teepee changed the topic of #openscad to: OpenSCAD - The Programmers Solid 3D CAD Modeller | This channel is logged! | https://openscad.org/advent-calendar-2021/ | 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
<teepee> yes, that's part of it. I love my color dots driven by an wemos d1 mini :)
<JakeSays> lol
<teepee> so much better than the expensive hue stuff
<teepee> aha, day of the tree
ur5us_ has joined #openscad
<JakeSays> bah. i sure wish oscad had constants that could be exported when use<>'d
<JakeSays> or better yet, enum types
<InPhase> JakeSays: Use functions for this.
<InPhase> One of the key values of use is namespace isolation of variable names.
<JakeSays> InPhase: yes, i am, but i really don't like it
<JakeSays> InPhase: that's why i said constant or enum, so they wouldn't be variables
<InPhase> All variables are constants. :)
<JakeSays> yes, i know
<JakeSays> so then export
<JakeSays> 'export foo = true;'
<JakeSays> or just enums :)
<InPhase> We could, but it subverts expectations a bit. Then as a user of use you have to scan all around to see if anything is being exported that's disrupting your values.
<JakeSays> well, that's where the compiler becomes your friend
<JakeSays> "hey, this is used!"
<InPhase> A syntax that I would find reasonable is the following: foo = use<foo.scad>;
<JakeSays> ahh that would be perfect!
<InPhase> Then foo.x could be an accessible variable.
<JakeSays> well, i still wouldn't want variables exposed
<JakeSays> i consider them private
<JakeSays> actually, no that'd be fine
<InPhase> Well we could consider a few options for that.
<InPhase> I think it's less hazardous to expose them in that scenario. There are no secrets really.
<InPhase> It would only disrupt things like tab completion.
<InPhase> They are not mutable so the thing used cannot be disrupted.
<JakeSays> tab completion is important
<JakeSays> so if a compile error happens in a use<>'d file the compiler doesn't report it?
<InPhase> It's supposed to be reported, along with where it happened.
<InPhase> There was an old bug back... 2019? that messed up line numbers.
<JakeSays> it's not showing up, and everything else in the file is reported as unknown
<InPhase> Pastebin it?
<JakeSays> which part?
<InPhase> Enough to reproduce the problem I guess. :)
<JakeSays> is this not legal? translate([1,1,1]) nw ? cylinder(r=r, h=s[2]) : cube([r, r, s[2]]);
<JakeSays> the use of the ternary operator
<InPhase> It's not.
<JakeSays> ah ok
<InPhase> Use if/else there.
<JakeSays> that's the compile error that should've been reported
<JakeSays> like this? translate([1,1,1]) if (nw) cylinder(r=r, h=s[2]) else cube([r, r, s[2]]);
<JakeSays> i'm guessing no since it doesn't compile
<InPhase> b=false; translate([7,7,7]) if(b) cube(5); else sphere(5);
<InPhase> The unintuitive semicolon is important.
<InPhase> You can add brackets to make that more intuitive.
<JakeSays> ah right. ok cool
Jack2162 has joined #openscad
Jack21 has quit [Ping timeout: 256 seconds]
fling has quit [Ping timeout: 256 seconds]
<JakeSays> match would be a handy construct as well
<InPhase> Well we have the abomination called search.
<InPhase> That was a well-intentioned thing that is pretty chaotic.
<JakeSays> oof
<JakeSays> what's the easiest way to round a single side of a cube
<peepsalot> easy on typing or easy on cpu to render?
<InPhase> JakeSays: Make it in 2D and extrude.
<InPhase> JakeSays: Unless you want 4 edges of one side.
<JakeSays> i just need to round one side of a box
<JakeSays> one of the four vertical sides
<JakeSays> well, vertical 'corner'
<JakeSays> wow. just came across bosl2. what a beast
<JakeSays> lol it has a logo interpreter in it
<JakeSays> it is impressive though
<peepsalot> JakeSays:
<JakeSays> peepsalot: ah thank you!
