teepee changed the topic of #openscad to: OpenSCAD - The Programmers Solid 3D CAD Modeller | This channel is logged! | Website: http://www.openscad.org/ | FAQ: https://goo.gl/pcT7y3 | Request features / report bugs: https://goo.gl/lj0JRI | Tutorial: https://bit.ly/37P6z0B | Books: https://bit.ly/3xlLcQq | FOSDEM 2020: https://bit.ly/35xZGy6 | Logs: https://bit.ly/32MfbH5
J2388 has joined #openscad
J23 has quit [Quit: Client closed]
teepee has quit [Quit: bye...]
teepee has joined #openscad
J2388 has quit [Quit: Client closed]
J2388 has joined #openscad
LordOfBikes has quit [Ping timeout: 260 seconds]
LordOfBikes has joined #openscad
J238897 has joined #openscad
J2388 has quit [Ping timeout: 260 seconds]
epony has quit [Remote host closed the connection]
<gbruno> [github] thehans opened issue #4617 (Non-determinism with manifold and minkowski) https://github.com/openscad/openscad/issues/4617
<gbruno> [github] thehans edited issue #4617 (Non-determinism with manifold and minkowski) https://github.com/openscad/openscad/issues/4617
<peepsalot> finally got around to writing that up. didn't help that thunderstorm knocked out internet for 12hrs last evening
<peepsalot> i wonder if fiber internet is generally less susceptible to downtime from lightning strikes
<peepsalot> still eagerly waiting for fiber install to finish in my neighborhood so i can finally say goodbye to spectrum
<peepsalot> née Time Warner
<InPhase> peepsalot: For below ground, probably not much difference, as it will come down to the power supply to the upstream coms equipment.
<InPhase> For above ground, a slight difference. Lightning to a cable line could damage equipment. Although only slight, because tree limbs are going to take them both out really easily, and that's probably the big hazard.
<InPhase> I've had below ground cable go out due to water damage before though. Below ground fiber I've never once had go out from any sort of weather event over... 5 years? As long as I ran some sort of power to the in-house equipment I got Internet.
<InPhase> A few brief outages from other unknown issues thought.
<InPhase> s/thought/though/
<InPhase> Maybe that was 7 years of fiber.
<peepsalot> dang, lucky
<peepsalot> i'm so jaded wrt US telecoms that I didn't think I would even get fiber to my house in my lifetime
ur5us_ has quit [Remote host closed the connection]
ur5us_ has joined #openscad
Guest46 has joined #openscad
ur5us_ has quit [Quit: Leaving]
snakedGT has joined #openscad
peeps[zen] has joined #openscad
qeed__ has joined #openscad
rogeliodh8 has joined #openscad
snaked has quit [Remote host closed the connection]
pa has quit [Ping timeout: 261 seconds]
peepsalot has quit [Remote host closed the connection]
rogeliodh has quit [Quit: Ping timeout (120 seconds)]
rogeliodh8 is now known as rogeliodh
qeed_ has quit [Remote host closed the connection]
pah has joined #openscad
Guest46 has quit [Quit: Ping timeout (120 seconds)]
use-value has quit [Remote host closed the connection]
use-value has joined #openscad
Guest20 has joined #openscad
Guest20 has quit [Ping timeout: 260 seconds]
L29Ah has quit [Ping timeout: 252 seconds]
la1yv has quit [Read error: Connection reset by peer]
la1yv has joined #openscad
Bocaneri has joined #openscad
Bocaneri is now known as Guest4705
fling has joined #openscad
Sauvin has quit [Ping timeout: 276 seconds]
fling has quit [Remote host closed the connection]
fling has joined #openscad
SamantazFox has quit [Remote host closed the connection]
SamantazFox has joined #openscad
guso78 has joined #openscad
snakedGT has quit [Quit: Leaving]
Guest4705 is now known as Sauvin
qeed__ has quit [Quit: qeed__]
L29Ah has joined #openscad
qeed has joined #openscad
epony has joined #openscad
fling has quit [Ping timeout: 255 seconds]
fling has joined #openscad
L29Ah has left #openscad [#openscad]
L29Ah has joined #openscad
Guest41 has joined #openscad
<Guest41> Is there an elegant way to define a path that would be intersection of a helicoid with (an extrusion of) a rounded-rectangle?
