<gbruno>
[github] kintel reopened issue #5982 (Documentation fails to warn that surface() ignores alpha channel on grayscale PNG with alpha, producing unexpected result.) https://github.com/openscad/openscad/issues/5982
snaked has joined #openscad
sculptor_ has quit [Changing host]
sculptor_ has joined #openscad
hyperair has joined #openscad
hyperair has quit [Ping timeout: 252 seconds]
<gbruno>
[github] kintel closed issue #5982 (Documentation fails to warn that surface() ignores alpha channel on grayscale PNG with alpha, producing unexpected result.) https://github.com/openscad/openscad/issues/5982
hyperair has joined #openscad
bitbasher has quit [Ping timeout: 276 seconds]
foul_owl has joined #openscad
foul_owl has quit [Client Quit]
hyperair has quit [Ping timeout: 252 seconds]
sculptor_ has quit [Ping timeout: 260 seconds]
teepee has quit [Remote host closed the connection]
teepee has joined #openscad
_whitelogger has joined #openscad
hyperair has joined #openscad
mmu_man has joined #openscad
teepee_ has joined #openscad
teepee has quit [Ping timeout: 244 seconds]
teepee_ is now known as teepee
sculptor_ has joined #openscad
sculptor_ has quit [Ping timeout: 272 seconds]
zauberfisch has quit [Read error: Connection reset by peer]
zauberfisch has joined #openscad
TheAssassin has quit [Ping timeout: 244 seconds]
TheAssassin has joined #openscad
ccox_ has quit [Ping timeout: 276 seconds]
pa has quit [Ping timeout: 252 seconds]
mmu_man has quit [Ping timeout: 272 seconds]
ccox has joined #openscad
pah_ has joined #openscad
pah_ is now known as pa
pa has quit [Changing host]
pa has joined #openscad
bitbasher has joined #openscad
mmu_man has joined #openscad
J25k41 has quit [Quit: Client closed]
J25k41 has joined #openscad
bitbasher has quit [Ping timeout: 276 seconds]
J25k41 has quit [Quit: Client closed]
J25k41 has joined #openscad
bitbasher has joined #openscad
J25k41 has quit [Quit: Client closed]
J25k41 has joined #openscad
howiemnt has quit [Ping timeout: 248 seconds]
sculptor has joined #openscad
J25k41 has quit [Quit: Client closed]
J25k41 has joined #openscad
sculptor_ has joined #openscad
mmu_man has quit [Ping timeout: 276 seconds]
sculptor_ has quit [Changing host]
sculptor_ has joined #openscad
sculptor has quit [Ping timeout: 260 seconds]
howiemnt has joined #openscad
sculptor_ is now known as Sculptor
sculptor_ has joined #openscad
sculptor_ has quit [Changing host]
sculptor_ has joined #openscad
Sculptor has quit [Ping timeout: 252 seconds]
sculptor_ is now known as Sculptor
J25k41 has quit [Quit: Client closed]
J25k41 has joined #openscad
mmu_man has joined #openscad
mmu_man has quit [Ping timeout: 245 seconds]
DedoVerde has joined #openscad
<DedoVerde>
Hi there! Cool to see there's an IRC channel for OpenSCAD. Classic all the way
kintel has joined #openscad
<kintel>
DedoVerde This where all the action is. Usually..