<peepsalot> i chose a $fn divisible by 4 so there's no discontinuity on the rounded edge
<peepsalot> if you use an "arc" function you can avoid any boolean ops and just define the polygon directly before extruding: https://bpa.st/2OXA
<peepsalot> easy one-liner module
<InPhase> I do have a whole smooth primitives library, but I didn't implement a cube only rounded on one edge. :)
ferdna has joined #openscad
<JakeSays> i needed it for this: https://pasteboard.co/IkmasdeuuIlS.png
<JakeSays> actually this: https://pasteboard.co/s28HtTbZRrCp.png
Jack216272 has joined #openscad
Jack2162 has quit [Ping timeout: 256 seconds]
king has joined #openscad
king has quit [Client Quit]
foul_owl has joined #openscad
ur5us_ has quit [Ping timeout: 240 seconds]
foul_owl has quit [Quit: WeeChat 2.3]
foul_owl has joined #openscad
foul_owl has quit [Client Quit]
foul_owl has joined #openscad
foul_owl has quit [Client Quit]
foul_owl has joined #openscad
foul_owl has quit [Quit: WeeChat 2.3]
foul_owl has joined #openscad
<M6piz7wk[m]> oh the square has to be inside difference
<M6piz7wk[m]> hmm
<M6piz7wk[m]> is there any smarter way to define a triangle that can be centered then using the polygon that is giving me an anxiety
<M6piz7wk[m]> ah cylinder with lowered $fn
<M6piz7wk[m]> yay
<M6piz7wk[m]> meh that doesn't work u.u
arebil has joined #openscad
<M6piz7wk[m]> is there a way to center a polygon
arebil has quit [Quit: My keyboard has gone to sleep. ZZZzzz…]
<InPhase> M6piz7wk[m]: circle(r=25, $fn=3);
<M6piz7wk[m]> i need like rectangle with a b being futher apart then c b
<InPhase> What are you considering "centered" for that?
<InPhase> https://bpa.st/TPKQ Vertex center is pretty straightforward.
<InPhase> A center-of-mass center would be a lot trickier.
PovilasCNC has joined #openscad
fling has joined #openscad
Jack2162728 has joined #openscad
Jack216272 has quit [Ping timeout: 256 seconds]
GNUmoon has quit [Ping timeout: 276 seconds]
<peepsalot> M6piz7wk[m]: do you mean triangle again, since you only list 3 points? good news is for triangles the average of the vertex coordinates is the same as the center of mass
<peepsalot> and there is similarly a formula you can use for a general polygon, which given just below that.
ferdna has quit [Quit: Leaving]
<M6piz7wk[m]> To clarify i am trying to make a generic alluminium extrusion function where i wanted to do a trapezoid for the outer geometry, but i couldn't figure out how to center `polygon()` (which seems like the optimal answer and would prefer it to be polygon i guess..) so i wanted to make that shape out of rectangle and two triangles which is painful to maintain..
<M6piz7wk[m]> <InPhase> "https://bpa.st/TPKQ Vertex..." <- checking
<peepsalot> if you just want to center something about its rectangular bounds, then that is dead simple
<M6piz7wk[m]> <InPhase> "https://bpa.st/TPKQ Vertex..." <- seems perfect, but reading the code seems more complicated then comfortable
<M6piz7wk[m]> peepsalot: it's trapezoid and i don't know how to calculate it's max width and lenght to deduce width/2 lenght/2 for centering
<M6piz7wk[m]> since it accepts coordinates on the axis that are painful to convert
fling has quit [Ping timeout: 240 seconds]
<peepsalot> trapezoids on that extrusion image are only centered on one dimension (edge)
<M6piz7wk[m]> i need their center centered on 0 0 0 coordinate to painlessly define the relations against other points
<M6piz7wk[m]> since the design is define one side and then for loop it for the rest
<peepsalot> if you want to rotate trapezoids to replicate that shape, you only need the center of the circular hole to be at the origin.
<peepsalot> trying to "center" the trapezoids themselves is needless complication
<M6piz7wk[m]> the design is scalable so i want to define center of the trapezoid and then lenght/2 + lenght*0.04
<M6piz7wk[m]> (offset from the edge might need checking)
Jack2162728 is now known as Jack
<Jack> 20 lines of code ..
