2026-02-02 09:14:04 <KipIngram   > I feel like my notebook fan is audible more frequently since the Fedora 43 upgrade.  :-(
2026-02-02 09:19:11 <tpbsd       > maybe the ACPI is broken or the speed control or temperature sensor ?
2026-02-02 09:20:57 <KipIngram   > Yeah, no idea.  Back in the old days when I would buy a Windows notebook and immediately nuke Windows and install Linux, I was accustomed to having to fight battles to get my battery life up and deal with such issues.  A few years ago, though, I started buying notebooks shipped with Linux, and that immediately got a lot better.  But I'm pretty familiar with this kind of thing changing from release
2026-02-02 09:20:59 <KipIngram   > to release, as they tinker with things and don't always have the same priorities I have.
2026-02-02 09:21:32 <KipIngram   > I am still in "balanced" power profile - it's not as simple as having gotten switched to "performance."
2026-02-02 09:21:41 <tpbsd       > is it a thinkpad ?
2026-02-02 09:22:17 <KipIngram   > No, it's a "Slimbook."  That's the moniker of a line of notebooks prepared by the Fedora organization itself, and supposedly optimized to run Fedora.
2026-02-02 09:22:28 <tpbsd       > Linux has always worked really well on Thinkpads I find
2026-02-02 09:22:35 <tpbsd       > oh
2026-02-02 09:27:22 <KipIngram   > My first Linux ready notebook was from Lenovo, just a product they offered.  I was quite happy with it.  I've been happy with the Slimbook too - my reasoning on that choice was that it still could be the case that Linux wasn't as strnog a support focus for Lenovo, even if they chose to offer it.  But you'd expect it to be a very strong support focus for Fedora - their whole point.
2026-02-02 09:27:40 <KipIngram   > It felt like the way to get the most dedicated support for the whole endeavor.
2026-02-02 09:28:05 <KipIngram   > Their whole goal would be to have Fedora make the best possible impression on the customers.
2026-02-02 09:33:35 <tpbsd       > sound logic
2026-02-02 09:36:09 <tpbsd       > I have a couple of Thinkpad X60s by Lenovo, probably circa 2002, Im astonished how perfectly MX-Linux runs them, everything works, even sound and brightness buttons, suspend on close screen, resume on open laptop etc
2026-02-02 09:37:11 <tpbsd       > theyre nearly worn out now, I used them so much, and they were second hand when I got them on ebay for $100 each around 2015
2026-02-02 09:37:37 <tpbsd       > I also have the docking bases, theyre outstanding also
2026-02-02 09:38:27 <tpbsd       > I fear that there will never be units of that quality again (they were about $4500 when new)
2026-02-02 09:38:59 <tpbsd       > theyre in lightweight magnesium skins, just awesome
2026-02-02 09:40:40 <KipIngram   > Ok, I discussed this with ChatGPT, and it drew my attention to Intel turbo boost.  It was on.  I have now turned it off; we'll see if that makes a difference.
2026-02-02 09:41:08 <KipIngram   > Most of the time I'm not needing tippy top performance from my computer - if I run something that does it's easy to turn it back on.
2026-02-02 09:42:24 <tpbsd       > ahh, that sounds like a good reason for the extra fan noise
2026-02-02 09:42:47 <KipIngram   > I found it useful to turn it off back at work, in my ssd testing.  My box would hold twelve SSDs, and I often wanted to test several at once.  Obviously, I didn't want the fact I was testing "other drives" to affect a particular drive's test results.  So I gave each drive a pinned core for its fio traffic generator process.  I found that with turbo boost on, those tests DID affect one anotoher.  I
2026-02-02 09:42:50 <KipIngram   > could see it in the results.  Turning it off made a very noticeable difference in "isolating" the tests from one another.
2026-02-02 09:43:33 <KipIngram   > Hmmm.  I probably should send a note on that to the kid I trained to replace me.  It's "built in" to my tooling, but he probably still ought to be specifically aware of it and the reasoning behind it.
2026-02-02 09:43:42 <KipIngram   > I didn't think to mention that while bringing him up to speed.
2026-02-02 09:44:48 <tpbsd       > KipIngram, I developed and finished two apps using languages I don't know at all, but I needed AI to do it. I finished, they work fine, no obvious problems but I didnt enjoy the process at all
2026-02-02 09:44:55 <KipIngram   > Well, it's hard to say for sure just a few minutes in, but it actually already does seem quieter.
2026-02-02 09:45:28 <KipIngram   > I tinkered a bit with using ChatGPT to actually help me produce code, but didn't wind up pleased with that process AT ALL.
2026-02-02 09:45:47 <KipIngram   > I use it more as a sounding board - it is good at "carrying a conversation" and I find that just chatting with it helps me sharpen my own thoughts.
