<FireBeard>
I just checked out your website, and I love that it has an actual 'webdesign' feel of the site, and I really hope I will not insult anybody with the following statement, but I think I can do better
<FireBeard>
you can tell me to gtfo, that is fine
<teepee>
I'd say get to the point
<FireBeard>
the point is that I'd LOVE to have a crack at making a new design for the site
<FireBeard>
was that not clear? ;)
<teepee>
nothing wrong with that as such. the critical question is how that's going to happen
mmu_man has quit [Ping timeout: 252 seconds]
<teepee>
no, it's not clear if you try to sell a service for example
<FireBeard>
oh no no, NO cost, it's free of course !
<FireBeard>
sorry
<FireBeard>
I would never charge anything for an open source project
<FireBeard>
I just like to make the site more friendly and nice to visit
<FireBeard>
is all
<FireBeard>
just a suggestion, I like your software [though, I have not tried it yet, but have followed a few tutorials, because I thought the way it works might be for me]
<FireBeard>
I can go if you want :)
<teepee>
the website could use an update for sure
<FireBeard>
if you like the current design, but tweaked up a little, I'd be happy to help
<teepee>
there was some updated design a while ago, but that never got to a state that went live
<FireBeard>
it's what attracted me to the site when I saw it: it's not a default squarespace or anyother site standard, but it can use some help :D
<FireBeard>
why not?
<teepee>
not sure, likely just free-time of the people involved at the time
<Murr>
teepee: Yes, I saw the image.tar thing in the logs of the original attempt. Glad this is fixed, and I'm happy to say the arm64 image works great
<teepee>
Murr: nice, I just tried a simple example on a RasPi and that worked fine too
<FireBeard>
I mean, a new design would be VERY easy to implement, if I'd adhere to the current image sizes
<FireBeard>
sorry, English is not my first language :/
<FireBeard>
I just like to help
<FireBeard>
you can oust me any time you want :D
<teepee>
Murr: the next run should be a bit faster, I've increased the parallel processing to 2 :-)
<teepee>
ah, now the effect is gone, I can have 3 columns
<teepee>
the emoji look a bit sad too ;-)
<syserrio>
teepee: I wasn't able to get the separate objects in the 3mf file with lazy-union -- I checked in the file with a text editor. I modified my program and now have the union of the objects at the root of the program as I read was necessary on a reddit post. I tried both manifold/new and CGAL/old rendering. Any tips? (I can always go back to
<syserrio>
colorscad, though)
<Murr>
I see what you mean by the effect, it's like the spinners have a hair trigger, if you don't click very quickly, they jump by 2
<teepee>
syserrio: just no top level union in the file, like: cube(); cylinder();
<teepee>
let me try that, to be sure
<Murr>
The emoji are due to the fact that only preloaded fonts work, and presumably this one isn't included in my image
<Murr>
Same limitation as the thingiverse customizer
<teepee>
but they show up as \u2938 instead of a broken glyph
<Murr>
but all feedback is good. I know this is still in extremely rough shape
<teepee>
so they don't reach openscad processing
<Murr>
Yeah, my parameter extraction messed up and did not process the escape sequences
<FireBeard>
teepee: are you interested at all?
<teepee>
there's an export command for parameter extraction into json
<FireBeard>
if not, I'm out :D
<teepee>
FireBeard: yes, if slow is fine and we can talk what features are set, the design side would be pretty open I think
<FireBeard>
ok, great, and slow is very fine with me !
<syserrio>
teepee: that did it!!! Thanks! Bambu asks me if I want to import the multiple objects.
<FireBeard>
I can show you some of my nonsense next time I'm active, if you want :)
<teepee>
syserrio: yep that's what I get in prusa too
<FireBeard>
because I have to prove myself too, of course !
<teepee>
sounds good to me
<InPhase>
FireBeard: Tip: First step would probably be to propose something with a mock-up and demo, and maybe a plan if there are any significant navigation or content organization changes. Also it's probably meritorious to preserve most of the main landing pages, with external links to them while doing so. And then it would be good to get feedback on that proposal before you deep dive into the whole thing
<InPhase>
without feedback.