<Jack> (i know it is not identical .. just to show you how i would do that)
<M6piz7wk[m]> image not found for the pasteboard
<M6piz7wk[m]> nwm works.. just blocks tor
GNUmoon has joined #openscad
ur5us_ has joined #openscad
peeps[zen] has joined #openscad
peepsalot2 has joined #openscad
peepsalot has quit [Ping timeout: 240 seconds]
peeps[win] has quit [Ping timeout: 240 seconds]
arebil has joined #openscad
arebil has quit [Quit: My keyboard has gone to sleep. ZZZzzz…]
GNUmoon has quit [Remote host closed the connection]
GNUmoon has joined #openscad
lastrodamo has joined #openscad
ur5us_ has quit [Ping timeout: 256 seconds]
TheAssassin has quit [Remote host closed the connection]
TheAssassin has joined #openscad
arebil has joined #openscad
lastrodamo has quit [Quit: Leaving]
lastrodamo has joined #openscad
ur5us_ has joined #openscad
ur5us_ has quit [Ping timeout: 252 seconds]
<gbruno> [github] brainstorm opened issue #4022 (PushPullFeeder assertion errors). https://github.com/openscad/openscad/issues/4022
<gbruno> [github] brainstorm edited issue #4022 (PushPullFeeder assertion errors). https://github.com/openscad/openscad/issues/4022
arebil has quit [Quit: My keyboard has gone to sleep. ZZZzzz…]
mhroncok has joined #openscad
<InPhase> M6piz7wk[m]: It's typically easier to create things in the manner you need it, rather than try to morph the wrong thing over into what you need. This is all four trapezoids starting out where they need to be: https://bpa.st/CJIA
<InPhase> M6piz7wk[m]: There were also other ways to do that. For example, you could make one big square, cut one square out of the middle, and cut out two rectangles at 45 and -45 degrees for the cross-slits.
fling has joined #openscad
miique has quit [Ping timeout: 245 seconds]
<M6piz7wk[m]> i see O.o i try to experiment with it thanks
arebil has joined #openscad
jfj39 has joined #openscad
<jfj39> Hi
<jfj39> With Ubuntu 20.04, in Freecad, to use scad file, i need to set in parameters the path of executable OpenScad; but i dont know were it is.  Does any one knows ? Thanks in advance
<buZz> edit the .scad file and set the parameters inside it
<InPhase> jfj39: That depends on how you installed it. But just open a terminal and type "which openscad"
<buZz> oh
<buZz> derp :P
<InPhase> buZz: Yeah, sounded like a misread. :)
<buZz> ^_^
<jfj39> Thanks a lot "which openscad" gave me answer :-)
<InPhase> jfj39: Note that the "which" command works for anything in Linux. It's very frequently convenient.
arebil has quit [Quit: My keyboard has gone to sleep. ZZZzzz…]
GNUmoon has quit [Remote host closed the connection]
GNUmoon has joined #openscad
PovilasCNC has quit [Ping timeout: 240 seconds]
jfj39 has quit [Quit: Client closed]
<JakeSays> Jack: that's cool but it doesn't model the original
<Jack> ..  it could if it was important
<JakeSays> well, when one is trying to reproduce something, yeah it's important
<JakeSays> Jack: hey have you ever used those pla+ filaments?
<Jack> you already had it done - so no need or?   Also i have no specifications and even don't know how it should be used -  it was just as you asked and said you want to make familiar with the lib  .. so i just made a poc  ..  or if you needed squares with different rounded edges ..
<JakeSays> Jack: ah ok. i appreciate it.
* Jack has barely used PLA - Ü
<JakeSays> Jack: this is what i was modeling: https://m.media-amazon.com/images/I/91WHiON-npL._AC_SL1500_.jpg
<JakeSays> oh and i finally figured out how to use support blockers in cura. lol
<buZz> Jack: PLA+ isnt PLA ;)
<buZz> its related though
<buZz> but they wouldnt go into the same bag for recycling, for instance
<Jack> i understood what is was - but not if it only should look like it  or if e.g. the dias really need to be a separable  object (print with multi material)
<buZz> ?