2026-02-02 09:45:55 <tpbsd       > so I decided to only use AI for Google type issues, like your laptop fan advice above and also configuring apps I dont know etc
2026-02-02 09:45:56 <KipIngram   > Then I write my code the way I always would.
2026-02-02 09:46:17 <KipIngram   > Right - it's good for quick info on things I'm unfamiliar with.  Small one-shot things like that.
2026-02-02 09:46:46 <KipIngram   > But this phrase "vibe coding" just doesn't excite me at all.  I found it to be fairly disastrous.
2026-02-02 09:47:05 <KipIngram   > Mostly because the AI would base its reasoning on deprecated information it sees out there.
2026-02-02 09:47:05 <tpbsd       > yeah, if I do any more with those apps, I'll make sure I learn them and do the work myself, asking the AI for help only where Im stuck
2026-02-02 09:47:24 <KipIngram   > Right.  It is useful in those ways.
2026-02-02 09:48:02 <tpbsd       > Ive also noticed that letting the AI do the work has taken the edge of my software design (it wasnt great to begin with as Im not a programmer, Im a electronics guy)
2026-02-02 09:48:35 <KipIngram   > I've been chatting with it about this new system I'm about to write, and having that conversation is part of what helped me arrive at the most recent layer of ideas I'm fairly excited about.  Getting things lined up right from the start to support clean and tidy meta-compilation.
2026-02-02 09:49:13 <tpbsd       > so I learnt that using AI has let me allow my software creativity deteriorate
2026-02-02 09:49:17 <KipIngram   > Forth has always been able to meta-compile, but typically it involves quite a lot of hoop jumping.  I want to plan right from the start for it and hopefully have it be more graceful.
2026-02-02 09:49:57 <KipIngram   > Yeah, I'm 100% sure there are tons of people out there trying to "use AI to be more than they are."  I think that's a really dangerous way to use it.
2026-02-02 09:50:10 <tpbsd       > yeah, I get good feedback discussing concepts etc with a AI at the design stage
2026-02-02 09:50:48 <KipIngram   > It's not that different from having a good friend or colleague to discuss things with, except the AI never runs out of patience.  :-)
2026-02-02 09:51:08 <tpbsd       > I do enjoy that part. Its the part where I let it do my design that I dint enjoy, that has left a bad taste in my mouth, it's just not me
2026-02-02 09:51:17 <tpbsd       > exactly!
2026-02-02 09:51:20 <KipIngram   > Right.  That just didn't work.
2026-02-02 09:51:25 <tpbsd       > a Ai has to listen
2026-02-02 09:51:39 <KipIngram   > I wound up spending as much time getting the AI output to work as I'd have spent doing the job myself.
2026-02-02 09:51:54 <KipIngram   > I never had it give me something that "just immediately worked."
2026-02-02 09:52:03 <tpbsd       > it's like the definition of home "home is the one place where they HAVE to let you in" !
2026-02-02 09:52:06 <KipIngram   > Not for something at all complicated.
2026-02-02 09:52:21 <tpbsd       > yeah, agreed
2026-02-02 09:52:54 <KipIngram   > And when I was trying to fix its code, it wasn't much different from inheriting code someone else had written, and I never enjoyed that anyway.
2026-02-02 09:52:58 <tpbsd       > I have always loved the design process and never saw it as work, I guess thats why I dont like AI doing it
2026-02-02 09:53:08 <tpbsd       > yeah
2026-02-02 09:53:09 <KipIngram   > It never worked exactly the way I thought it should work.
2026-02-02 09:53:34 <KipIngram   > Oh yeah, that early architectural design is the most fun part, and I spend more time on it than I probably should.
2026-02-02 09:53:44 <tpbsd       > but for tossing ideas around at the design stage, it's ok, even helpful
2026-02-02 09:54:10 <KipIngram   > Totally agree.
2026-02-02 09:54:27 <tpbsd       > and the cool thing is that a lowly 45GB AI Deepseek image is perfect at that stage also
2026-02-02 09:54:31 <KipIngram   > You know, I think this fixed the fan thing.  :-)
2026-02-02 09:54:44 <tpbsd       > I dont need a online AI for that part
2026-02-02 09:54:49 <tpbsd       > awesome
2026-02-02 09:55:17 <tpbsd       > just my local 45GB deepseek variant is excellent for a 'pair designer'
2026-02-02 09:57:08 <KipIngram   > So, gold star for ChatGPT this morning.  :-)
2026-02-02 09:57:11 <tpbsd       > and anyway, we all know that the big AI centres arent making nearly enough to pay their bills, so all that stuff is probably going away soon
2026-02-02 09:57:58 <tpbsd       > my online AI probably isnt as it's OSS and free, the user only pays for the power etc
2026-02-02 09:58:09 <KipIngram   > Yes - that something I've thought about a good bit.  Sure, they can show off these really impressive accomplishments, but to get them they need these huge facilities with tons of electricity and so on - that just does not translate into "world changing" the way they market the whole business.
