<Andy-pants>
teepee: I found it! Someone mentioned that it was in BOSL2. However, it looks like revarbat keeps it in a separate repo. https://github.com/revarbat/openscad_docsgen
<Scopeuk>
wnot if that has access to nc is a completely differnt question
<Scopeuk>
s/wnot/now
<teepee>
well, updating the docker images is probably a good idea anyway :)
<Andy-pants>
It's peculiar that they don't define an executor for the IRC orb. If there's no executor defined, does it just fall back to the executor for the current job?
<Scopeuk>
depends on how you layout your config, job steps run in the parent executor
<Andy-pants>
It's not heavily used. So, I assume that most folks who use it just do the workaround of adding netcat to the image in which the job is being executed.
<teepee>
not sure that's worth the effort, if I saw that right it claims 25 usages and it does not support SSL
<Scopeuk>
seams like they should probably have a better solution
<Scopeuk>
particularly as it is linked directly from the main notifications documentation
<Scopeuk>
doesent seam it fixed the old irc message either :(
<Andy-pants>
Their current orb implementation also means that the connection lifetime is only the lifetime of the job. So, it'd mean that it would need to connect, join, message, quit with each notification, right?
<teepee>
yep
<teepee>
well, the channel allows send without join
<Andy-pants>
ohhh, ok. That makes life easier. I was going to suggest a bot which has a persistent connection and listens to webhooks. That would allow the decoupling of the pipeline and IRC, allowing infrastructure independent notification endpoints (IRC, Slack, etc.) and can be modified without needing to modify the pipeline.
<Andy-pants>
But, it involves another piece of instrastructure the project would need to maintain and operate.
<teepee>
maybe that's still the better option for the long term
<teepee>
I need to get a server ordered for a long time anyway :)
<teepee>
this could even be an option for the existing bot
<teepee>
well, the one that's still on vacation :)
<Andy-pants>
Sounds like a perfect job for a Raspberry Pi (provided you can get one!). Or, if you want to be fancy, a serverless implementation. :)
teepee is now known as openscad-ci-bot
openscad-ci-bot is now known as teepee
<teepee>
true, but the other stuff could use a reasonable server so this would just be a tiny additional load on that
<Scopeuk>
gbruno could probably be updated for a webhook from circleci along with the GitHub ones
<teepee>
yeah, that too. I was thinking of poor absent othx though
teepee_ has joined #openscad
kintel has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
teepee has quit [Ping timeout: 255 seconds]
teepee_ is now known as teepee
Mcgiwer has joined #openscad
<Mcgiwer>
Hello. I'm a new user of OpenScad and I have a question:
<Scopeuk>
Mcgiwer welcome. in this forum it is usually best to directly ask the questions and see what happens, we are a fairly friendly bunch
<Mcgiwer>
I wanted to design a rotating cylinder with height of (as a example): 3 cm and width of 5 cm, with a hole in a middle with diameter of 2 cm. Basing on the documentation, I had attempted to create a code for it, but it seem I did something wrong in it. Bellow I list a code part I wrote:
muesli8 is now known as muesli
<Mcgiwer>
h = 2;
<Mcgiwer>
w = 5;
<Mcgiwer>
int = w/2;
<Mcgiwer>
cylinder(h,w,w);
<Mcgiwer>
translate([0,w/2,0]) {
<Mcgiwer>
rotate_extrude(angle=360) {
<Mcgiwer>
cylinder(h, w/2,w/2,center=true);
<Mcgiwer>
}
<Mcgiwer>
}
<Mcgiwer>
as result, I got a warning of ignoring 3d object for 2d operation
<Mcgiwer>
(in the second cylinder I had teporasry replaced the "int" with the "w/2")
<Mcgiwer>
^ temporary
<teepee>
yes, cylinder is already a 3d object, rotate_extrude() expects a 2d shape to be rotated around the Z axis
<InPhase>
You have to turn your head on its side to see what rotate_extrude needs for its input.