<JakeSays> Jack: it won't be printed. it's a 'vitamin' (odd name)
<buZz> plastic recycling indeed doesnt happen on appearance , but on material
<Jack> PLA+ is a PLA blend and i haven't touch any of these
<buZz> PLA is awesome, highly recommend
<buZz> PLA+ is ok , if you need some impact modified material and dont want to increase temps too much
<buZz> all repraps i ever built were PLA parts only (beside the vitamins)
<buZz> only very rarely had to replace one part with PETG or ABS for heat treatment
<buZz> Jack: its pretty common in reprap land
<buZz> (that term)
<JakeSays> Jack: i still need to resolve my petg issues. i have three roles of petg that i need to use up
<buZz> oh
<buZz> this isnt even #reprap , lol
<JakeSays> Jack: yeah the nop<whatever>lib uses the term to represent parts that aren't printed
<buZz> nophead :)
<buZz> well known reprap designer
<JakeSays> yeah that's it
<JakeSays> it's a very cool lib
<Jack> https://reprap.org/wiki/Category:Vitamin   they even have their own language  - Ü great
<JakeSays> lol yup
<Jack> .oO for most animals Vitamin C is NOT a vitamin but a metabolite ..
<Jack> so  a vitamin is not describing a part/substance  but a  capability or better an inability
<JakeSays> Jack: yes
<JakeSays> in the reprap world
<JakeSays> but i just use them to represent parts i'm building a case for
<Jack> there are "parts with a hole"  "parts that hurt when stepped on and require a case"  ...  Ü
<JakeSays> this is where that board is used: https://pasteboard.co/aZzc7yz4eDlQ.png
<Jack> categorization is such a fascinating topic
<Jack> looks like i could be used in KiCad
<JakeSays> that proto board has a bunch of breakout mounted sensors on it. i'd love to create an actual pcb with the sensors directly mounted to it
<peeps[zen]> nophead is the name of the author... of nopSCADlib
<JakeSays> peeps[zen]: right
<teepee> well, the name is actually Chris (at least that's what he's giving on the github page) :)
<teepee> and he has some very detailed 3d printing stuff at https://hydraraptor.blogspot.com/
<Jack> people hadn't surnames  in the beginning .. so it is also kind of  categorization   as they just used their job as surname to identify someone from a different village whom they don't know.
<teepee> he does, and I know what it is, but I'm not sharing it unless I can find where he put that online himself
<teepee> I do know someone from Globalfoundries who actually just has a first name
<JakeSays> it's not necessary anyway
<teepee> which is fun with some software and forms which insist on having both :)
<JakeSays> lol
<teepee> like his email address was <firstname>.<firstname>@...
<JakeSays> he's a serial killer!
<teepee> I'm sure he's not :P
<JakeSays> lol
<teepee> he's in Singapore now but I met him a couple of years ago, and yes, nice guy :)
<teepee> I know they say that of serial killers too in TV
<JakeSays> having a buddy at a foundry could be convenient.
<teepee> but I guess that's mostly just TV
<teepee> not sure how that helps, it's not like I can easily order chips
<teepee> it's interesting to have a peek into the cleanroom when walking to meetings ;-)
<teepee> not that I did in-person meetings for 2 years
<JakeSays> might be able to get a discarded wafer or something cool like that
<teepee> well, if someone fills out all the forms, it's possible
<JakeSays> it is?
<JakeSays> i'd love to have one
<teepee> a collegue I worked with there for some years did get one as gift when leaving
<teepee> obviously those are not having anything interesting on there :)
<teepee> but they certainly look cool, but the new ones are quite big at 300mm diameter
<JakeSays> lol right
<JakeSays> that's the one i want - a 300mm wafer
<JakeSays> and yeah whats on it is irrelevant.
<teepee> on one or 2 desks they had wafers sitting around
<JakeSays> my high school electronics teacher had an 8080 mask and two wafers.