2026-02-02 09:58:20 <KipIngram   > People are not going to be running around with those capabilities in their pockets.
2026-02-02 09:58:39 <tpbsd       > moonshot/Kimi-K2 which is a pretty big AI, around 700GB image
2026-02-02 09:58:40 <KipIngram   > Pervasive real world applications are going to be far, far more modest.
2026-02-02 10:00:01 <tpbsd       > exactly, tho  moonshot/Kimi-K2 is like a 1000 lb gorilla of AI, it's always right when I ask it anything, in fact at times it seems annoyed that Im so dumb
2026-02-02 10:00:17 <KipIngram   > What we will get is the worst stuff - it won't surprise me at all to see the government willing to field those big facilities so they can use AI to track and monitor us all more closely, for example.
2026-02-02 10:00:25 <KipIngram   > Because profit won't be their motive.
2026-02-02 10:00:32 <tpbsd       > yeah, I agree
2026-02-02 10:00:45 <tpbsd       > it will be used for control if possible
2026-02-02 10:00:54 <KipIngram   > It used to be the case that if they wanted to watch someone they had to assign agents and so on, so it was a self-limiting process.
2026-02-02 10:01:00 <KipIngram   > Those limitations are falling away.
2026-02-02 10:01:08 <tpbsd       > AI is already built into the war machine
2026-02-02 10:01:32 <KipIngram   > Back then it had to be really worth it to them to home in on a particular person, but it's getting to the point where they can do it with everyone, by default.
2026-02-02 10:01:33 <tpbsd       > yeah, sadly youre 100% correct
2026-02-02 10:02:07 <KipIngram   > And what's really depressing is that very few of us even seem to think about it.  The mass population just bops along thoughtlessly.
2026-02-02 10:02:35 <tpbsd       > I used to scorn the govt claiming to 'see' what everyone was doing in the 1970's, id laugh it off, yeah you and whose army
2026-02-02 10:03:04 <tpbsd       > now with computer tech and AI, Im not laughing anymore
2026-02-02 10:03:21 <KipIngram   > We here in the West actually do still live in a world where we can have what we want from our government, but only if a substantial segment of the population cares enough about it to push for it.
2026-02-02 10:03:36 <KipIngram   > And we're not engaging with this one.
2026-02-02 10:04:33 <tpbsd       > hey did I mention "fabric" ?
2026-02-02 10:04:38 <KipIngram   > And we're so divided these days that on practically anything we do engage with it's damn close to 50/50, which is the perfect formula for leaving the government able to do whateve rit wants.
2026-02-02 10:04:39 <veltas      > rendar: Just wondering about your choice of 'if-true', the only languages I see that kind of verbosity in is test automation languages
2026-02-02 10:05:02 <veltas      > Like robotframework etc
2026-02-02 10:05:02 <KipIngram   > We fight over EVERYTHING.  As a default response.
2026-02-02 10:05:44 <tpbsd       > KipIngram, I use fabric (written in GO) to access my online AI from the CLI at my terminal
2026-02-02 10:05:56 <KipIngram   > I'm not familiar with that.
2026-02-02 10:06:08 <tpbsd       > KipIngram, it's very conveninet as I live at the CLI
2026-02-02 10:06:10 <KipIngram   > Will googling "go fabric" get me there?
2026-02-02 10:06:21 <tpbsd       > heh, no
2026-02-02 10:06:31 <KipIngram   > Me too, me too.  I mean, I do have my browser open, but I spend a ton of time in my CLI workspace.
2026-02-02 10:06:44 <KipIngram   > CLI and browser together are 99.9% of my computer interaction.
2026-02-02 10:06:54 <tpbsd       > but this will "#> fabric "why is my laptop fan extra noisy"
2026-02-02 10:07:10 <tpbsd       > or
2026-02-02 10:07:17 <KipIngram   > I just meant to learn about this fabric thing.
2026-02-02 10:07:26 <KipIngram   > Not to access it.
2026-02-02 10:07:34 <tpbsd       > but this will "#> fabric "why is my laptop fan extra noisy" > fan-noisy-advice.txt
2026-02-02 10:07:45 <veltas      > Ironically Google's language 'Go' is not very googleable
2026-02-02 10:07:52 <KipIngram   > :-)
2026-02-02 10:08:19 <KipIngram   > Ugh.  It's 3am here - I shouldn't even be awake.  My sleep cycle is the pits lately.
