<dTal>
I just wish you could do that kind of thing in pure OpenSCAD
<teepee>
me too
<teepee>
although not possible to run this missing the include
<dTal>
you only need Naca_sweep from thingiverse
<teepee>
what's that then:
<teepee>
include </home/dtal/myduct>;
<dTal>
you generate that with the included python script
<teepee>
ahh, 2nd script, did not see that
<dTal>
and change the path :p
<teepee>
I might give a go later, meeting coming up soonish
fling has quit [Remote host closed the connection]
fling has joined #openscad
<J22>
why shouldn't you not beeing able to do that in pure SCAD ? .. have done that lately (a loop Airfoil that changes the profile)
<Scopeuk>
dTal the "sharks tooth" boundary is interesting, its something I'm used to seeing on hot/cold stream mix for large jet engines not used to seeing it on a "simple duct"
<dTal>
yes you can do it in pure openscad in the sense that you can do any computation in openscad, and feed the results to Naca_sweep - however you end up reimplementing basic things like translate and rotate, just as Parkinbot does in Naca_sweep. You can't do it in idiomatic OpenSCAD, since you can't extrude a varying profile
<J22>
dTal yes i changed the length of the profile .. i could counteract by keeping the thickness equal .. or just scale X
<dTal>
once you're at that point, it's way less painful to do it in numpy. "Doing it in OpenSCAD" would amount to trying to write that Python script in pure openscad. I'm not even sure it's possible since I made heavy use of first class functions
<J22>
well i needed 2 lines of code and you tell me your 2 page long scripts are less painfull?
<dTal>
J22: it's easy if your goal is to make something that vaguely looks similar, but my spec is to have a circular inlet lip of constant radius, and the shroud wall must meet it at a tangent
<Scopeuk>
J22 yes I'm used to seeing them used to control noise via turbulence control on large turbofans, was just surprised to see it used for a ducted fan which only has cold stream
<dTal>
Scopeuk: well it's neither tested nor scientific, but the logic here is a bit different - the lower part of the duct is only really needed to hold the stator blades, which in turn hold the hub. There are 5 blades, so you can save a bit of weight by chopping out some of the "bits in betwee"
<Scopeuk>
I was quite happy to accept "it looked cool" was just curious if there was a reason like that, very neat
* Scopeuk
has historically done a lot of work with large turbo fans (software only)
<dTal>
J22: if you try to compensate by changing the thickness you will get a lumpy duct, no?
<J22>
maybe it is quieter .. the jet fans use more spiky Chevrons
<dTal>
Scopeuk: that's super cool! We design electric ducted fans for unmanned aircraft (drones, in a word)
<J22>
dTal why lumpy if all parts are equal in thickness?
splud has quit [Ping timeout: 240 seconds]
<Scopeuk>
my understanding of it on a large jet is you use to to break the boundary layer between hot and cold stream from the engine and cause the turbulent boundary to be distributed as that reduces the noise generated compared ot having a single incredibly turbulant boundary
<Scopeuk>
hot and cold stream being the bits that went through the core of the engine (where fire is) or through the bypass (where fire shouldn't be)
<Scopeuk>
dTal neat
<J22>
i mean SCAD doesn't have a native way to create Airfoil profiles .. but then you could do this with a chained hull () of them
<J22>
however nice result dTal
<dTal>
J22: how will you maintain a constant leading edge radius? you said you could just scale the thickness, but if you take a foil and decrease the chord while keeping the thickness the same, the leading edge will squash (increased radius)
<Scopeuk>
there are lots of ways to generate something "similar" but not identical, I think I would be looking to generate a profile in 2d sweep it into 3d and then cut wit ha bunch of angled large cylinders to generate the chevrons. but I think that is only superficially similar
<dTal>
Aye cheers :) what's your Ü library?
<dTal>
it is not so googleable
<dTal>
chaining hull() is no good if the profile is not convex, which is true of almost all airfoils due to camber
<dTal>
You definitely got closer than I would have expected very quickly
<dTal>
very impressive library, however it proves my point - you reimplemented half of openscad, because you can't get geometry back from the engine
<dTal>
now program minkowski() :p
af has quit [Ping timeout: 272 seconds]
<J22>
oh yes and you are right if your profile is convex hull is not an option ( thought it may be convex like a naca 2412 - used that as you can put paper around )
arebil has quit [Quit: My keyboard has gone to sleep. ZZZzzz…]
<J22>
didn't your scripts do the same - creating points for a polyhedron ?
af has joined #openscad
<dTal>
and if you want to do NACA65 profiles you'll be in for a very annoying time - no closed form equation, it's all interpolation of an empirical curve. And you don't get interpolation in openscad so you'll have to write that routine
<dTal>
ye-es, ultimately the python script does just generate points and I do have to implement a bunch of tedious geometry. But doing that in Python is much nicer than in openscad, you can do things like curry functions with lambdas so I keep the parametric curve as a function right until the very end, which means (among other thing) that I can numerically derive it at any point, which I use to do offset()
<J22>
offset for points is also something you need to write for SCAD
<dTal>
If you have to manually manipulate all the points, can't use any of the nice functions, not even translate and rotate, what's the point of doing it in .scad?
