<kintel>
Does anyone know the oldest version of Python we need to support (for unit test purposes)? We're currently on python 3.4
L29Ah has joined #openscad
<InPhase>
That's 11 years ago. Not worth thinking any further back than that.
<InPhase>
kintel: Hmm. Probably our other .py files in tests should have the #!/usr/bin/env python3 shebang line.
<InPhase>
Then they would all use the same python.
<InPhase>
Hard-coding that to #!/usr/bin/python is a mistake.
<InPhase>
test_pretty_print.py and test_cmdline_tool.py have it wrong.
<kintel>
Was hoping for dataclasses, but I think I found a workaround
<kintel>
(python 3.7)
<kintel>
yeah, the shebang could be updated. I'll do that while I update test scripts
<InPhase>
Most of our scripts look like they should be extremely version agnostic and pretty robust against Python version changes within 3.x. The one I have the most concern about is mingw_convert_ctest.py. That pulls in some stuff I'm not familiar with that I don't have confidence in it staying consistent.
<InPhase>
Specifically that from winreg import *, and import mingw_cross_info.
<kintel>
heh, not gonna touch *win* ;)
<kintel>
^ the glob that is
<InPhase>
And "from winreg import *" is just asking for trouble on name collisions. :)
<kintel>
Context btw.: I want to switch to use manifold as the default backend, but I want more test coverage first. Specifically I want it to be trivial to write tests that run on both cgal and manifold without managing more tests, by allowing tests to run multiple variants with the same expected output
<kintel>
..and if I can find a good path forward, I could employ the same to run both preview and render tests together
J25k59 has quit [Quit: Client closed]
J25k59 has joined #openscad
kintel has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
kintel has joined #openscad
<teepee>
kintel: python 3.9 should be safe, that's debian 11, ubuntu 22.04 has 3.10
califax has quit [Remote host closed the connection]
califax has joined #openscad
nomike has joined #openscad
nomike has quit [Ping timeout: 260 seconds]
nomike has joined #openscad
nomike has quit [Remote host closed the connection]
guso78k has quit [Quit: Client closed]
stealth_ has quit [Quit: Leaving]
guso78k has joined #openscad
<guso78k>
teepee, this is what I meant with interactive editing : https://www.youtube.com/watch?v=9xBtzE5hhR4 . you mark values in the code which are enabled for dragging, like "180".
<othx>
guso78k linked to YouTube video "PythonSCAD dragging objects interactively" => 1 IRC mentions
<guso78k>
Its all about glUnproject, a lots of matrices and a small brother of GeomtryEvaluator.cc
mmu_man has joined #openscad
juri_ has joined #openscad
nomike has joined #openscad
buZz has quit [Ping timeout: 252 seconds]
Guest49 has joined #openscad
Guest49 has quit [Client Quit]
snaked has quit [Remote host closed the connection]
buZz has joined #openscad
buZz is now known as Guest8613
L29Ah has left #openscad [#openscad]
L29Ah has joined #openscad
L29Ah has left #openscad [#openscad]
L29Ah has joined #openscad
guso78k has quit [Quit: Client closed]
L29Ah has left #openscad [#openscad]
L29Ah has joined #openscad
J25k98 has quit [Quit: Client closed]
J25k98 has joined #openscad
guso78k has joined #openscad
guso78k has quit [Client Quit]
Knochi has joined #openscad
<Knochi>
Hi,
<Knochi>
has anyone managed to export a multicolor 3mf and make it work with bambustudio?
<teepee>
if you expect the colors to show up in bambu, that's impossible
<teepee>
with experimental feature "lazy-union" enabled you currently can export multi-object files that allow extruder selection in bambustudio
J25k98 has quit [Quit: Client closed]
J25k98 has joined #openscad
Guest8613 is now known as buZz
<Knochi>
It's technically not impossible. When you upload the scad to makerworld you can generate compatible 3mf. And you can export colored 3mf with the dev snapshot version. But when i do locally they open as single color
<teepee>
it is impossible as openscad uses lib3mf which only supports spec compliant files. coloring in most slicers is *not* spec compliant
<teepee>
so yes, we *could* write our own 3mf stuff, but that's pretty unlikely to happen
<Knochi>
no problem, i'll upload to makerworld as private and use their customizer ;-)
<teepee>
the only reasonable solution would be some spec compliant handling everyone can agree on
<teepee>
you don't need colors in slicer for multi-color printing
<Knochi>
yes it's a pitty that the don't tie to the specs
<Knochi>
What I do usually is export each color as single stl, import them all and tick the "handle as one object" box, this way the stay together. But need to accomendate for that in the project
<teepee>
i commented some time ago on a bambustudio issue discussing that, but nobody cared as it was already closed (with wrong reason in my opinion)
<Knochi>
ah have seen this issue :-D
<teepee>
lazy union gives that in a single 3mf-file
<Knochi>
that's totally sufficient. Now I understand! You are right, I don't need the actual colors, just the objects.
