> yeah it sometimes does weird stuff when it disconnects
<Iskierka> Fine for me, but I'm using that auto-switch-default backend option
<Iskierka> no need to remap as it just goes "okay, we're just playing to the default, what's the new default now that's gone?"
<Ellied> and I've had the most absurd trouble with Clementine doing something that forces the output back to local and not bluetooth, but only after specific songs (consistently)
<bofh> what the, that's weird as shit
<Ellied> other music players don't do it, just clementine
<Iskierka> this sounds like buffer overflow nonsense or such
<Iskierka> reading part of the song data as output device
<Ellied> sometimes, but not always, there's a dialog that comes up that says "invalid state" or something
<Ellied> oh, maybe I should just try increasing the buffer size or something
<Iskierka> I'd hope that's *not* it
<Iskierka> certainly as you'd think it would break in so many more ways than that if it was buffer overflowing
<Ellied> yeah probably
<Ellied> I'd love it if there were a plugin to put KDE system tray applets into the xfce panel. Then on my slower machines that are only just barely able to run the full Plasma shell wouldn't have to, but I could still use the nifty controls
<bofh> ehh, doubt it, audio out goes into a circular buffer and I hope alsalib isn't too broken that it starts shitting into places it shouldn't.
<bofh> the Plasma shell eats resources, I kinda hate it
<Ellied> whitequark told me she uses the Plasma desktop but with i3 as the window manager. Maybe I should try something like that.
<Ellied> might help a little.
<Ellied> dinnertime
<kmath> <whitequark> incognito tab google search https://t.co/KSt0TemFXx
* egg|zzz|egg pokes bofh with a squareRoot and a rootn
<bofh> Ellied: oh prolly.
<bofh> Ellied: I sometimes use the Plasma shell and sometimes xfce4-panel depending on system. No desktop b/c fuck those. And yeah I use i3 as well (or openbox on oen system).
<bofh> egg|zzz|egg: I don't have much to add since last convo, tho I could long-distance poke Atlas
<bofh> also that somehow reminded me that I haven't logged into synapsid website to follow folks so let me do that
<bofh> (same username as on diapsid website)
<Iskierka> for all people complain of plasma eating resources I'd be very impressed at any other OS beating the ~24 hour reasonable-use-idle battery life of my laptop
<Iskierka> windows at similar or more minimal settings is down at 21 hours
* UmbralRaptor really needs to add his synapsid accounts to Twidre.
<egg|zzz|egg> bofh: so there's apparently an incompatibility between principia and another mod, see #1549; we concluded that the other mod was doing something odd, so the user opened ihsoft/KIS#221, and then we looked at ihsoft's github page Ꙩ_ꙩ
<Qboid> [#1549] title: Seems to crash with KIS | Crashing when trying to add a part to spacecraft.... | https://github.com/mockingbirdnest/principia/issues/1549
<Qboid> [#221] title: Crash with principia when dropping item | I've reported this to principia and it seem you are doing the wrong thing (not them) (tm)... | https://github.com/ihsoft/KIS/issues/221
<Iskierka> Twidre?
<UmbralRaptor> …wait, people complained that the Pioneer plaques were porn?!
<UmbralRaptor> Iskierka: Android Twitter client. FLOSS.
<Iskierka> I've seen someone complaining that the plaques were what made the missions so expensive
<Iskierka> I worked out they're about $5000 of gold
<UmbralRaptor> lol
<egg|zzz|egg> !wpn UmbralRaptor
* Qboid gives UmbralRaptor a épée
<UmbralRaptor> Yay, stab!
<Iskierka> a/an fail
<UmbralRaptor> (Uh, I think Qboid is confused by the accent?)
<Iskierka> probably
<Iskierka> still gonna point it out :p
<UmbralRaptor> !u eé
<Qboid> U+0065 LATIN SMALL LETTER E (e)
<Qboid> U+00E9 LATIN SMALL LETTER E WITH ACUTE (é)
<bofh> egg|zzz|egg: OH GOD
<Iskierka> Not a simple binary offset. Solution: check unicode name, if matches "latin <small/capital> letter <vowel>", do an
<bofh> egg|zzz|egg: so they're doing vessels with no parts?
<bofh> egg|zzz|egg: also anything special with ihsoft's github page?
<Iskierka> There is actually a lag as you attach parts so I'm suddenly wondering if what they do confuses other bits of the game and takes a while to figure out how to handle an empty vessel
<Iskierka> although if they pushed that to KIS it's probably the wrong one as I think that's all handled by KAS
<egg|zzz|egg> bofh: so there's nothing particularly weird with that, every weird mod will break invariants (principia after all breaks everything that thinks Kepler orbits are the authoritative truth, breaks manoeuvres, etc.)
