2023-06-24 04:13:45 <veltas      > Found the C names-as-symbols example from 4chan https://i.imgur.com/nsvum0E.png
2023-06-24 04:21:05 <veltas      > Will have to repost when unjust is here
2023-06-24 04:37:05 <gordonjcp   > KipIngram: ah, okay, cool
2023-06-24 04:37:30 <gordonjcp   > KipIngram: and then you can use the Russian Farmer method with the addition steps hardcoded
2023-06-24 09:24:11 <KipIngram   > Yeah, it's all an exploitation of having just a single divisor that you know ahead of time.
2023-06-24 09:24:38 <KipIngram   > I really should bone up better on my finite fields - that stuff is key to a whole lot of encryption algorithms and so forth.  I know it a little, but it's... fuzzy.
2023-06-24 09:25:24 <KipIngram   > If the number of items in your finite field is prime, then some really interesting properties ensue, and those particular encryption algorithms are based on those properties.
2023-06-24 09:25:54 <thrig       > they should be a quicker study than infinite fields
2023-06-24 09:27:05 <KipIngram   > Part of the definition of a field is that a product operation exists -  you can take the product of any two items, and the result has to also be an item in the field.  If the size of the field is prime, then if you multiply by the same item over and over, the results will walk through every item in the field, except the zero item.
2023-06-24 09:27:22 <KipIngram   > This is what's going on with pseudorandom bit sequence generators.
2023-06-24 09:33:53 <KipIngram   > You can check it for simple cases.  Take the five-element finite field, which we can represent as (0, 1, 2, 3, 4).  Let's start with 2.  So we mark 2, and multiply by 2.  That's 4, so we mark 4.  2*4 = 8, mod 5 is 3, so we mark 3.  2*3 = 6, mod 5 is 1, so we mark 1.  2*1 = 2, so we're back to the start.  And we marked 1, 2, 3, 4 - everything except 0.
2023-06-24 09:34:20 <KipIngram   > They can prove that works out for every case where the size is prime.
2023-06-24 09:35:01 <KipIngram   > We could have started with 3 or 4 as well, and it will work in those cases too.
2023-06-24 09:35:42 <KipIngram   > Can't start with 0 because that sticks you at 0, and starting with 1 doesn't get you anywhere either.  But all other values will always work.
2023-06-24 09:36:52 <KipIngram   > I may be slightly off in how I stated it, but there's something along those lines that's an important result.
2023-06-24 09:37:56 <KipIngram   > I must be, in fact, because it doesn't seem to work if we start with 4.  4*4 = 16, mod 5 is 1, and 4*1 is back to 4.  Maybe the starting number has to be odd.  I'd have to look it up.
2023-06-24 09:38:27 <KipIngram   > Or prime - maybe the multiplier has to be prime as well.  That feels right.
2023-06-24 09:41:35 <KipIngram   > And now that I think about it I think these results may be associated with groups rather than fields.  Fields have added structure - you hwave two operations (+ and *) defined on fields, whereas groups have just one, *.
2023-06-24 10:07:01 <KipIngram   > gordonjcp: A lot of error detection and correction methods use the same math.  The stuff we're using these days on our drives involves fields of polynomials.
2023-06-24 10:07:09 <KipIngram   > And is way beyond me.
2023-06-24 10:07:47 <KipIngram   > There's a guy at the office who crafts FPGA circuitry to do all that stuff - I've tried to learn from him, but... not so far.
2023-06-24 10:12:12 <KipIngram   > Some poor kid has engaged me in a private conversation on Quora - he's trying so so hard to "hold onto" his notions about "classical particles."  The problem is that those ideas just break down at the microscopic level, and you have to give them up.  He doesn't want to.
2023-06-24 10:12:53 <KipIngram   > He really, REALLY wants the microworld to be little billiard balls.
2023-06-24 10:13:07 <thrig       > pretty sure there's a professor who has been arguing against Gödel's incompleteness for like forever
2023-06-24 10:15:36 <KipIngram   > Hard to give up things sometimes.
2023-06-24 10:16:04 <KipIngram   > Another siilar kind of thing that comes up on Quora a lot is "how can photons have momentum if they don't have mass?"