2026-02-02 10:08:30 <tpbsd       > fabric is just an app that enables AI access from the CLI
2026-02-02 10:08:39 <tpbsd       > ouch
2026-02-02 10:08:45 <veltas      > It did have some smart people on it though, I mean anything Ken Thompson worked on is automatically interesting to me
2026-02-02 10:08:46 <KipIngram   > I'll look into it - it sounds like something I'd like.
2026-02-02 10:08:57 <veltas      > (Golang, no fabric)
2026-02-02 10:09:02 <veltas      > I still need to try fabric
2026-02-02 10:09:05 <KipIngram   > Yeah, Thompson was usually into good stuff.
2026-02-02 10:09:23 <tpbsd       > KipIngram, it's awesome, browsers suck, one shouldnt need one to talk to a online AI
2026-02-02 10:09:31 <KipIngram   > He was part of the world where computers were tools, not toys.
2026-02-02 10:09:47 <veltas      > But also toys, I mean he made some early video games
2026-02-02 10:09:54 <tpbsd       > theyre tools here
2026-02-02 10:10:41 <tpbsd       > my retirement is spent making FLOSS projects
2026-02-02 10:10:44 <KipIngram   > That's true - fair point.
2026-02-02 11:02:23 <veltas      > Point of pedantry but a fair point of pedantry :P
2026-02-02 11:02:31 <veltas      > If referring to my point
2026-02-02 11:04:13 <veltas      > I emailed Bell Labs recently and they replied quite quickly, I was asking if I could get permission to use a historical copyrighted work of theirs, but realised I didn't need it after all
2026-02-02 11:04:20 <veltas      > I was a bit surprised at how quickly they got back
2026-02-02 11:04:30 <veltas      > Nokia Bell Labs as they're currently known
2026-02-02 11:05:12 <KipIngram   > Cool - I like it when folks are engaged like that.
2026-02-02 11:06:11 <KipIngram   > A couple of times I've emailed scientists whose work I learn about online and find interesting with questions, and have been pretty pleased to get actual replies from them in a lot of cases.  I think it may help that I can legitimately put "Ph.D." in my signature - I'm "in the club" so to speak.
2026-02-02 11:06:30 <KipIngram   > Donald Hoffman even offered to let me read a draft of the book he was working on at the time.
2026-02-02 11:06:34 <veltas      > Nice
2026-02-02 11:07:11 <KipIngram   > I'm sure part of it is that they're just as human as we are and are pleased when someone has a strong interest in their work.
2026-02-02 11:07:28 <veltas      > Dr Ingram
2026-02-02 11:07:46 <KipIngram   > Heh.  Once I left my univiversity job that moniker pretty much dieid.
2026-02-02 11:07:49 <KipIngram   > died
2026-02-02 11:08:04 <KipIngram   > It's just not that much of a "thing" in industry.
2026-02-02 11:08:17 <veltas      > I'm not a Dr but I can read papers etc
2026-02-02 11:08:24 <veltas      > I'm glad I didn't go into research personally
2026-02-02 11:08:35 <KipIngram   > I'm happy I did it, though - graduate school made me a much better thinker than I was post-undergraduate.
2026-02-02 11:09:19 <KipIngram   > I think I was just ready to approach the learning in a more mature way.
2026-02-02 11:10:08 <veltas      > What did you research?
2026-02-02 11:10:17 <iv4nshm4k0v > I've never finished my dissertation, but I consider my postgrad time well-spent.
2026-02-02 11:11:19 <veltas      > I was told I could go do a PhD based on my bachelor's result, and decided industry was a better choice, based on suggestion of basically everyone I respected at the time.  I did Computer Science and I don't think I really cared about any funded Computer Science research.
2026-02-02 11:11:51 <KipIngram   > I did my research in a lab called the "Center for Electromechanics" at The University of Texas at Austin.  The main focus was on "pulsed" EM systems, with applications like railguns and other types of accelerators.  Also did some electric vehicle research, and some fairly exotic one off projects, like a "single turn" tokamak for fusion, an electromagnetic torpedo launcher for the navy, etc.
2026-02-02 11:11:58 <veltas      > Also I really struggled with university environment generally, and found I fit a lot better into industry, so I think I made right call
2026-02-02 11:12:19 <KipIngram   > My specific research was in a type of accelerator that got less attention than railguns called a "coaxial accelerator" or "coilgun."
2026-02-02 11:12:28 <veltas      > Engineering basically?  Or physics?
2026-02-02 11:12:37 <veltas      > Regardless better than what I would have had to do
2026-02-02 11:12:45 <KipIngram   > Well, engineering formally, but with a fairly low-level physics based emphasis.
2026-02-02 11:13:05 <KipIngram   > Lots of electromagnetic theory, but basically with the displacement current term dropped from Maxwell's equations.
2026-02-02 11:13:13 <KipIngram   > They call that the "quasistatic approximation."