<DedoVerde>
Heh, well, IRC has always been a bit calmer and I'm fine with that
<Sculptor>
kintel, you mean, mostly github bot posting updages
<Sculptor>
updates
<DedoVerde>
lol ah yes, sounds fun
<DedoVerde>
so would this be considered the right place to ask stupid questions about render speed between 2021/nightly builds? n
<DedoVerde>
because I'm confused :P
mmu_man has joined #openscad
<kintel>
yup, in general, just ask, as the people answering may be in a different timezone
<Sculptor>
it's the right place to ask anything oscad related
<kintel>
(the channel is logged, so people read the log)
<DedoVerde>
will do, thanks! yeah i read that in the topic, I feel watched now, but I'll try to play it cool
<Sculptor>
ohh, why nobody said that to me when i joined 3 weeks ago
<Sculptor>
i'd talk less nonsense
<DedoVerde>
xD well but it's IRC tho
<Sculptor>
DedoVerde, i bought a 3d printer 34days ago, and found out about openscad 27 days ago
<Sculptor>
so far i was able to design anything i needed for 3d print in oscad
<Sculptor>
chatgpt can help as well with writing the code
<DedoVerde>
I'm extremely happy with OpenSCAD myself, it's one of the best tools I never new existed
<DedoVerde>
I was looking at various 3d/CAD software, thinking my cheapest realistic option would still cost me at least 500eu a year, be forced cloud garbage, dubious copyright terms, and require months of learning CAD. instead i just started coding and making models ;) it's awesome
<DedoVerde>
But, in short - if I have two instances of 2021 open, one with a lib, one with a model-specific .scad file calling the lib with certain params, and edit the lib, to change some diameter somewhere - the changes are INSTANT in 2021, in the model view in the other instance. In the Nightly, it takes about 10 seconds for them to show up. And I did select the "new/fast" renderer... However, the
<DedoVerde>
initial render (when I load the model.scad) is 10s vs 18s in 2021. Am I doing it wrong? If so, what is it? :D
<DedoVerde>
thanks if anyone has any idea or hints :D i also thought (dreamed?) there was a setting like "render on save" in a previous nightly (???), that's now not there anymore, if it ever was - maybe that could have helped? it seems it's a "kinda sorta full render" vs. "preview render" thing
<Sculptor>
DedoVerde, there's also a nice online viewer/editor/renderer that packs the whole code in the url - can use it for sharing designs with short url services https://tinyurl.com/marko-mac
<DedoVerde>
hey, that
<DedoVerde>
's cool, reminds me a bit of jsfiddle and the like :D
<Sculptor>
can tweak your designs on a phone
<DedoVerde>
could be really useful to share code if required to replicate my "issue", if it even is one... maybe I just lucked into an unexpected feature on 2021 there, IIRC it didn't quite behave the same way if both .scad files were opened in the same instance; i think then rendering took longer too... would be fun if i was just accidentally abusing a bug, but it's really really fast
teepee has quit [Remote host closed the connection]
teepee has joined #openscad
<DedoVerde>
oh yeah and if it makes any difference, I'm still running an older "high end" build, 5950x, 128GB DDR4, 6900XTXH, PCIE-NVME. so the horse power should be there more or less maybe hopefully, unless openscad prefers cuda :p
J25k41 has quit [Quit: Client closed]
J25k41 has joined #openscad
<Sculptor>
you mean business
<Sculptor>
wonder what's the newer one liek
<Sculptor>
DedoVerde, from my newb experience with oscad, good idea is to tweak $fa $fs $fn and boost them once you are happy with the design, prior rendering
Guest17 has joined #openscad
Guest17 has quit [Client Quit]
mmu_man has quit [Ping timeout: 276 seconds]
J25k41 has quit [Quit: Client closed]
J25k41 has joined #openscad
mmu_man has joined #openscad
J25k41 has quit [Quit: Client closed]
J25k41 has joined #openscad
<DedoVerde>
@Sculptor sorry for the late reply, was just idling and calling me old folks ;) I've optimized $fn as far as I could in terms of what my geometry allows, but I've not yet looked at the other two - docs open now, thanks!
<Sculptor>
yeah, but i mean you can use a switch
<DedoVerde>
oh, you mean to differentiate between preview and final? that sounds like a smart idea
<Sculptor>
yep
<Sculptor>
speeds up rendering
<Sculptor>
during the design phase
<DedoVerde>
have to look at that one too... in the 2021 version it really was instant though, but maybe it could help with the newer build... I was so surprised when I saw then it was so much slower, it just seemed like I had to be missing something
<DedoVerde>
actually that reminds me, I'll just update my GPU drivers as well, mine are from march...
<teepee>
DedoVerde: if you run the nightly-version, make sure to go to preferences and select Backend=Manifold
<DedoVerde>
not sure if openscad actually renders anything on (amd) gpus but hey, can't hurt, I hope
<teepee>
GPU does not make a huge difference, it's not helping for the meshing part
<DedoVerde>
ah no that's the "new/fast" renderer, I already had that checked; actually, I did not try the "old/slow" one on 2025 yet, curious if that changes anything, brb...