<dTal>
You're ultimately just using OpenSCAD as a way of dumping vertices into CGAL
<J22>
i think all modern airfoils are not described by geometry but by define the wanted pressure / lift and the geometry is simulated upon that
<J22>
yes sure SCAD doesn't have support for lofting or sweeping .. it is like InPhase close points
<Scopeuk>
its a common game, also the reason why when anyone in software asks for something to be "optimised" you have to dig pretty hard into what they actually want
<Scopeuk>
CPU time is different to optimising for memory usage and different again to minimise jitter
<dTal>
Are there OpenSCAD competitions? Who can reproduce a cunningly defined shape most accurately / shortest number of lines ?
<Scopeuk>
nothing "formal" but usually if you post something here and note that either it can't be bested or your not happy with it various people can't resist
<J22>
didn't game industry just used more and more space instead of optimizing at it was necessary when games couldn't update and need to fit on a CD (or 5 wingcomander)
<J22>
and then there is kkrieger
<Scopeuk>
games are generally bloated by texture data/graphical data at this point
<Scopeuk>
you can shrink stuff with compression etc but you lose performance elsewhere
<J22>
dTal i have seen some for fun contest on openscad on reddit
<myosotis>
I posted a 15 line 50ms render of a c-clip here
<myosotis>
within just an hour, the guys here had it down to 40 lines and an hour of render time
<Scopeuk>
myosotis you're model was the "victim" I was thinking of above
<Scopeuk>
s/you're/your
<myosotis>
this was a model that had already fulfilled it's purpose in life and was never to be printed again.
<InPhase>
Nerd sniping is a real condition.
<myosotis>
regular people just ignore the bait
<J22>
scopeuk if i had to guess it would be 1 Ω
<Scopeuk>
I remember the problem I also remember the full solution involves partial differentials, I can't remember the number, the reciprocals in the parallel function makes things "fun"
<myosotis>
just solve it practically
<dTal>
myosotis: lol so longer *and* took longer to render?
<myosotis>
and guess what value it's converging on
<InPhase>
To be fair I also made a 2 line version.
<myosotis>
the c-clip discussion was valuable, and I learned a lot of your guys contributions, I def. appreciated it
<myosotis>
have you guys watched the IT crowd? this convo reminds me of the countdown episode
<Scopeuk>
it's already a word!
<J22>
always fun when i explain to pupils that math has nothing to do with numbers - but with abstract logical thinking - and that is what you train .. not to sum up your bill
<myosotis>
the problem with math is that it's too pedantic
<myosotis>
the problem with non-mathematicians interpreting math...
<myosotis>
I thought I was kidding when I said pi=3 :(
<myosotis>
ignorance is bliss. now I'm very upset
<Scopeuk>
once of the biggest mistakes we make with children's education (secondary when they are a little older) is not making it clear to them that the huge set or seemingly random things they are being taught are giving them a basis to be able to work with all sorts of things they haven't been encountered yet. you individually may never pursue advanced
<Scopeuk>
mathematics directly but knowing some of the concepts will make things you do encounter easier to understand and it easier to communicate some problems
<myosotis>
j22 you are making my life immeasurably worse
<Scopeuk>
I'm going to exercises my own ignorance not read than and assume it was an even less successful sequel to the trillogy
<J22>
myosotis i stopped to tell those things - as it really make people depressed. ( and you have no idea how many of these things i know)
<myosotis>
the curse of the eternally curious
<J22>
without the news .. i wouldn't notice the war in Ukraine - or all the other bad things .. and when i was working with a news agency
<J22>
they told me that most of the news they have are not suited for selling as nobody could bring that information on TV or paper .. makes you think about the cave (allegory)
<myosotis>
wat cave
<Scopeuk>
news is a curious beast, largely its narrative based these days, that is the news networks fit stories into their existing messaging. "news wipe" was an interesting if satirical take on it
<J22>
“Beware that, when fighting monsters, you yourself do not become a monster... for when you gaze long into the abyss. The abyss gazes also into you.”