2026-02-02 11:13:22 <veltas      > Basically the best advice I've seen to CS students and professional programmers is to go 'applied' immediately
2026-02-02 11:13:23 <KipIngram   > Boils down to ignroming radio effects.
2026-02-02 11:13:28 <KipIngram   > ignoring
2026-02-02 11:13:51 <veltas      > I'd much rather work on programming related to your EM thing than e.g. study compilers for 4 years
2026-02-02 11:14:15 <KipIngram   > I've always quite enjoyed numerical modeling of physics systems.
2026-02-02 11:14:48 <veltas      > You can see this play out e.g. with UNIX where they did the bare minimum increments to solve different applications for other departments, starting with the patent office's typography
2026-02-02 11:14:48 <KipIngram   > And in these systems you had to go cross domain - you modeled the EM, but the currents caused heat and so you had to model heat flow and its effects on material properties and so on.
2026-02-02 11:14:55 <KipIngram   > And stress in the windings, etc.
2026-02-02 11:15:11 <KipIngram   > so it was a whole electromagnetic / thermal / mechanical mess.
2026-02-02 11:15:19 <veltas      > It sounds interesting, I confess my calculus probably isn't good enough to approach that without a lot of guidance
2026-02-02 11:15:23 <KipIngram   > With everything feeding back on everything else.
2026-02-02 11:16:02 <KipIngram   > The calculus came in during the planning - by the time you got to the programming it was mostly linear algebra.
2026-02-02 11:16:20 <KipIngram   > Huge matrix equations.  Though you were time-stepping differential equations, so in that sense it was still calculus.
2026-02-02 11:16:34 <KipIngram   > But that part itself was fairly simple at that stage.
2026-02-02 11:17:06 <KipIngram   > We had a Cray across the parking lot in another building, which at the time was a huge deal.
2026-02-02 11:17:14 <veltas      > I suck at linear algebra too (assuming it's the heavy kind I think, not the kind people tend to learn for 3D graphics)
2026-02-02 11:17:35 <KipIngram   > Linear algebra turns out to be one of the most important parts of math for physics.
2026-02-02 11:17:54 <KipIngram   > I was lucky and drew a good teacher for that course.
2026-02-02 11:18:18 <veltas      > I'm very good at 3D graphics linear algebra because I care about graphics, I'm terrible as soon as you ask me to e.g. prove the inverse matrix formulae or anything 'Jacobian' etc
2026-02-02 11:18:39 <veltas      > I did very badly at algebra at uni, I would like to learn it properly one day but for now I accept I'm illiterate at it
2026-02-02 11:18:40 <KipIngram   > Usually I'm a lot less interested in the theory side of math than in "using it," but I've gotten an opinion that linear algebra theory is pretty useful.
2026-02-02 11:19:34 <veltas      > It's necessary because a lot of problems are easier to handle if you can linearise them somehow, and sometimes linear in multi dimensions is better than non-linear in less/similar dimensions
2026-02-02 11:19:42 <veltas      > I get that bit at least
2026-02-02 11:20:00 <KipIngram   > But to show its importance, say you're studying vibrations in some solid structure.  That's a messy, complicated thing.  But if you do the right coordinate system change you can convert it from N degrees of freedom that are all coupled with each other into N separate decoupled problems.
2026-02-02 11:20:15 <KipIngram   > You can then solve those easily one by one, and then jump back to the old coordinate system with your solutions.
2026-02-02 11:20:36 <KipIngram   > That's all tied up with eigenvalues and eigenvectors - the eigenvectors are the basis vectors of the "easy" coordinate system.
2026-02-02 11:20:51 <veltas      > I do stuff like that with graphics, but mess it up a lot.  Reminds me I still need to debug some matrix maths on a graphics program I'm writing
2026-02-02 11:20:57 <KipIngram   > And the eigenvalues are usually the vibration frequencies of those separated systems.
2026-02-02 11:22:14 <veltas      > That hurts my head :)
2026-02-02 11:22:22 <KipIngram   > Have you ever used quaternions?  They're good for 3D graphics, because they dodge some of the pathologies you can run into with other methods.
2026-02-02 11:22:47 <KipIngram   > Geometric algebra can get you at some of the same improvements.
2026-02-02 11:22:49 <veltas      > I've implemented quaternions, they're really necessary for rotations
2026-02-02 11:23:02 <veltas      > I've come to 'understand' how they work a little bit by making mistakes with them and debugging
2026-02-02 11:23:16 <KipIngram   > Yeah.  I've come to believe that our standard methods for vectors and stuff aren't the best - we'd be much better off teaching our kids geometric algebra right from the start.
2026-02-02 11:23:40 <KipIngram   > It just handles things a lot more gracefully, and some of the stuff we teach (the cross product, most notably) only WORKS AT ALL in three dimensions.