<DedoVerde>
(CGAL vs Manifold is what I mean, sorry for not being precise)
<DedoVerde>
hm
<DedoVerde>
so short answer no, not really, cgal vs manifold in 2025 does not make a difference regarding the (extreme, 10-20x) difference in rendering time when changing lib.scad in 2025-instance1 and looking at the preview of model.scad in 2025-instance2
<DedoVerde>
however, cgal rendering in 2025 took 16 seconds for the initial render, and 18 in 2021, so yay, it's faster :D doesn't help me necessarily but still
<DedoVerde>
as a matter of fact, 2025-instance2 (the one with model.scad) freezes entirely (!) until the rendering is finished, that's caused by editing lib.scad in 2025-instance1
<DedoVerde>
same exact files in 2021, also two instances, one for model.scad (mostly just to see the preview - though still showing all sub-windows/MDI-children or whatever they are) and one for lib.scad, changing the same value in lib.scad, takes about 0.5 seconds, if that, to show up in the preview
<DedoVerde>
so the rendering thread blocks the gui thread entirely (if they are separate?), and takes forever, for some reason, might be the shortest description i could come up with. again, this is not meant to talk negatively about openscad, i love the hell out of it, just something seems off there when directly comparing 2021 vs nightly
<DedoVerde>
@Sculptor thanks, looking into that too! @teepee i see, oh well, thx!
mmu_man has joined #openscad
<DedoVerde>
ok, now i gotta know what happens if i edit the lib in 2025 and render/preview the model in 2021, assuming it'll be instant as well...
<DedoVerde>
oh hey and don't update graphics drivers while 2025 is open or you'll get a white, uncloseable window xD
<DedoVerde>
that's on me
<DedoVerde>
[21:13:10] <DedoVerde> ok, now i gotta know what happens if i edit the lib in 2025 and render/preview the model in 2021, assuming it'll be instant as well... << yes, it is
<DedoVerde>
it seems kind of unholy but maybe i'll just do that for now, use the 2025 editor (with moar darker backgrounds) and everything for the lib, along with the 2021 renderer for the actual product model...
<DedoVerde>
takes 2 more seconds on initial startup/model load but 1/20 for each revision, so...
<DedoVerde>
oh and if anyone out there has a plugin for visual studio, as in the ACTUAL visual studio, not the crappy free version, to work with openscad... my love and reverence would be eternally yours, in this life and the next
Sculptor has quit [Ping timeout: 260 seconds]
<DedoVerde>
or maybe I'll just install vs code as well, but if 2025 made that unnecessary, i certainly wouldn't complain either ;) if it makes sense I'll gladly file a bug report as well, though i'm not on any of the githubs/labs/whathaveyous yet... i only do proprietary usually. thanks anyway if anyone in a position to "fix" it (if anything's broken) reads this :D
<DedoVerde>
https://opencollective.com/openscad/donate << glad to see that that exists as well, will certainly be buying you guys a coffee or two once I make my first sales... and let me know if you ever switch to c#, i'd love to contribute. but i just hate c++, java, javascript etc. with a passion. c# is king of kings. and it has background workers so your gui threads don't block :P
<kintel>
DedoVerde When you say "render" are you referring to Preview (F5 or automatic if selected) or Render (F6) ?
<gbruno>
[github] Sparsh-N opened pull request #5984 ([Draft] Add search path and initial changes for MSVC based headless build support.) https://github.com/openscad/openscad/pull/5984
<DedoVerde>
yeah i figured i was too unclear on that as well, I'm afraid I'm not really familiar with the correct terminology yet, when it's "meshing", when it's "previewing" and when it's "rendering and all that"... to me, every time code becomes a 3d visualization, that's "rendering" - so by "first load rendering" i mean when I double click a .scad file, and by "rendering changes" i mean when I change
<DedoVerde>
a diameter from 1.5 to 2.5 and I can visually see the change in the model preview on my desktop
<DedoVerde>
br
<DedoVerde>
brb
J25k41 has quit [Quit: Client closed]
J25k41 has joined #openscad
<DedoVerde>
ah, i take it that gbruno is a bot, and those were the git changes alluded to earlier? :D @kintel so to try and be clear, I found for me, working with 2021, the fastest way was to have a lib.scad and a model.scad, whereas the lib was "constant" and the model was "variable"; if i then edited those two files in two separate instances of openscad2021, and changed something in lib.scad, the
<DedoVerde>
change was IMMEDIATELY apparent in the other instance, hosting model.scad; however, doing the exact same thing, the exact same way, with the exact same files, but in 2025, the changes take about 8-12 seconds to appear
<kintel>
In OpenSCAD, Preview (F5) is what we do to display your design on the screen without actually evaluating all the geometry. This is also automatic on save or when you edit dependent files if Design->Automatic Reload and Preview is selected
<DedoVerde>
oh yeah and by "making changes" I just mean changing a value and pressing ctrl+s, so "save", not an f5/f6 "real" render
<kintel>
There is one known regression from 2021 to 2025: The time to first frame on preview may take longer, but the frame rate while moving the object around should be a lot faster
<DedoVerde>
so if i have two instances, change something in instance1 (with the library), press "save" or ctrl+s, the change shows up in a fraction of a second in a second, completely different instance, hosting model.scad
<kintel>
Do you have a lot of triangles in your model?