<InPhase>
Basically, you take your round target goal, visualize it, and then you slice a sliver of it from the xz plane. That is the 2D shape that you need to make on the xy plane, and feed into rotate_extrude.
<InPhase>
Note that you only slice the xz plane for positive x.
<Mcgiwer>
I had tryed the diference function you have and it had displayed a circle in the middle, but only in the bottom view. In the top one it wasn't visible
<InPhase>
Mcgiwer: Your height was probably not large enough for it to go all the way through.
<InPhase>
Mcgiwer: You must always satisfy the overlap requirement, with things to be subtracted out extending past the surfaces that they are to remove.
<Mcgiwer>
ok, had found a solution - in the second cylinder I had mutiplied the hight thru a value of 2.01 to get pass thru
<Mcgiwer>
ok, second question, how to add hollow's on the side of the external cylinder?
ur5us_ has joined #openscad
<Mcgiwer>
(in the narrow part)
<InPhase>
It's not clear what you mean, but probably more difference removal.
<InPhase>
Maybe even with a for loop if they are distributed in a pattern.
Andy-pants has joined #openscad
<Mcgiwer>
quick afk. I try to make a screenshoot of the result
Mcgiwer has quit [Quit: Ping timeout (120 seconds)]
<teepee>
what's the view on this years GSoC?
Mcgiwer has joined #openscad
<teepee>
I did reactive the tasks, but I guess we would need to sign-up soon-ish
<teepee>
considering the amount of questions so far, I suppose we may not get any applications anyway
<InPhase>
teepee: I think we did better with the longer more full time project aims.
<InPhase>
teepee: I feel like there's a diminished return on the half-sized stuff, because there are fixed costs to get a person going on the right path.
<InPhase>
This is not really unique to GSoC. I see it at work all the time. But it just happens that full time GSoC is right at the threshold where half-time GSoC is the amount of work that normally is not that productive.
<teepee>
good point, let me close one of the 2 remaining short ones
<InPhase>
What is that url again for the project lists?
Mcgiwer has quit [Quit: Ping timeout (120 seconds)]
<InPhase>
You have caught me right as I need to leave, but I can review the list and think about options for it later in the day while you're hopefully sleeping.
<teepee>
well, it was long time restricted to students, it's not anymore
<teepee>
although I've not checked the details for this year
<InPhase>
Our job is to mentor them, google's job is program administrator and salary dispenser, and the students get real project internship experience with that mentorship, and ostensibly they then work fulltime across the summer months between semesters on some defined scope project to contribute to the codebase.
<guso7868>
However i want to spend my time for my primary occupation, suppose this is better than what google would pay to me.
guso7868 has quit [Quit: Client closed]
guso78 has joined #openscad
<teepee>
"be a student or a beginner to open source software development."
<InPhase>
When it goes right, we get some nice leaps forward on the project, requiring only that we put in some mentoring hours for a student and get them on the right code track.
<InPhase>
Right, I guess they dropped the strict student requirement.
<InPhase>
Although students can be a good choice as they are typically not also otherwise occupied over the summer.
<InPhase>
It's pretty important to success that they actually are putting in the specified hours.
Mcgiwer has joined #openscad
<guso78>
teepee, BTW did you eventually peek into the my latest textures PR ? some of the ctests are not pass and i already found the reason for that.
<Mcgiwer>
stupid auto disconnect :-(
<Mcgiwer>
ok, back. InPhase: Here is a screenshoot with drawn arrow to the side where I want to do hallow's: https://ibb.co/GkBD0vH
<teepee>
guso78: only quick glance, the language side might be the difficult part to figure out as we at some point want to have "real" material/color support
<teepee>
having one for preview and one for final model would be bad
<teepee>
but I guess we need to start the discussion somewhere
<guso78>
right now you can specify a picture as a texture and it would show equally as if you would specify it with color gray #808080. of course you can color your texture by specifying a color next to the texture.