<FireBeard>
ok, thanks. I'll probably be back in a day or three or so
<InPhase>
FireBeard: It's a many-stakeholders sort of thing. :)
<FireBeard>
InPhase: I will make a mock-up for sure :)
<teepee>
I guess I can try to extract some of the other design somehow, there's a zip backup somewhere
<FireBeard>
oh no no, no stakeholder whatever, I do it for free, with no ties or anything
<InPhase>
FireBeard: Also, it will be important to be communicative and very open to feedback, because of said many stakeholders.
<FireBeard>
just because I think the software is awesome
<InPhase>
FireBeard: No, I mean stakeholders like everybody involved in the project and interested in it would be impacted by the changes.
<FireBeard>
aha ok :)
<FireBeard>
ok fine :)
<FireBeard>
I just love doing a bit for open and free software
<teepee>
I'd love to see an easier way of posting news articles :-)
<FireBeard>
also, you might just HATE my designs ! :D
<FireBeard>
I hope not though, I think I can make the site a bit more welcoming
<Murr>
Very interesting, that export command. `--export-format` as such is documented, but I have not seen any mention of `--param`.
<teepee>
heh, we just did a design competition for our company logo, I did not really hate any of the suggestions
<FireBeard>
I'm out for now, thank you teepee and InPhase for the positive feedback, I'll be back
<Murr>
And since I have to implement this from scratch in FreeCAD anyway, I might as well keep my implementation for OpenSCAD. It's only a few dozen lines of code
<Murr>
But the official export would make a good regression test for my implementation
<teepee>
in case we ever get the native customizer format going it would be automatic
FireBeard has left #openscad [Bernard, it's hot in the worm.]
<teepee>
but who knows, if and when that happens :)
<teepee>
right, the -h text does not list the allowed formats
syserrio has quit [Ping timeout: 240 seconds]
Artea has quit [Ping timeout: 246 seconds]
J25k73 has joined #openscad
J25k11 has quit [Ping timeout: 240 seconds]
kintel has joined #openscad
<kintel>
The last website update project included a move to Wordpress. The idea was to make it easier for more/other people to participate in website activities as the current git-based approach might not be the easiest for people not familiar with git. However, the fear is that it kind of makes it harder for the existing folks to contribute, plus wordpress comes with its own set of issues and complexities.
<kintel>
Anyway, the topics being tackled back then was 1. general facelift, 2. better mobile support, 3. easier article publishing and 4. exploring opportunities for ad-financing, and optionally 5. bringing the mailing list/forum back to our own server (but no time was spent on that)
ccox has quit [Ping timeout: 252 seconds]
Artea has joined #openscad
kintel has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
pbsds3 has quit [Quit: Ping timeout (120 seconds)]
tja has quit [Read error: Connection reset by peer]
tja_ is now known as tja
mmu_man has joined #openscad
lastrodamo has joined #openscad
pbsds3 has quit [Ping timeout: 245 seconds]
pbsds35 has joined #openscad
ooxoo has quit [Quit: Leaving]
mmu_man has quit [Ping timeout: 244 seconds]
ccox has joined #openscad
RichardPotthoff has quit [Read error: Connection reset by peer]
mmu_man has joined #openscad
mmu_man has quit [Ping timeout: 260 seconds]
ACG has quit [Remote host closed the connection]
mmu_man has joined #openscad
mmu_man has quit [Ping timeout: 252 seconds]
mtm has quit [Ping timeout: 244 seconds]
mtm has joined #openscad
mmu_man has joined #openscad
mmu_man has quit [Ping timeout: 244 seconds]
mmu_man has joined #openscad
mmu_man has quit [Ping timeout: 244 seconds]
Guest56 has joined #openscad
Guest56 has quit [Client Quit]
mmu_man has joined #openscad
mmu_man has quit [Ping timeout: 252 seconds]
misterfish has joined #openscad
<church__>
Is minkowski heavy on plain flat surfaces/edges too? I'm thinking of using it to add clearance between prism "teeths" & prism-shape cuts. So far i tried to avoid minkowski usage as much as possible, but here it would simplify things a lot
<church__>
And minkowski with sphere sounds a bit overkill too, wondering what simpler shape to use for minkowski with
<J25k73>
you can use an octahedron to add a chamfer. Minkowski gets bad when used on concave surfaces - also each point makes it slower.