<kintel>
..or can you share it?
<DedoVerde>
let me try to answer :) i unfortunately can not share the model though, i'd love to, but it's "very proprietary"... i could try making a completely non-intention-divulging fake model with the same issues though...
<kintel>
Also, by "two instances" do you mean two different processes, or just two different windows? It matters a bit since caching only works within the same process.
<DedoVerde>
first of all, thank you so much for the interest... second, let me get back to you asap with answers
<DedoVerde>
[22:09:37] <kintel> There is one known regression from 2021 to 2025: The time to first frame on preview may take longer, but the frame rate while moving the object around should be a lot faster << i could not really answer that, in that i have no idea when the FIRST frame appears, but I can tell that CGAL rendering from 2021 to 2025 for the same model (in my case) became faster from 18 to
<DedoVerde>
16 seconds; doesn't seem related
<DedoVerde>
[22:08:22] <kintel> In OpenSCAD, Preview (F5) is what we do to display your design on the screen without actually evaluating all the geometry. This is also automatic on save or when you edit dependent files if Design->Automatic Reload and Preview is selected << i just do "ctrl+s" (or strg+s since I'm german :P), so doesn't seem related
<kintel>
Ctrl-S actually triggers F5 previews
<kintel>
CGAL rendering may be marginally faster, but that seems unrelated. Manifold rendering should be very much faster.
<DedoVerde>
[22:09:49] <kintel> Do you have a lot of triangles in your model? << Geometries in cache: 285 Geometry cache size in bytes: 11554752 CGAL Polyhedrons in cache: 0 CGAL cache size in bytes: 0 Compiling design (CSG Products normalization)... Normalized tree has 274 elements! Compile and preview finished. Total rendering time: 0:00:00.325 << does that help answer that? sorry, no idea how to view
<DedoVerde>
triangle count?
<DedoVerde>
aaaaah ok good to know, thanks!
<kintel>
Do F6:
<kintel>
Total rendering time: 0:00:00.075
<kintel>
Top level object is a 3D object (manifold):
<kintel>
Vertices: 79
<kintel>
Facets: 154
<DedoVerde>
lol ok that makes sense... sorry
<DedoVerde>
might take ~10 minutes though from my experience
<DedoVerde>
with this model
<kintel>
Even with Manifold?
<DedoVerde>
that's on 2021 with CGAL
<kintel>
Manifold should be 100x to 1000x faster
<DedoVerde>
actually i never finished an f6 render on 2025 yet... let me try right after
<DedoVerde>
lol ok just a sec then, let me cancel
<kintel>
..but that depends on your design, as there is one slow path left: minkowski sums of concave shapes
<DedoVerde>
yes, i... yes :D
<DedoVerde>
i do those a lot
<kintel>
heh
<kintel>
Also: 11554752 CGAL Polyhedrons in cache
<DedoVerde>
...sounds bad?
<kintel>
That's 11 million cached objects. This is very much corner case territory, and I'm surprised your computer isn't on fire!
<DedoVerde>
lol
<DedoVerde>
D
<DedoVerde>
xD
<DedoVerde>
yeah well thats just how i do you know
<kintel>
Also, it's an indication that you haven't enabled Manifold,
<DedoVerde>
i wish i could share the design, i swear it makes sense :D
<DedoVerde>
where would one do such a thing? i have a bunch of unions everywhere, which from what i read in the docs are completely useless, so... ah yes, ok, you mean the "manifold renderer" (?), i do indeed have that still disabled in 2025, true... earlier tests; let me revert and test
<DedoVerde>
so the (wrongly named by me) initial render time, as in when i double click the model.scad, goes down from 16 to 10 when moving from cgal to manifold
<kintel>
It sounds like you're generating a large number of tiny objects which are, somehow, different, then combining them in a way that involves massive concave objects and then perform manifold operations on them
<kintel>
That includes multiple known classes of problems, and I can see how that's interesting. I hope it involves protein folding :)
<DedoVerde>
well, i have a bunch of objects, not really that tiny though - cylinders, usually (more or less roughly) 10cm long, together forming a shape
<DedoVerde>
hahaha it sure does XDDD
<DedoVerde>
let me give you some data while i let it render
<DedoVerde>
ok, f6 render on 2025 with manifold runnin now; getting 8 billion errors, which is good i guess. ah, render finished, lol
<DedoVerde>
yeah, it really is 100x faster
<kintel>
Btw., once it renders (after 20 secs or so) in 2025, are you able to move the object around on the screen with some reasoable performance?