<guso78>
i think it should be possible to have same screen for preview and CGA
<guso78>
CGAL
<guso78>
you can specify a texture of bamboo , and it shows exactly bamboo. if you specify baboo and green as color, it shows the same picture more green(exactly by color green - gray)
<Mcgiwer>
does got any idea how to solve the question I had asked ?
<guso78>
Mcgiwer, you want to remove one side of your torus ?
<teepee>
or a hole poking into it in direction of the arrow?
<Mcgiwer>
guso78 no. I want to make hollow's on the side of cylinder marked with the side marked with the arrow
<Mcgiwer>
sorry, let me rephase it
<guso78>
i got it
<guso78>
where the arrow is you want to drill in ?
<Mcgiwer>
yes
<guso78>
you can difference with a cylinder which you place there
<guso78>
you make a cylinder with : cylinder(d=1,h=20, center=true) e.g.
<Mcgiwer>
I know, but I doesn't knoe how to select the side of the cylinder
<guso78>
but of course you have to position it correctly
<guso78>
you can do rotate([0,90,0] ) the_cylinder to rotate it correctly
<guso78>
then translate(x,y,z]) the_result to position the previous object correctly
<guso78>
you can just put %cylinder(d=1,h=20) to see in gray, how the cylidner would lookl like
<guso78>
then use rotate and later translate to position it correctly keeping the % sign
<guso78>
teepee, sometime you will find adequate time to look into the python PR O:3
<guso78>
many features base on that ...
<Mcgiwer>
guso78 on the screenshoot you can see the code I had used to create the cylinder on the snapshoot, but I don't know how to apply the command you gave to it
<guso78>
Mcgiwer, give me few minutes to establish the final code for you
<Mcgiwer>
ok, thanks
<teepee>
guso78: if the crazy work stuff will not go down soon-ish, I'm dead anyway ;-)
<teepee>
which I do plan to prevent
<Mcgiwer>
guso78 the screenshoot is a typo in the first line and a semicollon is missing
<Mcgiwer>
[stupid keyboard issues >.< ]
<guso78>
$fn=20;
<guso78>
h=3.5; // height
<guso78>
w=5; // width
<guso78>
int = w/4; // interhal hole
<guso78>
difference() {
<guso78>
cylinder(h,r=w);
<guso78>
cylinder(2.01*h, r=int, center=true);
<guso78>
translate([0,0,h/2])
<guso78>
rotate([-90,0,0])
<guso78>
cylinder(d=1,h=w+1);
<guso78>
}
<guso78>
teepee, i dont see your crazy work go down before next ten years. the coding style of openscad is still ways easier to me than learning ANY of the gui, where you need to find a turorial on youtube whenerver you want to explore a new feature ....
<teepee>
nah, there's a couple of things in progress already and the huge task backlog in the main project is almost gone too
<guso78>
main project is GF, right ?
<teepee>
BMW
<guso78>
sorry BMW :)
<guso78>
Bayrischer M...
<teepee>
although I do hope to see some live action in Saxony soon :)
<guso78>
BTW all the employees in your comany with higher levels use such a car ...
<teepee>
nope, my company car is a Prius
<guso78>
traitor :)
<teepee>
I have that since maybe 6 years before starting some BMW stuff :P
<teepee>
but we do still have an i3
<guso78>
very good excuse :)
<teepee>
but it's old and would not even reach half way to munich
<guso78>
it depends where you depart from. if you refer to tokyo , you might be true
<guso78>
BTW today i found out that the ctests and offscreen renderer use the vertex-buffer-object render experiemantal feature in openscad, whereas the GUI does not.
<guso78>
and most of the tests would fail without this experiemental feature .. but its clean-as-is ...
guso78 has quit [Quit: Client closed]
guso78 has joined #openscad
<teepee>
with kintel working on the GL stuff, there's likely/hopefully some modernizing changes coming