<teepee> you just can't wander off with those through the security gates :)
<JakeSays> .. actually it may have been an 8008 mask
<teepee> cool
<teepee> make a design and get google to produce it :)
<JakeSays> you could make out individual transistors. lol
<JakeSays> google?
<teepee> yes, google sponsors an open source program with a fully open source PDK on Skyworks 130nm (IIRC)
<teepee> the first batch is through but I think they planned a couple more
<JakeSays> ohhhh right. i totally forgot about that
<JakeSays> i remember looking at it, and the costs were very reasonable
<teepee> I have not followed closely, still need to catch up with the fun stuff
<JakeSays> i would love to make my own asic
<teepee> there's a video series from Matt Venn (I hope the name is correct) which shows some of the details
<teepee> I'm not sure if he's directly involved or "just" one of the open source customers
<teepee> oh, it's Skywater, not Skyworks, oops
<teepee> now we just need to convince Dag Lem to join the party and produce more C64 SID chips :D
<JakeSays> that looks like a fascinating video series
<JakeSays> someone needs to build a desktop asic printer. lol
<teepee> haha, yes, wasn't there someone creating transistors in the kittchen or something like that?
<teepee> botsmack!
<othx> OUCH!!!
<othx> teepee linked to YouTube video "Semiconductor Fabrication Basics - DIY Homemade NMOS FET/MOSFET/Transistor Step by Step" => 1 IRC mentions
<teepee> ahh, works only for videos, not the channel page :(
<lf94> "not some ugly nerd thing"
<teepee> ¯\_(ツ)_/¯
Jack has quit [Quit: Client closed]
Jack has joined #openscad
<InPhase> lf94: lol.
<InPhase> lf94: I thought something was familiar about that. EinPunkt was in this channel as Punkt back in early September working on modifying the same scad design.
<InPhase> I tried helping him but he was not communicating clearly at all, in the style of not providing adequate details of any of the problems, so it didn't work well.
<InPhase> He returned multiple times, usually seeking help with the same design. This was one example conversation, complete from start to finish. https://bpa.st/SILA
<InPhase> The very last conversation in which he participated was me saying, "<InPhase> Okay. Well I can't really help without complete self-contained explanations of what the issues are that you're having." He never replied, left, and never returned.
* InPhase shrugs.
<lf94> lol.
<lf94> wtf convo was that
<buZz> lol what a boring design too :D https://www.thingiverse.com/thing:3453133
<othx> buZz linked to "CNC 3018 Pro Z-Axis Upgrade by vespakoen" on thingiverse => 1 IRC mentions
<buZz> bonuspoints for putting the nonprinted part on foreground of photo
<Jack> buZz haha when i  first saw this i thought what a clean design looks industrial..  but now were you pointing out that it was the injection molded part i was judging  - Ü
<Jack> i think i like the last picture most
<buZz> :)
<gbruno> [github] t-paul pushed 1 modifications (Merge pull request #4023 from obliviand/feature/lib3mf-2.0
<gbruno> [github] t-paul closed issue #3709 (Build process cannot find lib3mf library (vv. 2.1.1 and 2.0.1)). https://github.com/openscad/openscad/issues/3709
<Jack> why has find/replace not a case sensitve option ?
<teepee> only the most unhelpful answer comes to mind ;-)
<Jack> nobody had time to do it ?
<Jack> is this in your control or comes from the qscintilla code
<lf94> InPhase: libfive maintainer and curv community are going to start a "SDF User Group" if you're interested. It's going to be hosted as sdfug.discourse.group
<lf94> I'm going to be the admin; I'm waiting to hear back from Discourse if we're accepted into their "open source tier"
<teepee> why advertise if it's named to specifically exclude?
<lf94> What do you mean? I'm mentioning it because InPhase was interested in SDFs
<teepee> Jack: the find code is pretty much OpenSCAD driven I think
<lf94> And well, libfive and curv are using SDFs for CAD.
<lf94> and OpenSCAD is CAD. So it's related.
