<InPhase>
erickrdini: You have hardcoded all of your values, so you will need to change all the locations of the numbers as well if you change the cube dimensions. It is a better idea to put the cube size in a variable, and calculate the number locations with math so that you don't have this problem.
<erickrdini>
Could anyone help my do that?
default_ has joined #openscad
sublim8 has quit [Killed (NickServ (GHOST command used by default_))]
<Scopeuk>
what I have done is converted your positions (27 or 33) to a fraction of the cube face by dividing though by your old size of 60 to keep them in the same relative position, and then multiplied that by the current cube size
<InPhase>
erickrdini: Please note that your design is unprintable.
<InPhase>
erickrdini: If you want to print it, 0 and 8 need vertical bars placed over the number (use cube after the difference) so that the center pieces of those numbers remain attached.
<InPhase>
And 6.
<InPhase>
The bars should grab from the same calculated values.
<InPhase>
Also, the 8 as you had it was upside down. If you want it right-side-up, switch that line to: translate([-num_offset, 0, 0]) rotate([0, 90, 180]) number("8");
bomb has joined #openscad
<erickrdini>
tks guys!
<erickrdini>
you help me a lot
ali12341 has quit [Remote host closed the connection]
ali1234 has joined #openscad
<erickrdini>
tks Scopeuk you help me a lot
<erickrdini>
you too InPhase ... tks
hendursa1 has joined #openscad
gunnbr_ has joined #openscad
gunnbr_ has quit [Client Quit]
Zenmaster43 has joined #openscad
<Zenmaster43>
I can't seem to get the ternary operators ? : to work. I always get a syntax error, ie. selectedVal = selector = 1? value1 : value2;
<Zenmaster43>
Never mind. I figured it out.
Zenmaster43 has quit [Quit: Client closed]
<InPhase>
:)
<teepee>
zen!
bomb is now known as honp
honp is now known as bomb
pah has joined #openscad
pa has quit [Ping timeout: 260 seconds]
<dalias>
is there any decent mesh fix tool with minimal deps?
<dalias>
want to import stl file but it's broken
Jack21 has joined #openscad
<Jack21>
wouldn't it be good to release scADVENT solstice - submissions days and then every day just one?
aiyion has quit [Remote host closed the connection]
<teepee>
dalias: admesh fixes some issues, but not always all openscad is picky about
<InPhase>
dalias: Mesh fixing is notoriously tricky.
<teepee>
Jack21: what do you mean, the previous calendars opened one door a day
<InPhase>
dalias: The thing that seems most promising to me is meshlab's mesh resampling routine, except for the problem that bad meshes often cause it to segfault.
<Jack21>
teepee: ah ok - i understood last time that it should be released 24.12
<InPhase>
That would be tricky as there is no 24th month in the year.
<Jack21>
and as we have things to print and decorate the tree with it..
<Jack21>
InPhase -- 24 Dec
<InPhase>
We have traditionally exploited the one-a-day release process to buy additional time to finish the designs for the later days.
<Jack21>
.oO why should i use month before days
<InPhase>
Because it's the international iso standard? :-P
<InPhase>
2021-11-05 is today.
<Jack21>
yes that is ok (with the year and dashes)
aiyion has joined #openscad
<InPhase>
Always most significant first, unless one is writing archaically with words like "On this 5th of November, remember remember..."
<teepee>
Jack21: no, the full repo went from private to public after christmas eve
<teepee>
the separate files according to each day via the webserver
<Jack21>
teepee ah this is what i had mistaken
<InPhase>
I guess it's time to start brain storming about some clever designs for it.
<Jack21>
already made 8 designs .. maybe someone like to check if they are ok - or what needs to change
<InPhase>
I should set myself a goal to design something that requires function literals.