<guso78[m]> New PR from Robert Schiele can so that. Using rotate_extrude He can so all Sorts of mechanical springs
<Guest41> HI gusto -- thanks for the input -- I'm completely new to openscad and wonder what a PR is? Otherwise just point me in its direction!
<Guest41> Iv'e found https://github.com/schiele
<buZz> PR is Pull Request
<Guest41> Ah a git thing
<Guest41> Ok -- so he has this feature in his github repository fork of openscad (https://github.com/schiele/openscad). I'll dig further. Thank you.
<J238897> Guest41 you know what a chained or segmented hull is?
<Guest41> J238897 -- no idea
<J238897> i wouldn't consider it as elegant but easier than to construct polyhedron if you are not familiar with it
<J238897> but i am also not sure what you try to do
<guso78> refer to this one... openscad has so much options to get extended with new features
<guso78> for your round rectangle i would place 4 cirlcles i the corners and build a 2d hull around it.
<Guest41> J238897 -- I've been building parametric 3D models of new generations of Ethernet transformers in a 3D Electromagnetic Simulator (CST Studio Suite) and evaluating performance. For one family the windings are formed on a core with a rounded-rectangular cross-section.
<Guest41> And found that VBA is a bit limiting for exploring the design space -- so wanted a simpler way to express my designs and explore the solid models before commiting to EM simulation.
<Guest41> gusto78 thank you
<J238897> you can use libraries like https://github.com/rcolyer/closepoints  or other with similar functionality
* InPhase salutes.
<InPhase> Guest41: If you can handle notions of vector rotations and coordinate translations, closepoints is really the best way right now to go about that sort of formula-driven structure.
<InPhase> Guest41: It has optional matrix transformation operators built into it, which you can use to help, or you can ignore them and just generate your own outline points, or mix and match to fit a particular need.
<Guest41> Brilliant J239997, InPhase and guso78 -- I've got much to digest, but have a clear idea of where to start -- thank you!
<J238897> Guest41 something like this https://imgur.com/a/AU5vuDu ?
<Guest41> J238897 -- just so
<J238897> the problem is that you need to segment the straight parts of the rounded rectangle to get the same pitch as in the corners
<Guest41> Yes -- but that can be done;)
<Guest41> J238897 that is very compact
<InPhase> Guest41: Did you mean a helix, or a helicoid? (You said helicoid, but a winding sounds more like a helix.)
<Guest41> A winding around a cylindrical core would be take the path of a helix.
<Guest41> In general a winding around an extrusion of an arbitrary cross-section would take a path that is the intersection of a helicoid with an extrusion of that cross-section.
<Guest41> That is just the path of the winding, though. The winding iself would in-turn be an extrusion along this path.
<InPhase> Are the wires themselves round?
<Guest41> Yes -- conductor and insulation are round -- and co-axail.
<InPhase> I need to run an errand promptly for my kid, but I can illustrate that in closedpoints pretty quickly in about 30-40 minutes from now. It should be like a 5 minute design.
<Guest41> the insulation is an anulus
<Guest41> bargain! you are very helpful!
J238897 is now known as J23
<J23> the verbose part is within the library
<Guest41> J23 -- that's the way I like it.
<Guest41> What library did you use?
<Guest41> I'm gona fetch a coffee...
<Guest41> back in 5
la1yv has quit [Read error: Connection reset by peer]
la1yv has joined #openscad
<J23> included <ub.scad>  // top of the code
<Guest41> J23 oh -- ub.scad is something you wrote and uploaded together with the snip I can see?
<J23> are you on the link where you see the rendered model on the right?
<Guest41> yes
<Guest41> or the lib from there
<J23> the link  https://ochafik.com/openscad2/  ?
<J23> you can see a drop down where you can select libraries .. however you can also take it from git
pah is now known as pa
<Guest41> J23 Oh yes, I see the dropdown now. Its a library mention at https://openscad.org/libraries.html no less.
use-value has quit [Remote host closed the connection]
use-value has joined #openscad
<J23> Guest41 https://imgur.com/a/PmtDjFC  this is the best i can do .. else i need to build the path from scratch  ( or write a subdivision module )
<Guest41> Looks good. How did you do it?