<teepee> I'm interested too, but this is still the OpenSCAD channel which is not SDF at this point
<lf94> I didn't want to create a "code cad" discourse because I believe a separate community is needed for that.
<lf94> It was not to exclude, you know how much I want to unite everyone in the code cad world :).
<lf94> The past few months I'm disappointed at the lack of Code CAD projects to be honest.
<lf94> There should be an abundance. There are so many ideas!
mhroncok has quit [Quit: Leaving.]
<teepee> I don't
<teepee> but lots of ideas is good
gunnbr has joined #openscad
<teepee> Jack: looks like we just use QString.indexOf which has a flag for case sensitive
<Jack> teepee  just was writing that the highlighting is case sensitive
* Jack was wondering that some replaced were white and some yellow background
<teepee> yeah, indeed it seems the actual find is part of qscintilla but the highlight is not
<teepee> and yes, that call is casesensitive = false
<teepee> let me quote dr.who even though I've not seen that for a while...
<teepee> that's extremely very not good
<lf94> I found a nice imgui text editor to embed in Curv. OpenSCAD should consider something similar.
<lf94> No more needing Qt
<lf94> Could also bring it one step closer to running in browsers
ur5us_ has joined #openscad
teepee_ has joined #openscad
teepee has quit [Ping timeout: 276 seconds]
teepee_ is now known as teepee
<gbruno> [github] t-paul unpinned issue #3640 (Next Release). https://github.com/openscad/openscad/issues/3640
othx has quit [Remote host closed the connection]
othx has joined #openscad
othx has quit [Remote host closed the connection]
othx has joined #openscad
othx has quit [Remote host closed the connection]
rvt has quit [Quit: ZNC 1.7.2+deb3 - https://znc.in]
othx has joined #openscad
rvt has joined #openscad
<teepee> othx: status?
<othx> Gthx.NET version 2.23 2021-11-11: OK; Up for 9 minutes, 16 seconds; mood: pretty good.
rvt has quit [Quit: ZNC 1.8.2+deb2+b1 - https://znc.in]
GNUmoon has quit [Ping timeout: 276 seconds]
<gbruno> [github] t-paul pushed 9 modifications (Merge pull request #4013 from ChrisCoxArt/fix_deprecations_and_warnings
<Jack> lf94  cadhub is cool but i don't want seeing oSCAD replaced by a "cloudbased" browser version
<InPhase> I would likely never swap my usage with an online system. But I would very much like compatibility with one for purposes of sharing designs.
<teepee> I see no point in replacing it
<teepee> but having a browser based playground would be a nice addition
<teepee> especially considering there already was a version running in browser already
<Jack> sure options options Ü
<Jack> for some rendering a thin client could be fun
<Jack> in the beginning i had designs rendering for 3 days
la1yv has quit [Read error: Connection reset by peer]
la1yv has joined #openscad
<ali1234> this new 3blue1brown video has a nice explanation of convexity https://youtu.be/ltLUadnCyi0?t=938
<othx> ali1234 linked to YouTube video "Alice, Bob, and the average shadow of a cube" => 1 IRC mentions
<Jack> and there is still that ugly image in the documentation
GNUmoon has joined #openscad
GNUmoon has quit [Remote host closed the connection]
GNUmoon has joined #openscad
GNUmoon has quit [Remote host closed the connection]
GNUmoon has joined #openscad
<LeviHarris-Brown> teepee I tried printing threads on my resin printer. They all feel a little bit tight (as in resistance, but I could still hand tighten them if I gripped hard after being >30mm deep in the thread), but so far M3.5 has worked well. I don't know about anything smaller as I don't have any threads I know for sure are smaller, but I'll update you again when I have some. I suspect the slight tightness could be made better by a slightly
<LeviHarris-Brown> better calibrated printer
<LeviHarris-Brown> (it was you that asked for the update?)
<ccox> for a resin printer and fine threads - it can help to use a "tough" or slightly flexible resin (found this out after about 20 iterations of optical SMA connectors)
<dalias> for threads make sure you follow the specs for the tolerance grades
<teepee> M3.5 is not bad, on a resin printer I would expect at least M3 to work fine. But it may need some tuning.