2026-02-02 11:24:04 <KipIngram   > So the minute you want to go higher, say to 4D for relativity, your default tools don't work anymore.
2026-02-02 11:24:08 <KipIngram   > But GA does.
2026-02-02 11:25:01 <KipIngram   > The crux of it is that we like to think of a rotation as being "around an axis."  But that's a 3D thing - rotations are REALLY association with planes (rotation "in a plane.")
2026-02-02 11:25:14 <KipIngram   > And it's only in 3D that you can associate a unique axis with that plane.
2026-02-02 11:25:26 <veltas      > It's not surprising to me the cross product falls apart in higher dimensions, based on using it and knowing how it's defined/works
2026-02-02 11:25:31 <KipIngram   > That axis is the "dual" of the plane.
2026-02-02 11:25:41 <KipIngram   > But in 4D the dual of a plane is another plane.
2026-02-02 11:26:08 <KipIngram   > And the whole stinky business of "polar" and "axial" vectors goes away in GA too.
2026-02-02 11:26:16 <KipIngram   > It's another artifact of the same issue.
2026-02-02 11:26:56 <KipIngram   > A vector like a velocity is a "polar" vector (I think - I often get them backwards).  Cross products are "axial" vectors.  Just another way of saying that a cross product is just a different sort of thing.
2026-02-02 11:27:14 <KipIngram   > Geometric algebra explicitly recognizes the PLANE the rotation is in, as a thing called a "bivector."
2026-02-02 11:27:54 <KipIngram   > I'd seriously recommend you look into GA if you do a lot of this graphics stuff.
2026-02-02 11:28:27 <veltas      > The 4D one?
2026-02-02 11:28:35 <KipIngram   > https://www.youtube.com/playlist?list=PLpzmRsG7u_gqaTo_vEseQ7U8KFvtiJY4K
2026-02-02 11:28:45 <KipIngram   > No, for 3D stuff 3D GA is fine.
2026-02-02 11:28:56 <veltas      > I know about the stuff you're talking about
2026-02-02 11:29:02 <KipIngram   > Ok, cool.
2026-02-02 11:29:13 <KipIngram   > Sorry - more people don't than do.
2026-02-02 11:29:32 <veltas      > As I said my understanding of linear algebra for graphics is good, I just don't really know the theory and 'algebra' part, and the harder linear algebra stuff like scrolls / tensors etc
2026-02-02 11:29:46 <veltas      > Proving or deriving inverse matrix etc
2026-02-02 11:29:57 <veltas      > Is all a bit beyond me but very 'basic' to an algebraist
2026-02-02 11:30:13 <KipIngram   > Tensor stuff talks about "covariant" and "contravariant" quantities - that's more of the same "dual" business.
2026-02-02 11:30:32 <KipIngram   > We get to ignore it most of the time because the distinction dissolves in Cartesian coordinates.
2026-02-02 11:30:51 <veltas      > What about covalent? :P
2026-02-02 11:30:57 <KipIngram   > :-)
2026-02-02 11:31:03 <KipIngram   > Ugh - my chemistry is weak.
2026-02-02 11:31:12 <KipIngram   > Which is pretty bad, since my father was a chemist.
2026-02-02 11:31:26 <veltas      > I'm looking forward to my kids needing help with maths/science homework
2026-02-02 11:31:35 <veltas      > Would be nice to refresh on some of it
2026-02-02 11:31:52 <veltas      > Honestly I could do with refreshing my times-tables even
2026-02-02 11:32:02 <KipIngram   > I'm good with the more physics connected aspects of chemistry - how we get the atomic orbitals and so on.  But once you get into the real business of mixing stuff up in beakers and so on I'm quickly lost.
2026-02-02 11:32:51 <veltas      > I was lucky enough to do an internship at Pfizer for a week after winning an essay competition to design a new trial heart medicine based on a presentation they gave
2026-02-02 11:33:46 <veltas      > But I was never actually that good at chemistry, I think I'm just a lot better at writing scientifically than most of my peers were at the time and also I didn't work with anyone else (I think everyone else's essay ended up looking the same)
2026-02-02 11:34:52 <veltas      > The 1 week was just spent synthesizing the drug
2026-02-02 11:35:16 <veltas      > I remembered thinking the lab smelled like a petrol station (organic chemistry, no open flames in the lab etc)
2026-02-02 11:39:33 <KipIngram   > :-)  My dad was an organic chemist specifically, so I have a feeling I know that smell.
2026-02-02 11:39:59 <KipIngram   > Somewhere in the dim mists of my deep past.
2026-02-02 13:03:32 <veltas      > tpbsd: 50C sounds unbearable
2026-02-02 13:04:43 <tpbsd       > veltas, it is, Ive experienced 46C at this very desk one summer day a few years ago and it was so hot it killed my airconditioner
2026-02-02 13:05:16 <veltas      > I suppose you've got techniques beyond AC to reduce the health risks on especially hot days?