2023-06-24 10:16:26 <KipIngram   > The problem there is that people take the fact that they were taught momentum is mass times velocity and try to make it the exclusive definition of momentum.
2023-06-24 10:16:40 <KipIngram   > It's not - momentum is just a broader thing that that - we weren't taught the whole story in high school.
2023-06-24 10:17:06 <KipIngram   > Mass times velocity is ONE WAY mmentum can show up.
2023-06-24 10:18:17 <KipIngram   > Mass times velocity is really wrong anyway - it's only valid at small velocities compared to the speed of light.  It's really m*v/sqrt(1-v^2/c^2).
2023-06-24 10:18:51 <KipIngram   > And that very conveniently becomes undefined when v = c, so it says nothing at all about the momentum of a photon - you wind up with 0/0, leaving the door wide open.
2023-06-24 10:21:22 <KipIngram   > It's basically why photons travel at the speed of light - if they didn't they'd have no momentum and no energy and we wouldn't be able to detect them.
2023-06-24 10:21:56 <KipIngram   > The only way for a massless particle to have any effect on the world at all is for it to travel at c.
2023-06-24 10:22:02 <gordonjcp   > KipIngram: isn't Quora just StackExchange for right-wing nutcases with a bizarre view of the world? ;-)
2023-06-24 10:22:37 <KipIngram   > I don't know - I really only pay attention to questions related to math and physics.  The stuff that shows up in my inbox is related to the kind of questions I've answered in the past.
2023-06-24 10:22:37 <gordonjcp   > KipIngram: anyway mass times velocity *is* a good approximation for momentum
2023-06-24 10:22:51 <KipIngram   > So I don't really partake of any "political traffic" on Quora.
2023-06-24 10:23:11 <KipIngram   > It's a good approximation in the situations we typically encounter day to day, yes.
2023-06-24 10:23:20 <gordonjcp   > KipIngram: just as out = ((in-out)*cutoff)+out is a good approximation for a 1-pole integrator
2023-06-24 10:23:23 <gordonjcp   > however
2023-06-24 10:23:23 <KipIngram   > Which is why our intuition is locked in on it so strongly.
2023-06-24 10:23:30 <KipIngram   > It's "always worked before" for us.
2023-06-24 10:23:33 <gordonjcp   > both blow up pretty badly as you get close to Nyquist
2023-06-24 10:24:01 <gordonjcp   > KipIngram: Newtonian vs. Relativistic physics are the same thing, mostly
2023-06-24 10:24:11 <KipIngram   > Yeah.  If we lived in a world where we could actually perceive details in "near the velocity of light" cases, our intuition would be very different.
2023-06-24 10:24:25 <gordonjcp   > KipIngram: it's just that you have to use more complex Relativistic methods when your approximations break down at half the sample rate
2023-06-24 10:24:32 <KipIngram   > But all that stuff happens too fast for us to tell anything about it, without sophisticated instruments, so we get no "training" as we grow up.
2023-06-24 10:24:46 <gordonjcp   > the real, continuous, analogue world has a sample rate
2023-06-24 10:25:04 <gordonjcp   > it's just it's related to Planck's Constant
2023-06-24 10:25:05 <KipIngram   > Yeah, totally agree with what you're saying.
2023-06-24 10:25:25 <KipIngram   > Humans using their naked senses are just very crude "instruments."
2023-06-24 10:25:40 <KipIngram   > We get only limited insight into the world as we just look around at it.
2023-06-24 10:26:12 <gordonjcp   > it's telling that every single one of our naive approximations - which work well for "in the middle" things - blow up at scales that are roughly 2.5 times Planck
2023-06-24 10:26:16 <KipIngram   > And modeling things as little billiard balls and so on - that works great.  It's no surprise those ideas get "built in" to our thinking.
2023-06-24 10:26:25 <KipIngram   > Yep.
2023-06-24 10:27:02 <KipIngram   > It's like expecting the whole world to operate just like the little village you grew up in.  Just a very unlikely scenario.