<teepee>
the only extra benefit of the colors might be auto-assigning extruders
Knochi has quit [Ping timeout: 240 seconds]
J25k98 has quit [Quit: Client closed]
J25k98 has joined #openscad
<InPhase>
teepee: Do you happen to know if the slicers are using a mutually similar alternative to the 3mf standard for colors? Or are they each rolling their own?
<teepee>
bambu and prusa use very similar but not identical ways (same place, different xml tags)
<teepee>
I have not looked at others
<InPhase>
Is there a reason they can't use the normal 3mf format for expressing this? Is it inadequate?
<InPhase>
I haven't gotten into diving into multicolor, but it is on my ambition list coming up. I want to obtain a multicolor printer, and then I would like to actually use it. :)
<teepee>
the prusa one might be an unofficial extension so it might be even spec compliant (e.g. using slic3r namespace)
<teepee>
bambu just adds extra xml tags which is technically allowed by the xml schema but the lib3mf spec forbids using that just in random way
<teepee>
but I don't know an official way of reading custom extensions via lib3mf
<hramrach>
speaking of colors it seems that some STLs are colored but openscad makes only colorless
Adel has joined #openscad
<teepee>
not possible, STL has no place to store colors
<teepee>
supposedly there are "extensions" but I'd not call that STL anymore
<hramrach>
argh, it's a different format, indeed
<hramrach>
too many formats
<hramrach>
yes, it would lilely be STP, not STL
<hramrach>
tried exporting wrl instead, and it's somehow broken, it displays different garbage in different applications
kintel has joined #openscad
<kintel>
Our WRL export isn't well tested (or automatically tested), so it's likely to bitrot. Ideally we'd add WRL import and get some automated testing going.
<guso78k>
@hramach, my implementation can import and export step, but due to OpenSCAD's nature all roundings will be immediately converted to approximatins https://imgur.com/a/e3ttkvO
<othx>
J25k98 linked to YouTube video "BambuStudio Colored OBJ Import" => 1 IRC mentions
sparsh-n has joined #openscad
sparsh-n has quit [Client Quit]
Adel has joined #openscad
stealth_ has joined #openscad
Guest57 has joined #openscad
Adel has quit [Quit: Client closed]
Guest57 has quit [Client Quit]
<kintel>
J25k98 Is that just a regular OBJ file with surface materials? If this actually works, that would be great motivation to update our OBJ exporter!
mmu_man has quit [Ping timeout: 268 seconds]
TheMaker has joined #openscad
mmu_man has joined #openscad
<hramrach>
there is some python library for generating colourful step files, that also works
<teepee>
kintel heh, I did draft a slighly stronger worded version ;-) Happy to throw that away.
<teepee>
I suppose we should make a standard reply at some point. it's like the 4th or 5th of those inquiries?
<kintel>
heh, standard reply would be great. I couldn't be bothered to think this time :)
dingodoppelt has quit [Remote host closed the connection]
<kintel>
Ideally, we encourage them to publish their analysis so that we can just share that
dingodoppelt has joined #openscad
<kintel>
..but last time we tried that, they went quiet. I guess someone earns good wages for having their job done for them ..
<teepee>
yeah, that and not even looking before what the application actually is
<teepee>
which is disturbingly lazy
<InPhase>
teepee: That was a very abrupt blow-off by the bambu lab dev in that issue.
<teepee>
I can't imagine any good results whatever we would answer with that preparation
<kintel>
it's cover-your-ass security
<teepee>
InPhase: yes, and I think it's also wrong
<teepee>
but I was a bit late to the party
<kintel>
J25k98 I believe Olivier's export is some highly experimental stuff. It might make it into OpenSCAD eventually, but I guess it needs significant love.
<teepee>
not the 3mf part, AFAIK it uses the bambu stuff generating the xml manually
<teepee>
hmm, ok. but good question for the slicer githubs :)
<teepee>
in regard to APIs 2020 would be pretty new I suppose
<J25k98>
prusa is member of the 3mfConsortium or
<teepee>
yes, I belive so
<InPhase>
I didn't deep dive, but glanced through the pdf. Labeling the VOLUME rather than the triangles with color is much much better.