<teepee>
I have 3 potential multi-part designs, mostly sure I can't finish all 3 :)
<teepee>
candle-arch (germany: schwibbogen), heng lamp style thingy, big christmas tree with led lights
<teepee>
well, the "led lights" part is sort-of renduntant as everything is supposed to have leds :)
<InPhase>
I don't mind a little fireworking now and then as long as people are using them safely. But some of those comments sound like people are using them to a lot more excess than I'm used to. I'm not sure if they're exaggerating, or there's some sort of chronic firework excess going on in the UK?
<Jack21>
module led() sprinkles leds over every design - Ü
<teepee>
no idea, it's guy fawkes day or something so really some kind of fireworks day/season
<teepee>
(according to wikipedia that is :)
<teepee>
yeah, led() would be nice, but it's probably tricky
<teepee>
for the christmas tree, I want to use those coated copper wire strips so there's a wire guide with a cut-out for each of the leds
<InPhase>
Here, fireworks are July 4th and January 1st, and sometimes also Chinese New Year depending on where in the U.S. you live (typically only a little bit in major cities). And then once in a while for other unrecognized events, which are possibly a small handful of birthdays.
<othx>
Jack21 linked to YouTube video "TOP 6 CRAZY HOMEMADE ROCKETS (FIREWORK)(720P_HD)" => 1 IRC mentions
<InPhase>
My only major problem with fireworks is the distinctly American one that they are sometimes indistinguishable from gunshots, depending on what pattern they are launched with.
<Jack21>
SMD 0603 (1,6 mm x 1,6 mm) RGB but without controller chip
<teepee>
without controller makes it difficult for very distributed designs
<teepee>
those rgb strips I linked are addressable ws2811 compatible
<teepee>
at least the ones I tried before, not actually tested the one I just linked, but it would be surprising if that one would be different
<teepee>
I have one running via a d1-mini with esphome, that's just crazy simple, cut annoying controller, solder 3 wires to the esp8266, upload esphome, done :)
<teepee>
now I need to try with the esp32c modules
<Scopeuk>
the apa102,s are actual spi for data iirc which is nicer for a lot of micros
pah has joined #openscad
<buZz>
yeah, they work directly from a pi's header
pa has quit [Ping timeout: 264 seconds]
<Scopeuk>
have been meaning to play but as I have no real plan tend to be bad at commiting money to it
<Scopeuk>
also I appear to not have brought any pi's yet
Jack21 has quit [Quit: Client closed]
pah is now known as pa
mhroncok has quit [Quit: Leaving.]
snaked has joined #openscad
<dalias>
teepee, thanks for the admesh tip. it didnt fix this but looks useful to have around
Jack21 has joined #openscad
<InPhase>
Classic mesh repair story: "I wasn't able to fix it, but remain hopeful someday I will be able to fix one." :)
<teepee>
I think specifically it does not fix zero faces
<InPhase>
One thought I've had but never tried, and haven't seen, is that mesh repair might be consistently possible if you can ignore faces and normals, and just turn the vertex set into a set of nurbs points.
<InPhase>
It might round a few pointy things, but would return generally the shape I think.
<teepee>
CGAL has quite some mesh processing too now, maybe the remeshing would help too
<teepee>
in theory that should consider zero faces :)
<InPhase>
Fundamentally there are going to be parts of broken meshes without correct answers, because there isn't correct input. One just has to choose in what manner the lack of well defined input will be modified into some sort of defined output.
<InPhase>
There does not exist an algorithm for "do what I intended".
ur5us has joined #openscad
ur5us has quit [Remote host closed the connection]
<erickrdini>
the number eight is not like the others!!
<erickrdini>
what i need to do!
<erickrdini>
tks
<teepee>
´yeah, looks like it's slightly off to fully cut through
<erickrdini>
yes
<erickrdini>
i think is something wrong in the code, cant find
<teepee>
it's probably the rotation
<teepee>
one option is maybe making the numbers twice the height
<teepee>
and use linear_extrude(..., center = true);
<teepee>
nope, interestingly that does not help
<teepee>
oh, it does when also using the 17/40 for the number 8 translate
<erickrdini>
when i do that the number goes away
<teepee>
linear_extrude(height = 2*number_height, center = true)
<teepee>
try with twice the number height and the center
<InPhase>
erickrdini: Did you mis above where I explained the 8?