<dalias> there's supposed to be some gap, and you probably want to model for the middle of the allowed range, not the edge
<dalias> teepee, I can do M2.5 on FDM :-)
<peeps[zen]> is M3.5 a thing?
<dalias> oriented in the XY plane. oriented vertically would be hard
<dalias> peeps, i wasn't aware of it until recently but apparently it's a thing you can buy
<LeviHarris-Brown> I've printed all the way down to M1.2, I just don't have anything here to test it at the moment.
<LeviHarris-Brown> M3.5 is pretty common for things like plug sockets etc
<dalias> and some ppl like it for electronics assembly or something
<dalias> i'd generally prefer m3 or m2.5 for that
<LeviHarris-Brown> yeah quite common there as well
<teepee> yeah, I never used M3.5
<LeviHarris-Brown> dalias: how on earth do you do 2.5 on FDM? I don't have an FDM printer, but that seems really difficult?
<dalias> with a line width of 0.25 or so and orientation in the xy plane
<teepee> grr, now Docker is down :(
<teepee> welcome to cloud services
<dalias> that's about the minimum you can expect an 0.4 nozzle to resolve, but it's enough to work. would do better with 0.2 nozzle
<LeviHarris-Brown> teepee: when AWS went down the other day it was pretty hilarious seeing people who had bought cloud reliant smart bulbs not be able to turn their lights on. We like smart devices, but why on earth would you ever buy something like that that doesn't run locally, nevermind that is literally dependent on the internet to function
<teepee> yeah, I saw that about chrismas lights and cleaning bots :)
<dalias> because you like 0.2-5s lag turning your lights on :)
<teepee> problem is in some cases you may not even know the details of what they use
<teepee> even better was that cat feeder which moved to "pay per month" for use or it would turn into a very big and expensive paper weight
<LeviHarris-Brown> teepee: we have all smart lights, but they're all Zigbee Hue bulbs. And any other similar device I make sure is based on an ESP32/ESP8266 and just flash ESPHome to it. It's pretty easy to avoid cloud based ones if you spend a few minutes searching, but yeah sadly many people don't have the knowledge to do that, so eventually the company bricks their devices and creates tons of e-waste :/
<LeviHarris-Brown> teepee: wow that's crazy. Reminds me of that juicero, AvE did a teardown on it and it must have been like >$1500 per unit for make
<dalias> getting way OT but i think the problem is lack of regulation of the market. you shouldn't be able to sell such devices, either at all, or without a giant consumer warning that they may be bricked or changed to pay-to-keep-using at any time
<LeviHarris-Brown> yeah, the lack of security in the IoT sector caught up with them, and is now being legislated. I think the same will happen with functionality in the next few years
<Jack> i think this right to repair would also allow to repair broken mechanisms
<teepee> which does not help if the hardware is locked down
<Jack> the point is that hardware can't be locked down if you have the right to repair it
<JakeSays> i sure wish it wasn't. i have several phones i could do some cool things with
<Jack> i mean we see this in cars or more tractor where the old ones without electronic are worth a lot
<teepee> unfortunately that's not how it works in most cases, it would be nice if that's the case
<JakeSays> my pixels aren't supposed to be locked but i haven't tried doing anything with them yet
<Jack> problem is that people buy that shit ..  they buy DRM shit  and use Smartphones  ..
<ccox> The larger problem: most consumers have no idea they are buying risky items, they have no concept of what cloud dependent services mean (and imply could happen).
teepee has quit [Remote host closed the connection]
<Jack> yes and it is so convenient not to think ..  if you want something open it normally require more knowledge
<ccox> It cannot be a free market if not all the costs and implications are known. Thus the seller is ripping off the consumer based on the asymmetry of knowledge.
teepee has joined #openscad
<Jack> teepee  51×12 LEDs ?  is the nuclear reactor included?