2023-06-24 10:28:17 <KipIngram   > Anyway, I wasn't aware of any particular "political slant" to Quora - you may be totally right.  It's just not the corner of Quora I dip into.
2023-06-24 10:29:12 <KipIngram   > My wife just yesterday read off to me a list of Quora questions that had landed in her inbox, and I had to admit hers were crazier than mine.  Still not really "political," but bizarre nonetheless.
2023-06-24 10:34:23 <gordonjcp   > KipIngram: if you look at the political side, a lot of it is questions from people baffled because they don't live in an Ayn Rand novel
2023-06-24 10:38:51 <KipIngram   > I'm a big freedom fan, but I've always found there to be something a bit cold about Rand's philosophy.  Mechanistic.
2023-06-24 10:39:25 <KipIngram   > In some way it feels like it takes a good concept and pushes it to far to an extreme.
2023-06-24 10:39:32 <KipIngram   > too
2023-06-24 10:40:22 <KipIngram   > It's like she's so fixated on the individual liberty thing that she loses sight of all the things that make liberty worth having in the first place.
2023-06-24 10:41:36 <KipIngram   > I think that's very connected to my objections to large corporations - I think they also lose sight of the very same things.
2023-06-24 10:42:41 <thrig       > but but next quarter profits
2023-06-24 10:45:00 <gordonjcp   > KipIngram: it probably appeals to teenagers who think the world revolves around them and money comes from some magic source
2023-06-24 10:45:14 <gordonjcp   > KipIngram: it falls apart rapidly the moment you have one other person to take care of
2023-06-24 10:45:43 <gordonjcp   > it's a very North American way of thinking tbh
2023-06-24 10:50:31 <KipIngram   > Yeah.  I'd never thought about the connection between that and "corporate" before, but I think it fits - Rand generally has her protagonist behave almost as though they "run themselves like a corporation."
2023-06-24 10:51:11 <KipIngram   > Selfish, methodical, precise - no room for silly little things like compassion, friendship, family, community, etc.
2023-06-24 10:54:17 <KipIngram   > Whereas my feeling is that if you live your life without those "silly things" then you havne't had much of a life, never mind your bank account balance.
2023-06-24 11:00:46 <KipIngram   > I actually think a lot of things appea to teenagers - you can find plenty of them over on the other end of the spectrum too, where that lack of understanding about things having to be paid for by SOMEONE leads them toward visions of some magic world in which everyone is just "automatically provided for."
2023-06-24 11:01:07 <KipIngram   > They're just missing some reality checks at that age, and it can lead them in different directions.
2023-06-24 12:58:09 <KipIngram   > I just re-watched this:
2023-06-24 12:58:11 <KipIngram   > https://www.youtube.com/watch?v=GLG9g7BcjKs
2023-06-24 12:58:24 <KipIngram   > I come back around to it every six or eight months.
2023-06-24 12:58:57 <KipIngram   > I'd probably disagree with that guy on a ton of detailed issues, but once he gets a couple of minutes into this I think he delivers a message that everyone who wants to think of themselves as liberal should listen to.
2023-06-24 12:59:04 <KipIngram   > He NAILS a major problem.
2023-06-24 13:00:05 <KipIngram   > A problem that is the primary reason I likely don't label myself as more liberal - I'm actually fairly open to a lot of those ideas, but I object so much to the whole "holier than though" attitude the camp projects that I just can't bear to think of myself as part of it.
2023-06-24 13:00:39 <KipIngram   > It's just a simple truth: if you attack people, they'll go on the defensive.  Nothing about that should surprise anyone.
2023-06-24 13:02:24 <KipIngram   > If you actually expect to change someone's mind on an issue, don't go in with guns blazing.  Have a CONVERSATION.
2023-06-24 13:02:31 <KipIngram   > A rational one.
2023-06-24 13:02:45 <KipIngram   > And that means listening as well as talking.
2023-06-24 13:35:22 <thrig       > less talk, more click-generating outrage
2023-06-24 13:35:43 <unjust      > go woke! go broke!
2023-06-24 13:47:37 <KipIngram   > Hey, is it possible to create a hyperlink, on a browser page or in a pdf document, that will open a console window on some server, assuming I have a working username and password?