<DedoVerde>
except for that stupid preview when changing stuff of course. damn this thing is fast (mostly) now
<DedoVerde>
Total rendering time: 0:00:12.400 Top level object is a 3D object (manifold): Status: NoError Genus: 533 Vertices: 232081 Facets: 466290 Bounding box: Min: -72.25, -72.25, 0.00 Max: 72.25, 72.25, 94.00 Size: 144.50, 144.50, 94.00 Camera: Translation: -19.95, 30.16, 51.08 Rotation: 34.70, 0.00, 22.20 Distance: 679.97 FOV: 22.50
<DedoVerde>
Rendering finished.
<kintel>
Design->Automatic Reload and Preview: If you uncheck that, it will turn off the render-on-save
<kintel>
..and you can just press F5 instead
<DedoVerde>
aaaaaaaaaaaaaaaaaaah that's that option i had seen but somehow was not in the preferences anymore - because it never was, lol
<kintel>
ok, 400K triangles in 12 secs. Sounds reasonable
<DedoVerde>
so facets are triangles. i'm learning
<kintel>
I think what happens is: You have a crazy large design, but you get rid of a lot of detail along the way so that the final mesh is relatively small (400K triangles)
<kintel>
However...
<DedoVerde>
sorry, need more whiskey, since clearly problem solving is much faster with it. just a sec...
<kintel>
When previewing, we don't know that, because we don't actually evaluate the geometry, so we attempt to give your a faster preview
<kintel>
..but it involves rendering all your source objects individually, and before doing that, we create what's called Vertex Buffer Objects to upload to the GPU
<kintel>
This process can be pretty expensive for the first frame
<kintel>
..but: We still need to perform the minkowski sum, as we cannot yet preview minkowski sums
<kintel>
..which leads me back to:
<kintel>
Also, by "two instances" do you mean two different processes, or just two different windows? It matters a bit since caching only works within the same process.
<DedoVerde>
this is all extremely helpful and an awesome read, but between that, the whiskey and multiple instances of 2021 and 2025 running, i need a sec to straighten everything out before i descend into relaying garbage info :D thanks a ton once again in the meantime
<DedoVerde>
[22:29:24] <kintel> Also, by "two instances" do you mean two different processes, or just two different windows? It matters a bit since caching only works within the same process. << just that one really quickly: i mean whatever the .exe does when you double click two different files, i'd assume it launches a process for each but maybe you have a logic implemented (i once had to do that
<DedoVerde>
too) that catches it beforehand and makes it all handled by a single process in multiple windows, so in short: no idea. me double click two files. me see two windows. ugh ugh.
<DedoVerde>
task manager says...