<church__>
initially i was thinking of tetrahedron .. funny, that sphere with $fn 3 results in prism, not tetrahedron
<InPhase>
church__: If by plain flat, you mean 2D, it works fast and efficient on that. But offset is even better.
<church__>
i was thinking on 3d prism. No concavity, few flat surfaces, made with polyhedron. But of varied slope angles and slightly rotated in two planes (thus using minkowski for even clearance around seems way simpler coding wise then calculating even clearances normal to each surface)
<church__>
LOL, though "sinking" triangular pointy prism into surface with manual approximated values even simpler and certainly not coding intensive
dingodoppelt has quit [Read error: Connection reset by peer]
dingodoppelt has joined #openscad
<church__>
hmm, interesting, same way sphere with $fn=4 doesn't end up in cube either. So i guess simplest way to get tetrahedron or octahedron would be with $fn=3 cylinder with d2=.001 and calculated height of sqrt(2/3)*side.
<feep>
one month I stared at my octree code, too scared to implement non-manifold detecting code cause I thought it'd be way too annoying
<feep>
"then you tried it and it was easy?" no then I realized that the correct answer was "just do not create non-manifold octrees" lol
<feep>
which took like two hours and works depressingly well
mmu_man has joined #openscad
<InPhase>
feep: Avoiding the making of non-manifold stuff does tend to be a much easier solution most of the time. The diversity of chaos that can count as non-manifold can be pretty wild, and a lot of it is just not fixable. :)
<feep>
InPhase: yeah <3
<InPhase>
"Blur your eyes and try to guess what the user meant" is a tricky algorithm to write.
<feep>
tl;dr I was doing "recurse more deeply on cells that contain borders" but that didn't work for cells that only contained one face but where the *edge* of the cell was closer to another face, so the optimizer would move it there. the fix was "recurse more deeply on cells that have more than one face, *or* that have one face and another SDF range overlaps with its SDF."
<feep>
edges*
<feep>
so now it just recurses until the *whole* cell is closest to the face that's in it.
<feep>
which~ fixes~ everything~
<feep>
of course, because I wrote this with AI I now have +90 unpushed commits to review, most of which I don't remember lol
<feep>
but tbh that's still much less threatening.
<InPhase>
church__: Oops, that should be L-2*roundby
<InPhase>
But basic idea. :)
rawgreaze has quit [Ping timeout: 252 seconds]
J25k73 has quit [Quit: Client closed]
J25k73 has joined #openscad
mmu_man has joined #openscad
rawgreaze has joined #openscad
mmu_man has quit [Ping timeout: 244 seconds]
<church__>
InPhase: seems that minkowski on such non-concave, non-round, flat objects, and when one you "drag around" also relatively simple, also very fast, fractions of seconds. So i guess in this case i'll go with minkowski :)
<feep>
I have smooth intersection~
<church__>
ugh. intersections (and differencing with them). got hit by those. luckily found "fix" in way of render(). otherwise they loved to blow csg object tree with crawling resulting performance or hit normalizing object limit
<church__>
everything started to fly right after search & replace all intersection* with render() intersection :)
<feep>
mh. it should be possible to make a d10 by intersecting 10 boxes, right?
<feep>
I'm trying to get sonnet to do it, but it's producing shapes that have *very* little to do with d10 lol
<church__>
InPhase: stupid idea .. but just had a thought - if object i need to make clearance around for is not that "easy to minkowski on" .. but shouldn't simple scale() do the job with less perf. hit?