<InPhase>
We don't print triangles.
<teepee>
but how does it match to the triangles?
<teepee>
we do have colored triangles
<InPhase>
<mesh volumeid="3"> <vertices> ... </vertices> ... </mesh> is in one of the examples
<InPhase>
pdf page 11
<InPhase>
The volume then has preceding markers of its properties by the id.
pa has quit [Ping timeout: 252 seconds]
pah has joined #openscad
pmn has left #openscad [WeeChat 4.7.0-dev]
<teepee>
uff "I.T Security Analyst" not understanding the concept of a locally installed application?
<teepee>
if that's the "IT for education", I don't have much hope for mankind anymore
howiemnt has quit [Remote host closed the connection]
howiemnt has joined #openscad
<InPhase>
teepee: I could sure tell stories in an unlogged context. lol
<teepee>
no need, mood is in the wrong place already ;-)
<InPhase>
:)
<teepee>
I mean that's not even coming from US
<kintel>
People keep saying ChatGPT will steal a lot of jobs, I'm becoming less sceptical :/
<teepee>
I can just imagine the mails we might get in a couple of month from there
<teepee>
indeed, those mails might have even made more sense coming from some LLM
<kintel>
We just need a chatbot at the other end of that email address ;)
<teepee>
haha, yes, let the bots decide security
<teepee>
we should make a signal account for that :D
<teepee>
it would be cool to hear from the teacher though
L29Ah has quit [Read error: Connection timed out]
<kintel>
I think this is just some policy, and people use it to avoid using their brains
<teepee>
yeah, but the teacher request is likely real
<kintel>
yeah definitely
guso78k has joined #openscad
<teepee>
it would be cool to have some "for teachers" section on the website
<kintel>
To be fair, we should probably have a privacy policy, and put it on the website, just to help people understand it.
howiemnt has quit [Read error: Connection reset by peer]
<teepee>
how does that work worldwide
<teepee>
we even have different ones for germany and spain on our website and that's all EU / DSVGO
howiemnt has joined #openscad
<kintel>
It would be our minimal policy: Our policy is to not store and sell user data. Feel free to translate that into the legal framework of your country :)
<teepee>
I guess some general statement might work but it's probably impossible to have something legally binding
<kintel>
A "policy" is simply something we strive to follow, not really legally binding.
<kintel>
But yeah, it needs a little bit of thinking
<teepee>
true, complying with DSVGO is probably not a bad start anyway
<teepee>
we don't have google on the site and the logs only store IPs in hashed format
<kintel>
Just not the kind of work I tend to prefer in my spare evenings ;)
<teepee>
also no cookies, so I think we are green even in EU terms
<guso78k>
teepee. your stored IP's until now tell you exact openscad usage statistics ;)
<teepee>
the new theme selector is local storage which I believe is fine as it's on the user side
<teepee>
no they don't
<teepee>
also again, we *don't* store real IPs *at all*
<teepee>
they are hashed already in the reverse proxy sitting in front of everything we own
<teepee>
that said, I don't know about the github stored site
<teepee>
something to check at some point
<guso78k>
ahh, ok, I was concluding from my simple web server to everybody else;s
<teepee>
but then, that's not related to the app
<guso78k>
but I am happy, that openscad finally merged the PR with the print services checkbox with off default. thx teepee!
<teepee>
well, I do own part of an IT company in DSVGO country, so we have to have someone taking care of tha
<teepee>
meaning, I'm pretty aware that stuff exists as we pay for the consulting
<teepee>
and in general I think it's a good thing, even if it's not perfect.
<teepee>
but just exposing those sites that "we care about your privacy, click here to transfer all your data to our 638 partners"
<teepee>
that's already a very good result
L29Ah has joined #openscad
lostapathy has quit [Quit: WeeChat 4.1.1]
lostapathy has joined #openscad
SamantazFox has quit [Quit: Bye]
manu78 has joined #openscad
<manu78>
Hi can anyone help me to create a small tool ?
manu78 has quit [Client Quit]
SamantazFox has joined #openscad
howiemnt has quit [Remote host closed the connection]
howiemnt has joined #openscad
peepsalot has quit [Remote host closed the connection]
peepsalot has joined #openscad
TheMaker has quit [Quit: Konversation terminated!]
kintel has quit [Quit: My Mac has gone to sleep. ZZZzzz…]