<myosotis>
that works for positive things as well though... I think people always forget that
<J22>
CAT PICTURES!
<J22>
and one of the reasons i don't watch horror films ( or advertisements) cuz - once it enters your brain you can't control it
<myosotis>
I just won't watch a lot of fiction because humans are limited by our imagination, and reality has no such limitations
<Scopeuk>
it would be like seeing std::string i; in a piece of c++
fling has quit [Ping timeout: 240 seconds]
fling has joined #openscad
N4buc0 has joined #openscad
<J22>
funny how this led me to reading about a "mach stem"
<J22>
WP should get a feature "people who read this also liked .. "
<Scopeuk>
I mean thats just hte automated version of people taging in "see also"
<J22>
and probably much worse .. as the "see also" has some obvious connection
<Scopeuk>
that does appear to be yet another constructive wave interference, ever come across the x51 waverider? aero lift by sitting on top of its own ultrasonic shockwave
<Scopeuk>
I think they still claim its a uav and not a missile but it does rather resemble a missile
<J22>
i remember how the thrust of an SR-71 mostly came from the shockwave cone within the inlet
<Scopeuk>
the engines on the SR71 are mad so many operating modes and moving parts
<J22>
and the success of a scram jet seems to be the pre cooler ..
N4buc0 has quit [Remote host closed the connection]
<Scopeuk>
was about to link to that, the temperature shift across is is bonkers
<Scopeuk>
they were trying to sell the pre cooler for use with turbo fans, iirc their test rig involved firing an F16 engine at full afterburner at the inlet of the precooler and having it drop the air to near cryogenic temperatures
<J22>
.. here my new AC for summer in 2030 after climate change .. ( also Sabre is not a scram jet)
<Scopeuk>
its combined cycle iirc
<Scopeuk>
turbo fan to mid altitude air breathing transitioning to scram and then liquid oxygen injection for single stage to orbit? its been a while
<J22>
i think it changed due to different versions
<J22>
but the name is for Air breathing rocket
<Scopeuk>
if you haven't read "Ignition! An Informal History of Liquid Rocket Propellants" that's well worth a go in this vein
<myosotis>
things that oxidize better than oxygen scare me
<J22>
everything above 12% you need to register as terrorist here
<dTal>
"The fact that its oxidizing ability surpasses oxygen's leads to corrosivity against materials often thought as incombustible. ClF3 is known to be capable of burning sand, asbestos, glass, ashes of materials that have already been burned in oxygen, and all but the most highly fire-retardant materials"
<buZz>
J22: its fun to remove all labels from containers and only put the molecule rendering on it
<J22>
buZz 3Dprinted scad models for the blind
N4buc0 has quit [Remote host closed the connection]
N4buc0 has joined #openscad
<buZz>
braille.scad
<buZz>
i'm sure someone made something like that :P
<myosotis>
I went to an ADA conference (usa) and saw this blind presenter navigate the internet using a screenreader. I was shocked to find his browsing speed roughly equivalent to a sighted person
N4buc0_ has joined #openscad
N4buc0 has quit [Remote host closed the connection]
arebil has joined #openscad
<J22>
well they are not slowed down by images and videos
arebil has quit [Client Quit]
<myosotis>
he could browser faster than me on well designed sites, but poorly designed websites were basically a black hole. it was pretty eye opening for me (am web dev-ish)
<Scopeuk>
myosotis it very much depends on the website, some are wonderfully well put together and accessible. others are a mess of constructs that ruin everything
<Scopeuk>
also Facebook like to arbitrarily change keyboard shortcuts and element names that break application integration for assistive technology, particularly in whatsap
<Scopeuk>
granted I mostly hear about it from someone that looks after an application that is usually for the inverse problem, people who can see and visually access the content but face challenges with the input side of it
<myosotis>
the bigger sites seem to care less somehow
<Scopeuk>
its a lot priority and they have the resources to move very quickly
<Scopeuk>
hopefully its something that will gain more traction moving forwards either through industry standardisation or legislation
<Scopeuk>
we now put so much of accessing our world behind the internet that its difficult to fully operate without it or at least it makes things more difficult
<Scopeuk>
a handful of standard javascript hooks for scroll page fixed amount, zoom and jump to next section/main nav/main body/next control would go a long long way in making it easy to stick a common interface on it, you can then let the user customise that interface to their need
pie_ has quit [Quit: pie_]
<buZz>
myosotis: at a callcenter i worked as dev, i implemented all accessibility stuff into our dynamic script stuff, with the help of a 'temp worker' , blind dude that hung out in my office for a month or something :)