<InPhase>
<InPhase> Also, the 8 as you had it was upside down. If you want it right-side-up, switch that line to: translate([-num_offset, 0, 0]) rotate([0, 90, 180]) number("8");
<teepee>
if you put a # in front of the linear_extrude() for the text, you'll see how that part looks like
<InPhase>
If you're going for Scopeuk's syntax, change only the rotate. Then the -17/40 should work
<InPhase>
You rotated it such that the fat part of the 8 was at the top.
<InPhase>
And in so doing that also offset it differently than the others.
<erickrdini>
trying
<erickrdini>
it works
<erickrdini>
tks
bomb has quit [Remote host closed the connection]
<InPhase>
The x rotation comes first in the 3 axes of rotation, so your first rotation was flipping your extrusion around the x axis, putting it in the negative-z direction. This threw everything else off.
<InPhase>
You failed to notice because it was still shaped sort of like an 8. :)
<InPhase>
But that's why the fat part was facing the other direction.
<Jack21>
erickrdini: if you add ,convexity=5 to the linear_extrude you cube looks better (only preview)
bomb has joined #openscad
Guest79 has joined #openscad
Guest79 has quit [Client Quit]
ur5us has quit [Ping timeout: 260 seconds]
ali1234 has quit [Remote host closed the connection]
<erickrdini>
i got this code and i dont want to show the order side trougth the letter what i have to add?
<dalias>
jack21, you can always write your own prng function explicitly
<erickrdini>
I want the lyrics to come out deep so that it can be read without difficulties but not leaked
<Jack21>
dalias: how can it change with every preview but use the same value when render?
<erickrdini>
guys do u understand me?
<dalias>
i dont know if the builtin random does or can, but it seems like in a functional language it's inherently hard for a prng to be deterministic because the order of expression evaluation is not deterministic
<dalias>
and thus each invocation of the prng function should *have its own seed*
<dalias>
rather than having a seed that's hidden internal state
<Jack21>
erickrdini first you need to add convexity to the linear_extrude(convexity=5);
<Jack21>
else have you rendered you letter cube? did that also looked wrong?
<Jack21>
(you also haven't set a font )
<erickrdini>
humm
<erickrdini>
no i havent
<erickrdini>
i add convexity=5 put can see the other side yet but i little less
<erickrdini>
*a little less
<erickrdini>
than before
<erickrdini>
how can i set the font plz
<Jack21>
then add convexity=10
<Jack21>
font .. go on help ↦ fontlist
<erickrdini>
ok
<erickrdini>
and where i add it plz
<erickrdini>
tks
<erickrdini>
added convexity 16 and it didnt work in the stl i can see the other side
<Jack21>
you select a font you like and click on copy - now either you replace text(font=font⇐ this one) with it or you create the variable font= and place it there
<erickrdini>
tks for that
<Jack21>
ok so you don'- want the letters punch through the whole side - just impress a little?
<Jack21>
it is easier when you work with variables so you only need to change one thing - that is what parametric design means
<erickrdini>
do u think gonna show the words clearly?
<Jack21>
for 3D print you only need 2 or 3 layer so yes 1mm is enough - but if the letters are too small maybe your printer can't print the details
<erickrdini>
OK
<Jack21>
also you could choose a bold font
<erickrdini>
sorry for that ok
<erickrdini>
tks i see you guys around
ali12341 is now known as ali1234
<Jack21>
but i think it will be ok -- you should print this upwards vertically so all sides are equal
<erickrdini>
ANY ADVICE TO DO THAT
<erickrdini>
sorry my caps was on
<Jack21>
a nice touch is to use nail polish over the letter (in the groove) and after drying you sand them so it looks clean and only the vanish in the grooves is left
<Jack21>
you load that into your slicer and turn it .. you also could add rotate([90,0,0]) in front of the first difference
<erickrdini>
ok
<erickrdini>
thats a lot
<erickrdini>
for your support
<erickrdini>
tks*
<InPhase>
erickrdini: Erase $fn=16 in text, and put $fa=1; $fs=0.4; near the top of the scad file.