2023-06-24 13:48:04 <KipIngram   > I'd like to click the link in my pdf document and have a console just open up (I'd actually like to have it run vim on some particular file).
2023-06-24 13:48:29 <thrig       > maybe if something knows what to do with a ssh link?
2023-06-24 13:49:08 <KipIngram   > I've created a report generator for my test runs that makes a 60-page or so pdf file, one page per test step with graphs and so on.
2023-06-24 13:49:43 <KipIngram   > I also collect log files during those tests, and the developers sometimes want to look at them.  I was hoping I could set it up so that clicking on a link on some test case page would send someone right to the log file for that page.
2023-06-24 13:50:01 <KipIngram   > I just learned that you can attach text files to a pdf, which is a cool idea, but these log files for an entire run are quite large.
2023-06-24 13:50:11 <KipIngram   > I don't think I want to attach that much stuff TO the pdf.
2023-06-24 13:50:12 <thrig       > put the log files on a website?
2023-06-24 13:50:26 <KipIngram   > Can't do that cuz "company policy."
2023-06-24 13:50:55 <KipIngram   > Even if I figure out how to do this, I need to give it careful thought - no telling where these pdf files might get circulated.
2023-06-24 13:51:25 <KipIngram   > I wouldn't want just anyone that happened to get a pdf file to be able to get onto that server, unless I bolted the account down pretty heavily.
2023-06-24 13:51:25 <Zarutian_iPad> internal website?
2023-06-24 13:51:45 <KipIngram   > Policy applies to that too - all this stuff is behind our VPN boundary anyway.
2023-06-24 13:51:56 <KipIngram   > They're "sensitive" about web servers.
2023-06-24 13:52:39 <thrig       > snowden.corp.int
2023-06-24 13:52:58 <KipIngram   > I usually copy these files to Box (our internal Dropbox), so I'm also looking into whether there's a way to automate that.  The files are available via browser, but I'm not sure how to figure out what there url is because it gets some magic number in it that Box makes up on the fly.
2023-06-24 13:53:11 <KipIngram   > I know how I've named the files, but the url winds up not being based just on that.
2023-06-24 13:54:13 <KipIngram   > In other words, for any given file I could get on my browser and discover an address I could then put into a link.
2023-06-24 13:54:20 <KipIngram   > But I don't know how to do that from Python.
2023-06-24 13:55:41 <KipIngram   > Linking to Box is the better idea, because the files are more guaranteed to last long term there.
2023-06-24 13:55:58 <KipIngram   > But it's unlikely a developer will be wanting to look at a really old log file.
2023-06-24 13:56:32 <KipIngram   > I did this already done work with the matplotlib-connector Python package and the pdfpages package.
2023-06-24 13:56:39 <KipIngram   > It was quite straightforward.
2023-06-24 13:57:00 <KipIngram   > Started on it early yesterday morning and got it basically complete by the end of the workday.
2023-06-24 13:57:25 <KipIngram   > I plan to add some further things, but it's already quite nice.
2023-06-24 13:58:29 <KipIngram   > Each page is organized with two columns of graphs - on the left I have the current drive generation shown against a baseline of the same case from the previous generation, and on the right the same thing except it's the current generation / current version vs. current generation / prior version.
2023-06-24 13:59:17 <KipIngram   > I've got graphs that show latency vs. iops as queue depth is varied, latency histograms, and the simple vs. time graphs of various quantities for each test case.
2023-06-24 14:01:25 <KipIngram   > I used to just make a folder full of individual graphs.  When i looked at them I'd open them all at once in Mac Preview, and that resulted in a nice experience.  But the other guys look at them after I upload them to Box, one image at a time, and that's not an experience the enjoyed.  Getting them into one document like this will give them an experience more like the one I've had.
2023-06-24 14:01:54 <KipIngram   > I also don't think Box kept the individual files in the right order, either, so I guess I see that was somewhat chaotic.
2023-06-24 14:02:24 <KipIngram   > Anyway, showed the current results to "the Fellow" yesterday and he approved, so that's nice.
2023-06-24 14:02:48 <KipIngram   > He's in a good mood right now anyway because he just became a grandfather a couple days ago.