2026-02-02 13:06:06 <tpbsd       > and altho I didnt have a thermometer, I rode a single cylinder 350cc motorbike thru Marble Bar in the northwest of Western Australia in summer in a heat wave
2026-02-02 13:06:25 <tpbsd       > it was nearly fatal for me
2026-02-02 13:06:38 <veltas      > Wow
2026-02-02 13:07:46 <tpbsd       > all that kept me going was I discovered I could smell water and so rode from one very old briny well to the next. They were dug by Afghan camel drivers in the 1900's
2026-02-02 13:08:42 <veltas      > That's fascinating, all of it
2026-02-02 13:08:45 <tpbsd       > I had to push my motorbike to get it moving in 1st gear as it was so hot. Water drops on the gearbox ran around like on a stove hotplate
2026-02-02 13:09:24 <tpbsd       > so Id soak myself in briny water from the wells by lowering my helmet as a bucket
2026-02-02 13:09:34 <tpbsd       > thats all that kept me going
2026-02-02 13:09:57 <tpbsd       > if I stopped I was immediately covered in flies all over me
2026-02-02 13:10:00 <veltas      > Good thing that cylinder didn't fail
2026-02-02 13:10:23 <tpbsd       > oh and another interesting thing, birds kept dropping dead out of the sky
2026-02-02 13:10:54 <veltas      > Shame it didn't cook the flies too
2026-02-02 13:10:55 <tpbsd       > very insightful, and yeah I would be dead if that happened
2026-02-02 13:11:03 <tpbsd       > hehe, yeah
2026-02-02 13:12:24 <tpbsd       > on the way back thru Marble bar about a week later it wasnt as hot, not life threatening like a week before, but a guy I just met was on a brand new yamaha 400 and riding behind me
2026-02-02 13:13:17 <tpbsd       > I heard the engine die as it ran out of fuel, saw him bend down to turn his fuel on to reserve and BANG!
2026-02-02 13:13:25 <tpbsd       > the engine siezed
2026-02-02 13:13:43 <tpbsd       > so I had to tow him to the next town
2026-02-02 13:13:47 <veltas      > Oh dear
2026-02-02 13:14:07 <tpbsd       > that kind of heat and new two strokes are a no no
2026-02-02 13:15:06 <tpbsd       > nothing kills the damn flies except fly spray and old age!
2026-02-02 13:17:06 <tpbsd       > and it's funny, after soaking myself in the briny water including a cloth over my face, id push the bike off (and get to a maximum of 35 mph) and think, "hey it's actually nice weather, it cant have been as hot as I just thought it was"
2026-02-02 13:18:10 <tpbsd       > but as the water evaporated and my clothes dried Id think "geepers, I was wrong, it's not nice weather, it's so hot Im feeling faint and cant breathe"
2026-02-02 13:18:36 <tpbsd       > Marble Bar is the hottest plave in Australia
2026-02-02 13:18:40 <tpbsd       > place
2026-02-02 13:19:21 <tpbsd       > there were zero cars on the road, they would have boiled
2026-02-02 13:19:34 <tpbsd       > and dead birds were all over the road
2026-02-02 13:26:33 <rendar      > veltas: oh i see, do you think  the string 'if-true' is too long?
2026-02-02 13:30:39 <KipIngram   > rendar: It feels to me like "if true" is what "if" already is.  But I think we've discussed that before, a few weeks ago.
2026-02-02 13:31:14 <KipIngram   > So that makes my reaction be "yes, it's too long."  It contains redundant content.
2026-02-02 13:33:41 <veltas      > Yeah too long in the sense it's just longer than necessary
2026-02-02 13:34:04 <rendar      > KipIngram: well, in my forth you have `if-true A B else C D end`    or   `if-false X Y else Z end`
2026-02-02 13:34:18 <veltas      > And also idiosyncratic, I don't see a lot of languages that do it that way, and mostly not in general purpose programming languages
2026-02-02 13:34:35 <rendar      > basically the if-true pops out a cell from the stack, checks if its truthy, and if so executed A B otherwise C D
2026-02-02 13:34:45 <veltas      > Oh I think we all understand what it means
2026-02-02 13:35:13 <rendar      > veltas: because in other languages you have a condition! in a stack-based languge, the condition is *BEFORE* the if's word
2026-02-02 13:35:22 <rendar      > e.g. `2 2 eq? if-true A else B end`
2026-02-02 13:35:45 <veltas      > Why not use   2 2 eq? if   and   2 2 eq? not if    ?