<DedoVerde>
well shit, no idea, since i have multiple 2025s and 2021s open, let me get back to you xD but it does look like one instance per file
<DedoVerde>
aka processes
<kintel>
In any case, there is caching - any geometry or sub-geometry which didn't actually change, will be cached
<DedoVerde>
which would explain why it's so damn fast
<kintel>
If your minkowski sum(s) happens to be cached, you will save a lot of time for subsequent edits
<DedoVerde>
it doesn't seem to change anything but the specific cylinder diameter i changed - in 2021; in 2025 it renders everything
<kintel>
yeah, so that's curious. ..and might be a new bug/regression. But there are too many moving parts right now. I think the best bet would be some sort of example highlighting the issue
<DedoVerde>
yeah no, that's where I'm out; when i google minkowski sum, i just get a wikipedia entry with a bunch of math; aka extremely obfuscated garbage code with 27 billion different symbols that mean a completely different thing dependent on whether they're written before, above, below, at a 26.3% angle to the top right of a number. math is hatred
<DedoVerde>
fuck math
<DedoVerde>
so that i can't answer but everything else i'll do my best to
<DedoVerde>
[22:36:51] <kintel> yeah, so that's curious. ..and might be a new bug/regression. But there are too many moving parts right now. I think the best bet would be some sort of example highlighting the issue << let me try
<kintel>
There might also be an O(n^2) algorithm somewhere that only starts taking effect at your scale
<DedoVerde>
i mean, technically, the "first bug" would be that the rendering, that can be triggered from outside (however that happens bzw - filesystemwatcher on the lib?! no idea...) by changing a value and ctrl+s'ing, blocks the whole GUI thread
<DedoVerde>
the second thing would be the time it takes
<DedoVerde>
just to be technical here
<DedoVerde>
but yeah, your explanations sound extremely on point from there
<DedoVerde>
sorry to ask but - are you a dev by chance? as in, openscad active dev? sounds like you know your stuff :)(
<DedoVerde>
haha well i guess we speak the same language at least i guess :D expect for that math stuff, ewww, please, do not mention that near me or my son
<DedoVerde>
*except
<DedoVerde>
you are a wizard and i appreciate that
<DedoVerde>
but me, no math, just code. simple
<kintel>
you still need some basic trigonometry :)
<InPhase>
Did someonen say math?!
<DedoVerde>
sorry, i'll try to get back to topic; just stoked to have the chance to talk to you
<InPhase>
Oh, no math. Nevermind.
<DedoVerde>
ewwwwwwwwwwwwwwwwwwwwwwww no math bah
<DedoVerde>
be gone
<DedoVerde>
demon
<kintel>
np, if you can come up with an example model that can help us fix this, it's all worth it :)
<kintel>
I'm almsot done cooking dinner, but I have raw chicken on my keyboard now
<DedoVerde>
xDDDD
<DedoVerde>
i like this place
<DedoVerde>
i'll be bugging you all a lot more
<InPhase>
kintel: That's what the rubbing alcohol spray on my desk is for...
<kintel>
I have leftover hand sanitizer from covid times, but it's somewhat sticky and smells like almonds. Not convinced
<DedoVerde>
i'll try to make an example, i really don't mean to be a... anything about it, but this thing could be huge for me, so I just can't in good conscience share it just like it is, until i decide to open source it, which i will, as soon as i can. but i have a LOT of big money competition in this field, which could just steal the idea, so... sucks, but I'll make something
<DedoVerde>
and patents, in germany, are impossible. i checked, in the USA it would cost me less than 1/5, and i'd be protected immediately - here, you get no protection until the patent is granted (3 years), and its prohibitively expensive. x_X
<kintel>
hm, so not protein folding, but datacenter A chip layout?
<kintel>
*AI
<DedoVerde>
haha no, just... well, i don't want to compromise you nice clean logs, but, let's say - my nickname hints at my profession
<DedoVerde>
brb, cig time ;)
<DedoVerde>
OpenSCAD can not only deliver in terms of being the obvious Go-To solution for clean-room, high-end engineering and chip design - it's also the ideal way to generate structures to help nature thrive. :D
<DedoVerde>
and yes, you may use that tag line xD
mmu_man has quit [Ping timeout: 265 seconds]
kintel has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
<DedoVerde>
hahahah
<DedoVerde>
ok so just for the record: if i DISABLE design => automatic reload and preview, it works as advertised - it (the model.scad) doesn't care if I edit the lib(.scad), but then, obviously, changes in the lib also don't show up in the preview in the other instance (as started from double-click)
<DedoVerde>
i gotta create that test model...
<DedoVerde>
good night y'all...
DedoVerde has quit []
mmu_man has joined #openscad
bitbasher has quit [Ping timeout: 252 seconds]
Sculptor has joined #openscad
sculptor_ has joined #openscad
Sculptor has quit [Ping timeout: 248 seconds]
J25k41 has quit [Ping timeout: 272 seconds]
sculptor_ has quit [Changing host]
sculptor_ has joined #openscad
sculptor__ has joined #openscad
sculptor__ has quit [Changing host]
sculptor__ has joined #openscad
mmu_man has quit [Ping timeout: 252 seconds]
sculptor__ is now known as sculptor
<ali1234>
^ this sounds a lot like the bug with reloading that made it reload after every keypress, which was subsequently fixed, maybe got fixed a bit too much?