2023-06-24 14:04:52 <KipIngram   > Anyway, I need to try to make it work with Box.  That solves the security problem, because then only people who have access to the Box content to start with would be able to follow the links.
2023-06-24 14:17:18 <thrig       > so much for thinking outside the Box
2023-06-24 14:22:18 <KipIngram   > :-)  Nice.
2023-06-24 14:41:57 <gordonjcp   > jeeeeeeeeeesus
2023-06-24 14:41:58 <gordonjcp   > so
2023-06-24 14:42:02 <gordonjcp   > that fucking submarine
2023-06-24 14:42:27 <gordonjcp   > the bolts to hold the hatch on had a torque spec, but they were done up with an ordinary ratchet
2023-06-24 14:42:38 <gordonjcp   > they only did up 17 of the 18 because the guy couldn't reach the top one
2023-06-24 14:42:41 <gordonjcp   > fuck *me*
2023-06-24 14:43:09 <thrig       > pesky regulations
2023-06-24 14:43:16 <gordonjcp   > not only that
2023-06-24 14:43:36 <gordonjcp   > but they used carbon fibre and resin that they bought cheap from Boeing because it was out of date
2023-06-24 14:43:37 <Zarutian_iPad> "normalization of deviancy"
2023-06-24 14:43:52 <thrig       > something something o-rings
2023-06-24 14:48:22 <thrig       > on the bright side, it prevented Boeing from using said in a plane
2023-06-24 15:46:57 <KipIngram   > Yeah, it's amazing how much stuff is coming to light.  My wife works in the subsea industry, so she's been following it with "professional interest."
2023-06-24 15:47:21 <KipIngram   > The thing is, though, the CEO of the company running it was on board, so at least he put himself right there on the front lines.
2023-06-24 15:47:42 <thrig       > skin in the game can be uncommon these days
2023-06-24 15:47:56 <KipIngram   > I'm on a mailing list for SETI, and yesterday got an email revealing that one of the SETI board of trustee members and his son were on board.
2023-06-24 15:48:25 <KipIngram   > Yes - I still think the guy made some severe mistakes, but I can't feel quite as bad toward him as I would if he'd been chilling in his living room when it happened.
2023-06-24 15:49:03 <KipIngram   > Apparently at some point he told someone that Boeing had helped design it - Boeing is vociferously denying that at this point.
2023-06-24 15:50:03 <KipIngram   > And he had also said he bought the carbon fiber from Boeing "at a great price" because it had "past its expiration date."  Boeing denies that too.
2023-06-24 15:50:21 <KipIngram   > Boeing wants nothing to do with this - I expect no one wants to be connected up to it in any way.
2023-06-24 15:50:41 <thrig       > boeing has had a number of own goals they might be MAX sensitive about
2023-06-24 15:55:14 <KipIngram   > I'm just cut from a different kind of cloth - I'm perfectly happy to look at pictures of the Titanic sitting right here on my sofa.
2023-06-24 15:58:02 <KipIngram   > Some people, though, seem to need something "edgy" going on in their life all the time.
2023-06-24 16:34:56 <veltas      > unjust: https://i.imgur.com/nsvum0E.png
2023-06-24 16:35:54 <unjust      > veltas: thanks for digging that up
2023-06-24 16:37:18 <veltas      > No problem
2023-06-24 16:37:18 <unjust      > kind of reminds me of LOLCODE somehow
2023-06-24 16:47:35 <KipIngram   > It's... cute.  :-)
2023-06-24 16:49:12 <veltas      > KipIngram: Yes that's literally what they said on 4chan
2023-06-24 16:49:16 <veltas      > When it was posted
2023-06-24 16:49:25 <veltas      > I assume it's from Twitter or something though
2023-06-24 17:29:33 <gordonjcp   > https://media.jehaisleprintemps.net/talks/baguetteonsnails/
2023-06-24 17:29:51 <gordonjcp   > ^ bit of an in-joke, posted at an unconference by an Ubuntu dev from France with a very strong accelt
2023-06-24 17:29:54 <gordonjcp   > *accent