<J23> it is basically the same code,  but i used a subdivided square ( "sq()") as input and then a modifier module to round the edges ..  but it turned out that  this was omitting points that are in line - so i need to change that function in the library so these points are still there
<J23> maybe a better way would be to change the lift after each corner
<J23> but the error is also only visible for high pitch .. if the windings are close to each other it is barley noticeable
<Guest41> J23 great. I'm encouraged. I will commit to learing enough to do the whole job with OpenSCAD.
<J23> just read how polyhedron are build .. so you generate the points and faces  -  but there is a reason why people wrote libraries to do this ..  it is much easier to use chained hull
<Guest41> thank you all for your help, it is much appreciated!
<J23> so this works fnx=10;
<J23> path=[for(i= [0:40],j=[0:fnx-1])shape[(i*fnx+j)%len(shape)]+[0,0,j*.15+i*3]];
<J23> so  each corner is lifted and then the next has a bigger lift to compensate the  segment distance in the straight section
Guest5742 has joined #openscad
Guest5742 has quit [Client Quit]
Guest41 has quit [Quit: Client closed]
Guest41 has joined #openscad
Guest41 has quit [Client Quit]
Guest41 has joined #openscad
<Guest41> Thank you J23 -- that code works for me. I'll use it.
<Guest41> BFN
<J23> you are welcome ..  keep in mind that  fnx= fn/4+1;    in case you render this  or  fn changes
<Guest41> Ah ! Ok. I'll comment that.
L29Ah has left #openscad [#openscad]
<J23> https://bpa.st/YVFNA   here the full code - also with a circular wire around
L29Ah has joined #openscad
<Guest41> Thanks J23 -- got it.
Guest41 has quit [Quit: Client closed]
<InPhase> Aww, he left. I got distracted, but then finally got it done. https://i.imgur.com/6j3v16X.png https://bpa.st/C26PI
<guso78> i should probably also focus on the available libs a bit more ...
<guso78> its really nice finish  ...
<guso78> BTW hope your kid is fine ...
<InPhase> Exactly spacing out the points for the rounded square and angling it just right for the upward tilt was slightly longer in code than I anticipated. I'll save that example and think about ways to abstract it better. It should be a more general primitive component I think.
<InPhase> guso78: Oh, yeah. He just needed a ride.
<guso78> i will also take it as input for my path_extrude (sanjeev has also shown recently the ability that there are many ways to bend along the way)
<InPhase> I like to test at extreme values to assess, which is where I quickly spotted I wasn't angling adequately.
<InPhase> You actually don't notice it at the values I put there for that image, even done sloppily, but if you make it really tall keeping the same winding count, it gets stretched and deformed from a cylindrical shape, unless you have the right tilt angle.
<guso78> traditional extrude  suffers from the fact that the profile(x-section ) is squeezed in the corners and you compensate for that by using any  points in the corner region.
<guso78> any->many
<InPhase> Just a rounded square is typically fairly easy, but the moment that helical pattern got mixed in, it got trickier. The arc-length spacing along the curved path and the spacing along the straight path also determines the point to point slope upward, so those have to match up to get the "natural" smooth rise.
<guso78> InPhase, yes you need to account for the altitude and the rotatational component when matching angles between "pieces"
<InPhase> It kind of made it a leaky abstraction, whereas I was trying to separate it. But, it's all coupled. :)
epony has quit [Remote host closed the connection]
epony has joined #openscad
<guso78> https://imgpile.com/i/hkNXvE right now trying to understand "apparently" obfuscated code. you create a tar file from empty directory, then you untar it again to get back what is known to be void ...
<guso78> I must have missed  something here ...
<InPhase> It's not empty.
<InPhase> That appears to be a tar-based copy via a temporary file.
<InPhase> tar is used sometimes as a portable way to recursively copy with preservation of permissions and dates and such.
<InPhase> That source examples directory is full of examples.
<guso78> InPhase, i got it. script started from wrong Working directory. i like the approach. tar is very useful to store file permisions and owners along each file.
<InPhase> Perhaps the mkdir confused you, but that's mkdir -p, meaning "no error if existing", so that an examples directory will always exist whether or not somebody has it in the repository. But, it's in the repository.
<InPhase> Oh I hate the tar thing, and think it's a terrible hack. ;) But it does have a good history of being portable, so it's sensible to use.