<egg|zzz|egg> bofh: no, I mean look whom they work for
<kmath> <brianklaas> Trump named a politician who has denied climate change science and who has no scientific credentials...to run NASA. https://t.co/mQ4AuQZlnd
<bofh> oh for fuck's sake
<bofh> egg|zzz|egg: wait, I thought Principia makes (perturbed?)Kepler orbits the authoritative truth? :P
<bofh> egg|zzz|egg: oh, another googler
<egg|zzz|egg> bofh: well the authoritative truth is our internal state, should you change the orbits, they will get overwritten by principia at the next frame (so hyperedit or the stock orbit editor cheat don't work)
<egg|zzz|egg> bofh: I mean that explains why they use the Google Java styleguide for their C#
<egg|zzz|egg> (we mostly use the Google C++ styleguide for ours :-p)
<bofh> egg|zzz|egg: well okay, but that's... kind of obvious? Principia has to control the entire internal state to, like, work. :P
<egg|zzz|egg> bofh: well you could imagine storing the data in the stock orbits, and evolving those (if you only use one-step methods and don't have multiprecision that works conceptually)
<Ellied> I like desktop environments, but by now it's more on an academic level, just interested in seeing what the modern-day developer thinks is a good user experience.
<egg|zzz|egg> bofh: also did I tell you about our generated interface code
<bofh> egg|zzz|egg: nope
<Ellied> and they also let me do certain tasks very quickly that my peers (most of whom use macs) expect me to be able to do quickly, like connect to wifi or print or do shit with bluetooth.
<bofh> oh, I do use an env, just not a desktop icon/shell program :P
<egg|zzz|egg> bofh: so most of our mod is C++, but KSP mods are actually .NET dlls that get loaded by the game; therefore there's a p/invoke interface
<Ellied> and ironically enough, knowing me, sometimes I *don't* want my screen to draw unnecessary attention by looking weird.
<egg|zzz|egg> bofh: it used to be that we wrote a header file for that, and then the corresponding extern declarations in C#
<Ellied> I already use the terminal way too much.
<egg|zzz|egg> bofh: and of course we'd screw that up and refer to nonexistent functions and fail at runtime :-p
<egg|zzz|egg> bofh: then we decided to add journaling (for profile guided optimization and also just debugging from in-game sessions)
<egg|zzz|egg> journaling at the interface of course, that's the natural place
<egg|zzz|egg> bofh: so then we needed a proto message definition for every interface function, so you'd have to keep in synch the header, the C# imports, and the protos
<egg|zzz|egg> bofh: so we kept the protos, and we generate C# and C++ structs and function declarations from that (as well as code that fills the protos when journaling)
<Ellied> there's a silly little program called Latte Dock which is mainly designed to let you configure Plasma to look like OSX, but also has the interesting characteristic of being a shell-agnostic panel to which Plasma applets can be attached
<bofh> egg|zzz|egg: oh god this is hilarious and sounds like terrible preprocessor abuse
<egg|zzz|egg> bofh: with some fancy annotations to say "this integer is actually a pointer so declare it as such, and also the function will delete it so keep track of that when replaying the journal" https://github.com/mockingbirdnest/Principia/blob/master/serialization/journal.proto#L251-L252
<egg|zzz|egg> bofh: nono, not preprocessor
<bofh> egg|zzz|egg: I'm getting libswscale flashbacks now, thanks for that
<Ellied> I've attempted to use it in Debian (compiling from source) with kind of shitty results, but so far the version Manjaro ships is great.
<egg|zzz|egg> bofh: not preprocessor, an actual bit of C++ that uses proto reflection to generate C++ and C# https://github.com/mockingbirdnest/Principia/blob/master/tools/journal_proto_processor.cpp
<bofh> oh god I don't know if I should be more or less horrified.
<bofh> > protobuf
<bofh> of course
<Iskierka> egg abuses computers
<egg|zzz|egg> bofh: so, compile the proto, generating the pb.cc -> compile the profile generator against that, generating the interface.generated.{h,cs} and the profiles -> compile the interface itself (which uses the journaling code, which of course uses the pb.cc that were compiled in the first step because we journal into protos)
<bofh> congrats for somehow being several orders of magnitude more complicated than libswscale, and probably comparatively bad to boost::spirit
<egg|zzz|egg> bofh: like, it's actually much much nicer than the alternative of writing all that crap by hand (and being able to replay things that you encounter in-game is fairly useful)
<bofh> Okay, true, I guess really what I'm marvelling at is that this is all *required*.
<egg|zzz|egg> bofh: it does make the build script a bit convoluted (this is the step that generates the header, C# decls, and profiles: https://github.com/mockingbirdnest/Principia/blob/master/Makefile#L147-L148)
<Iskierka> well they could've just written it all in C#
<Iskierka> not sure it'd run nice though
<egg|zzz|egg> it would probably have awful performance tbh, but the strong typing of quantities and reference frames is why we went with C++
<egg|zzz|egg> bofh: I mean it's not required; absent the replaying thing we would probably still be struggling to keep the things in synch by hand, whereas here we have code generators that we touch twice a year
<egg|zzz|egg> bofh: also obligatory mention of this line in the Makefile https://github.com/mockingbirdnest/Principia/blob/master/Makefile#L295
<bofh> UM.
<bofh> what does it *do*?
<egg|zzz|egg> bofh: it says make: *** No rule to make target 'the_universe', needed by 'an_apple_pie'. Stop.
<bofh> Ahh.
<egg|zzz|egg> bofh: I mean there's that one time when I did have a rule to make the_universe and I ended up having to wait nearly 14 Ga for make to finish
<bofh> ROFL
<Iskierka> surely the universe is made in the first planck time of its existence
<Iskierka> anything beyond that is creating things that must exist within a universe
<egg|zzz|egg> bofh: Iskierka: yes, making the universe is quick, but since it can make the universe it then proceeds to make an apple pie, and that takes eons