<buZz>
it was nice that he got all my farfetched humor easily ^_^
<buZz>
but his screenreader/webnavigate speed was amazing, i dont even see sighted ppl move that fast
<buZz>
a well setup accessible webpage vs a improper setup page was amazing too, he was already fast but could go ~10-20x faster if layed out to certain rules
<buZz>
like how you should design cities without highways running through them if you value growing offspring
<buZz>
:P
<J22>
maybe a reason is that they have to .. e.g. a seeing person can use a keyboard with one finger .. while a blind has to type blind and remember all positions. Then also using a mouse works only for visual .. but using a computer without a mouse is much faster
<J22>
a programmer usully uses lots of shortcuts a normal text person wouldn't need as jumping around and inserting / changing / selecting parts is not usual if you write a text
<buZz>
i think a eye is maybe too much data for most ppl , and ear is easier to filter towards meaning
<buZz>
leading to more skilled interpretation when deprived of vision
<buZz>
but , sadly i never followed my gut feeling to go study clinical psychology :P
fling has quit [Ping timeout: 240 seconds]
fling has joined #openscad
pie_ has joined #openscad
<myosotis>
I think blind people can navigate properly structured web content faster than sighted people is the same reason productivity goes up if you live in the cli
<myosotis>
maybe it's more about areas of focus than content type, since cli vs gui is both visual, but the cli focuses your attention very narrowly
<myosotis>
or a less cluttered context surrounding the activity
trashbird has quit [Ping timeout: 252 seconds]
pa has quit [Ping timeout: 252 seconds]
pah has joined #openscad
ndnihil is now known as nihil
nihil is now known as ndnihil
fling has quit [Remote host closed the connection]
fling has joined #openscad
trashbird has joined #openscad
pah is now known as pa
<dTal>
so why don't sighted people use computers the way blind people do, if it's more efficient? Their abilities are a strict superset of blind people's
<myosotis>
the same reason AI's make the same decisions over and over again without exploration built in.
<myosotis>
people get stuck in their ways, it's more comfortable to do what you already know
<J22>
dTal same reason why most ppl use QWERT layout .. which was invented BEFORE 10 finger typing .. welcome to neo-layout.org
<J22>
one of the main benefits of QWERT was a simple selling trick .. you can spell TYPEWRITER from letters in the first line
<myosotis>
i learned to type reasonably well on dvorak
<myosotis>
enough to come to the understanding that knowing how to use existing equipment is MUCH more valuable than knowing a possibly-more-efficient layout
<J22>
you can have your layout on usb but more importand it is simpler than speaking different languages .. you just switch between
Guest3347 has joined #openscad
<J22>
but an interesting point as Apple is not compatible because they using "better" solutions
Guest3347 has quit [Client Quit]
<J22>
or why is windows / office everywhere .. and there are better options .. just because everybody knows it and uses it and writes software / driver for it
fling has quit [Ping timeout: 240 seconds]
fling has joined #openscad
N4buc0_ has quit [Quit: Leaving]
fling has quit [Quit: ZNC 1.8.2+deb2+b1 - https://znc.in]
fling has joined #openscad
ur5us has joined #openscad
lastrodamo has quit [Quit: Leaving]
gunnbr has joined #openscad
fling has quit [Ping timeout: 240 seconds]
fling has joined #openscad
myosotis has quit [Remote host closed the connection]
myosotis has joined #openscad
fling has quit [Ping timeout: 240 seconds]
fling has joined #openscad
<peepsalot>
teepee: you around? i'm considering organizing resources (icons, etc) further
<peepsalot>
so moving from top level or elsewhere into subdirectories of /resources: "sounds", "html" (currently 3 src/gui/*Dialog.html files), and "icons".
<peepsalot>
those cover the actual files from openscad.qrc, and then some other "resources" could be questionably moved there, like: fonts, examples, color-schemes, templates, shaders
<teepee>
maybe it would be useful to separate between "used at build time" and "will be installed"?
<teepee>
right now it's all a bit mixed up
<teepee>
it would simplify things quite a bit if make install could be essentially just a whole tree
<teepee>
but then, it's not very critical, moving stuff out of the top level folder is a good step I think
<teepee>
lol, quassel just said "day changed", at 1am
<teepee>
someone did not get the daylight saving message :)
fling has quit [Ping timeout: 240 seconds]
<peepsalot>
hm, yeah i didn't consider distinction of installed files. i was sort of thinking some of those installed would maybe make sense to subgroup as "config" (color-schemes and templates, maybe shaders)
fling has joined #openscad
<peepsalot>
on second thought, I don't want to muck up too much at once so I think I'll just stick to the qrc specific files for the moment.