<peeps[zen]> guso78: what target platform are you concerned with in that script?
<guso78> Hi Peeps!
<guso78> right not bare linux
<guso78> now i found why its  not working for me
<InPhase> peeps[zen]: This is "release-common.sh" so I assume that's for all of them.
<guso78> gigl@fedora openscad]$ ldd openscad | grep not
<guso78>     libfive.so => not found
<guso78> "not" library was not found in my computer :)
<peeps[zen]> InPhase: it is, but for example the windows cross build exits the script early and handles most of the remaining code via cmake install / cpack
<peeps[zen]> last time i touched that script i removed 263 lines
<InPhase> Oh...
<guso78> peeps[zen], yep, this worked for me already!
<guso78> Its just a pity  that windows does not like the installer file and encourages the user *not* to run it because its not signed :') :')
<guso78> peeps[zen] you made the script much more  general using "ldd" to output the actual libray dependancy list
<peeps[zen]> did I? don't recall that part
<guso78> peeps[zen], I  see a lot of explicitely named library names are not present anymore present and now is see 'cp' cannot stat 'not found' library in my box instead XD
<peeps[zen]> I should have linked the full PR diff rather than the single commit: https://github.com/openscad/openscad/pull/4067/files#diff-a4362ecf51fd34a0619cd427ad45f151096710a4d4fab0e5e6cf74c347a38f87
guso78 has quit [Ping timeout: 260 seconds]
<peeps[zen]> but those changes were only meant to affected packaging of UNIX_CROSS_WIN
<peeps[zen]> I would like to other platforms using that . however, for example, even though cpack has a .deb handler, that's not kosher for actual debian release, they want you to use their own build tools iirc
<peeps[zen]> but if any mac users wanna try creating a bundle or whatever they do
epony has quit [Ping timeout: 240 seconds]
fling_ has joined #openscad
fling has quit [Remote host closed the connection]
L29Ah has left #openscad [#openscad]
fling_ is now known as fling
<peeps[zen]> the main point is that cmake should have intimate knowledge of all the dependencies, because it just compiled and/or linked them. so doing packaging in a shell script which doesn't know anything is a bit of a backwards approach
epony has joined #openscad
L29Ah has joined #openscad
<peeps[zen]> guso78[m]: if you are adding libfive as a depdency, it should check in CMakeListst.txt, and fail during configuration, not during packaging
<peeps[zen]> *dependency
<peeps[zen]> or maybe i'm misunderstanding the specific issue
<InPhase> There's already a PR with that added as a dependency, so hopefully we're not doing that twice. :)
<InPhase> Merge conflict messes would result.
Guest20 has joined #openscad
Guest20 has quit [Quit: Client closed]
caveman has joined #openscad
<caveman> does openscad use opencascade?
<InPhase> Nope. Very different approach.
<InPhase> I believe at least some parts of FreeCAD use it. But this also follows a different philosophy of gui-interaction centered design.
fling has quit [Remote host closed the connection]
fling_ has joined #openscad
<InPhase> OpenSCAD is programmatic CAD from start to finish.
<InPhase> The gui is basically a feedback system for interacting with the results of the code, plus functioning as an optional IDE.
<caveman> interesting. yes, i know about the programmability. however, i was wondering about the backend. apparently the backend is also totally different, which is good.
<caveman> i'm new to CADs. my experience is limited to: onshape, freecad. all i care about: (1) draw accurately, (2) assemble parts to test their joints, movements, etc, and (3) run FEA (e.g. calculix).
<caveman> i'd appreciate if you could share your experience (huge time saver for me) with respect to openscad and its ability in doing (1), (2) and (3).
<caveman> otherwise, i'll duckduckgo it (if you have better things to do than educating random internet people)
<juri_> never tried to run FEA stuff against my 3d printed objects.
<juri_> is any of it Free Software?
fling_ is now known as fling
<caveman> yes. calculix, which is used in freecad to implement their FEA (they call it FEM). it's very neat; helps me test my structures before embarrassing myself in reality failures.
<caveman> not sure if there are better ones than calculix. one limitation with calculix might be that it only operates on meshes, which i am not sure if it can support fancy joint properties, such as rotations.
snaked has joined #openscad
rawgreaze_ has joined #openscad