2026-02-02 13:36:35 <rendar      > veltas: you mean `2 2 eq? if A B else C D end`      <- because in this way that 'if' may seem related to A B, but it's not
2026-02-02 13:37:35 <veltas      > WHat about:   if  2 2 eq?  then  A B  else  C D  end
2026-02-02 13:38:07 <veltas      > There 'if' just sets up the stack ready for 'then', and 'then' is your 'if-true' semantically
2026-02-02 13:38:31 <rendar      > no
2026-02-02 13:38:39 <rendar      > this is against a stack based language
2026-02-02 13:38:41 <rendar      > consdier this:
2026-02-02 13:38:46 <veltas      > But I will say for RPN it's generally better to just assume people understand it's all backwards and if it reads well then go with it
2026-02-02 13:38:52 <rendar      > `def my_condition if-true A else B end end `
2026-02-02 13:39:00 <rendar      > i can run `2 2 eq? my-condition`
2026-02-02 13:39:14 <rendar      > with your solution i can't, or, if i can, its waay more complicated
2026-02-02 13:39:18 <veltas      > True
2026-02-02 13:40:05 <rendar      > i think if-true and if-false are very clean and neat in a stack based lang, i think i can shorten them as ift or off
2026-02-02 13:40:06 <rendar      > iff*
2026-02-02 13:40:31 <veltas      > I think the big problem with "if-true" is a lot of people will end up asking the questions that KipIngram and I are asking
2026-02-02 13:41:44 <rendar      > for instance?
2026-02-02 13:42:00 <veltas      > For instance this conversation
2026-02-02 13:42:58 <veltas      > if-false I'd say makes a bit more sense because it reads a lot better than "not if" or the dreaded "0= if" which is what I tend to write in real life because the Forth standard gave up making a nice language a long time ago
2026-02-02 13:42:59 <rendar      > well, i get your points
2026-02-02 13:43:28 <rendar      > the point of a stack based language is that you have the condition BEFORE the if keyword
2026-02-02 13:44:39 <veltas      > I mean for years people have said Forth shouldn't be "if ... then" but "if ... endif" etc
2026-02-02 13:44:46 <rendar      > but maybe i coulod use just 'if' for 'if-true'
2026-02-02 13:45:00 <rendar      > yeah i think they are right
2026-02-02 13:45:04 <veltas      > Because a ton of people get confused by the meaning of "then" vs how it's usually used in languages
2026-02-02 13:45:14 <veltas      > It's correct but totally idiosyncratic
2026-02-02 13:45:52 <rendar      > yeah i guess so
2026-02-02 13:46:26 <veltas      > How will you do 'end'?  I was looking at pushing an xt to the stack so that 'end' could depend on the preceding syntax
2026-02-02 13:46:46 <veltas      > And you can use 'end' to finish any syntax then, as long as it pushes the correct 'end' semantics to the stack
2026-02-02 13:47:00 <rendar      > i think `if A B else C D end` would be fine
2026-02-02 13:47:22 <rendar      > and `if! A B else C D end` for if-false
2026-02-02 13:48:31 <veltas      > Confusingly the convention for "if-not" or "not-if" in Forth would be to write:  if-  or  -if
2026-02-02 13:48:46 <veltas      > If you didn't know, I don't know who knows what on here
2026-02-02 13:49:08 <veltas      > Every other person has never used Forth, or spent 40+ years writing Forth
2026-02-02 13:49:18 <rendar      > yeah
2026-02-02 13:49:52 <rendar      > but your points a very right
2026-02-02 13:50:06 <veltas      > I think maybe Thinking Forth has a list of naming conventions along this line?  I can't remember exactly which book.
2026-02-02 13:51:11 <veltas      > I tend to use ! to mean "abort on error", i.e. to make FILE-OPEN simpler I often define : FILE-OPEN! FILE-OPEN ABORT" Failed to open file" ;
2026-02-02 13:51:17 <veltas      > I'm not aware of other people doing this though
2026-02-02 13:51:56 <veltas      > sorry, OPEN-FILE I think it might be called
2026-02-02 13:52:05 <veltas      > But you get idea
2026-02-02 13:52:27 <veltas      > I don't tend to use file wordset, it's just dreadful
2026-02-02 17:16:44 <DKordic     > PicoLisp has "when", "unless", "while", "until" as variants of "if".  IMHO their result is "(Maybe a)", which PicoLisp does not define.  "Nothing" when Condition was not met.
2026-02-02 17:17:50 <DKordic     > if-not is distasteful.  Why should it be better than "not if"?
2026-02-02 17:20:01 <DKordic     > "then" is poetic!  No one recomended "[ if ... ]"?!
2026-02-02 17:37:10 <veltas      > I recommended it
2026-02-02 17:37:23 <veltas      > Among a slurry of recommendations
2026-02-02 17:37:55 <DKordic     > Oh, sorry then.
2026-02-02 17:44:53 <veltas      > No problem