<InPhase>
erickrdini: Then look at how much prettier your letters get.
<InPhase>
Then capitalize the months! ;)
<Jack21>
InPhase: printing at that size probably wouldn't change much
<erickrdini>
i erased and out it and i got error
<Jack21>
(assuming it is mm not inch)
<Jack21>
erickrdini make sure you erased the ,
<InPhase>
Jack21: Those values are my defaults specifically because they're the limits of what will show up on a standard printer.
<erickrdini>
erase $fn=16 and replace for $fa=1; $fs=0.4
<erickrdini>
?
<InPhase>
erickrdini: Put $fa=1; $fs=0.4; at the top of the file, not in the text. It will apply to everything if you take the $fn out of text.
<InPhase>
You set it once, and then everything is printer-level smooth.
<erickrdini>
i did and dont work
<erickrdini>
brb
erickrdini has quit [Quit: Client closed]
<Jack21>
Inphase: these are sure fine and smooth values but when you print a part of a big curve you end up with .1 while your printer should be able to print better
<InPhase>
Jack21: With a 0.4mm extruder nozzle?
<JakeSays>
that would be one heck of a printer
<InPhase>
Oh, big. You mean the 1-degree?
<Jack21>
the nozzle doesn't matter - it is about the gantry
<JakeSays>
how does the nozzle not matter
<Jack21>
yeah -- as said for 99% it is better than needed
<Jack21>
jakeSays i can print a line with 5mm nozzle or 0.15 nozzle .. this doesn't change how straight the line is
<InPhase>
Jack21: There have been a handful of mechanical snug parts where I've dropped to $fa=0.5; $fs=0.2; and if I'm printing with good temperature, good filament, and good conditions, there's a very subtle improvement in smoothness. But at least on my 2016 era printer there is a positional jitter limit that means nothing is getting smoother than that regardless of what I tell it to do.
<JakeSays>
ah. true that.
<JakeSays>
heh. a 5mm nozzle would consume a lot of plastic
<InPhase>
Someday I want to own a concrete printer. :)
<InPhase>
I don't know what I'll do with it, but it sounds cool.
<InPhase>
Maybe print a walkway.
<JakeSays>
InPhase: lol
<Jack21>
jake says not if you print 0.1mm layer - in the end printing a cube(50) with 5mm walls would consume the same amount like a .5mm nozzle
<JakeSays>
Jack21: ye, i am aware.
<InPhase>
If I get a combo CNC jackhammer, robo shop vac, and 3D concrete printer, maybe I can replace my driveway with some quick programming.
<JakeSays>
InPhase: LOL
<Jack21>
cnc jackhammer - does that exist?
<InPhase>
The dream has to start somewhere.
<Jack21>
Ü
<JakeSays>
my brother is a general contractor. he said that if i could build a printer that would print the framing and exterior walls of a house it'd be revolutionary.
<JakeSays>
i asked him why. he said because he wouldn't have to deal with sub-contractors. lol
<InPhase>
JakeSays: Well there are concrete printers doing that.
<Jack21>
i have seen very nice equipment - with no one there able to use it
<JakeSays>
InPhase: right, but out of concrete, and the houses are weird looking
<InPhase>
But if you want wooden, I've considered that an MDF printer sounds doable. But this is probably not structurally sound.
<JakeSays>
oh i think it would be
<JakeSays>
you can get wood filament
<Jack21>
there are also these brick laying robots
<InPhase>
Woods structural strength requires long fibers. You would need a printer that can handle pumping out the longer fibers.
<buZz>
Jack21: lego bricks? :)
<JakeSays>
InPhase: plastic molecules are pretty long.