<LeviHarris-Brown> <Jack> "the point is that hardware can't..." <- Right to repair wouldn't really allow us to easily reprogram devices. And it's not trying to do that either. That type of legislation would be ruthlessly fought as well, e.g. just imagine how hard video game manufacturers would fight against that. They already go through ridiculous amounts of effort to secure the system
<Jack> lobbyism is a pain in all areas
<Jack> or in cars where you can program but you need a 3k$  device to do it
<ali1234> not nearly as hard as the mobile phone networks and spectrum regulators
teepee has quit [Remote host closed the connection]
<Jack> smart people will always be i minority of under 20%  (the level of smartness may rise though)
teepee has joined #openscad
<teepee> oh, did that message go through, having some disconnects :/
<Jack> last msg was twitter link
<teepee> yeah, and at the time I pressed enter it did ping with the "disconnecting"
<teepee> but you don't need to push 20ma per color, that would give some extra warm light, literally
<teepee> I've seen a big self made led wall and standing in front was fun, and warm :)
<Jack> over 500LED will need some power and will heat this up ..  we had printed (PLA) a 64LED device that just melted down and deformed
ur5us_ has quit [Ping timeout: 240 seconds]
<ali1234> it doesn't need to use a lot of power as the driver chip is capable of pwm
<teepee> when it goes white, you could actually feel it from maybe 2 meters away :)
<ali1234> looks like 32x96?
<teepee> could be, 2 rows of 16x16 panels
<Jack> 5000 my guess
<ali1234> 32x32 panels are more common
<teepee> self made
<ali1234> unless someone soldered together a load of 8x8 ws2812 panels but that would be real expensive
<ali1234> even if you don't factor in your time, you can't build LED panels cheaper than chinese factories
<ali1234> the only problem is they only make standard size and shapes
<othx> Jack linked to YouTube video "LED Wall Virtual Production - Bye Bye Green Screen" => 1 IRC mentions
<ali1234> 64x64 LEDs 2mm pitch is like $25 shipped
<ali1234> that's absolutely ridiculous
<teepee> that was ~7 years ago, china panels were not that easy and cheap yet
<teepee> *easy to get
<Jack> if you build a 16k LED Wall - it is still an investment
<ali1234> well yes, that's why the individual panels are so cheap though
ferdna has joined #openscad
<ali1234> such a screen would be 64 metres wide and would contain 35000 64x64 subpanels, you're looking at well over $1m once you include all the driver circuitry
<Jack> for filmproduction they are bit higher
<ali1234> yeah, naturally you have labour expenses and other stuff. and everyone involved has to make a profit or its not worth doing...
<JakeSays> Jack: i'd love to have a wall that big
<Jack> 73 million LED
<JakeSays> so i have this part: https://pasteboard.co/rwp9wjIsFwnM.png and i was wondering if y'all had an idea of how to print it w/o supports (w/o turning it upside down)
<Jack> jakesays what is the angle .. could work but wouldn't look too nice
<JakeSays> Jack: which angle?
<JakeSays> of the top of the clips on the side?
<Jack> those which are not 90°
<JakeSays> the tunnels print fine w/o support. it's just the two green bits that need it
<Jack> you can do 70  but it will be rough
<JakeSays> the bottom surface of the clips need to be flat
<JakeSays> there's a cap that clips over them
miique has joined #openscad
<Jack> or postprocessing  and sanding
<JakeSays> i need about 90 of them so the less work the better
<JakeSays> if i flip it over it prints w/o supports but then the red holes in the top are distorted
<Jack> put the tips down and use little support
<JakeSays> what do you mean?
<ali1234> make it two pieces that clip together
<Jack> the green tips down
<JakeSays> Jack: that's what i meant by flipping it over. the clips face down and print fine w/o support, but red holes are messed up
<JakeSays> ali1234: hmm. that's an interesting idea
<JakeSays> also i bet the holes would print better upside down if i printed the infill first
<JakeSays> this is the result: https://pasteboard.co/hSmxNFOEJ9W9.jpg
rvt has joined #openscad
<ali1234> what is it?
<JakeSays> it's a solderless system for holding ws2812b rgb leds
<ali1234> like for fairy lights?
lastrodamo has quit [Quit: Leaving]