<linext>
point being quick-render would be very nice to have in WebGL if possible
<linext>
also, FYI the amd 7950x renders 2x as fast as the i9900k
snaked has quit [Quit: Leaving]
snaked has joined #openscad
peeps[zen] has quit [Quit: Connection reset by peep]
peepsalot has joined #openscad
ur5us has quit [Ping timeout: 246 seconds]
J233179513341 has joined #openscad
J2331795133 has quit [Ping timeout: 260 seconds]
peeps[zen] has joined #openscad
peepsalot has quit [Ping timeout: 255 seconds]
Virindi has quit [Ping timeout: 246 seconds]
guso78 has joined #openscad
guso78 has quit [Ping timeout: 260 seconds]
Virindi has joined #openscad
JakeSays_ has joined #openscad
guso78 has joined #openscad
JakeSays has quit [Ping timeout: 264 seconds]
<guso78>
JordanBrown[m], if you wrote an openscad model and exported the CSG Tree to POVRAY, you could customize the faces nicely and you would have a nice raytrace picture.
<JordanBrown[m]>
Right, that would be the idea.
<JordanBrown[m]>
There would be two interesting aspects: converting the tree, and somehow annotating it with textures, et cetera.
<guso78>
But when you decided to change your model, you start altering the openscad model, export again and then you would have to *REDO* all the POVRAY customiztions in POVRAY which is something i think is very bad. Adding pass-though to povray parameters to openscad appears a bad way either. After all, i doubt that the function overlap between openscad
<guso78>
and povray is more than 50%
<JordanBrown[m]>
You want to arrange that all of the information comes from the OpenSCAD program, because trying to customize the OpenSCAD output independently in POV-RAY would be hard.
<JordanBrown[m]>
I think, without having even started to do it, that if you attached annotations to the tree you could supply what POV-RAY needs without having OpenSCAD know much about it.
<guso78>
i like the way to add just one texture parameter which i do in my openscad fork. main purpose is to enjoy myself and my fork but i don't deny to share after all I dont want to stress teepee with even more PR's he has been busy for weeks now with the python pr and we could not even manage to rebase the python-support branch to the latest main
<guso78>
branch.
<JordanBrown[m]>
I think you can drive that rebase, giving him a PR that he can just approve.
<JordanBrown[m]>
But I don’t know the exact git incantations.
<JordanBrown[m]>
Bedtime soon, so I’m going to drop off.
<guso78>
JordanBrown[m], one week back i told him it would be a good idea to "rebase" and short immediate replay was "Yep" . Now one past and nothing happend. i must assume, that actually he does not like the stuff.
<JordanBrown[m]>
He’s busy with the real world.
<guso78>
JordanBrown[m], true, but he approves many other pr's and mine does not even affect the main branch
<JordanBrown[m]>
Did you give him a PR, or just say it should be done?
<guso78>
JordanBrown[m], its not about approving a PR. its the task to rebase the python-support branch with latest main branch. yes i am sure he understood and he replyed with "yep"
<guso78>
its just few clicks, but it appears to me he is "delaying" by intent. his decision.
<JordanBrown[m]>
Right, and I believe the way that rebase happens is a PR against the Python branch.
<guso78>
yes, only python branch is affected, main branch is safe
<JordanBrown[m]>
Rebase is not just a few clicks - it’s merging and testing.
<guso78>
not sure how it works in the github GUI. i would do it in a git branch in my linux console. 3 commands and testing is done automatically after pushing back to github
<JordanBrown[m]>
Somebody needs to merge the state of the main branch into the Python branch, test to make sure the Python branch still works, and make a PR against the Python branch. Then he can approve that with a few clicks.
<guso78>
never mind, i got many other intersting tasks. will focus again on my face tracer :)
<guso78>
JordanBrown[m],i have did the ctest of the python branch locally and it was pass
<JordanBrown[m]>
If it merges clean, then yeah, but I would personally want to test it a bit before pushing.
<guso78>
(after few iterations)
<JordanBrown[m]>
That’s after merging the stuff from main?
<guso78>
i merged main in locally and it was clean, but i merged back to actual state of python-support branch because i thought its a good idea to exactly to to where i come freom
<guso78>
from
<guso78>
JordanBrown[m], is it better to merge actual state of main and push again to python-support branch ?
<JordanBrown[m]>
I *think* that what you need to do is a git pull of main, then rebase your python work on top of that, then push -f that, then PR that. But I am far far far from a git expert.
<JordanBrown[m]>
It might involve merges rather than rebase per se.
<guso78>
i will ask teepee about that.
<JordanBrown[m]>
Yes.
<JordanBrown[m]>
Anyhow, as I said, g'nite...
<guso78>
probably he will read all log and will approach me ...
use-value has quit [Remote host closed the connection]
use-value has joined #openscad
<teepee>
guso78: obsolete with fast-csg only?
pbsds has quit [Ping timeout: 246 seconds]
<guso78>
teepee, generally lookging into the code: each time when its BACKGROUND , then its not HIGHLIGHT. so being HIGHLIGHT is a subset of NOT BACKGROUND
teepee_ has joined #openscad
<guso78>
Hence I believe the first term "IS BACKGROUND" can be removed.
<guso78>
sorry: BACKGROUND is a subset of NOT HIGHLIGHT
teepee has quit [Ping timeout: 255 seconds]
teepee_ is now known as teepee
pbsds has joined #openscad
use-value has quit [Remote host closed the connection]
use-value has joined #openscad
snaked has quit [Quit: Leaving]
guso784 has joined #openscad
guso78 has quit [Quit: Client closed]
guso78 has joined #openscad
YesMan has joined #openscad
guso784 has quit [Quit: Client closed]
<YesMan>
Hello does anyone knows how the SimScale Community account and it's 10 successful sim limit works? Does your available successful sim pool replenish or is it 10 for life?
<teepee>
YesMan: nope, never heard about that and I suspect you are in the wrong channel
<YesMan>
Sorry, do you know of a larger scope cad channel on libera? thank you
noonien has joined #openscad
califax has quit [Remote host closed the connection]
califax has joined #openscad
Guest25 has joined #openscad
Guest25 has quit [Client Quit]
<teepee>
YesMan: there's #CAD but it seems to have only a single person
<teepee>
you can search via. /msg alis list *cad*
Lagopus has joined #openscad
J233179513341 has quit [Quit: Client closed]
J233179513341 has joined #openscad
Crocmagnon has joined #openscad
Crocmagnon is now known as undefined
undefined is now known as Crocmagnon
Crocmagnon has quit [Remote host closed the connection]
Crocmagnon has joined #openscad
Crocmagnon has quit [Remote host closed the connection]
<nbr0wn>
Anyone here familiar with the progress_report internals?
<JordanBrown[m]>
I am not.
<teepee>
nbr0wn: you mean the code creating the progress bar?
<nbr0wn>
Specifically the way that the progress counts are calculated. I've dug in far enough to extract them, but it seems like the counter backtracks sometimes and then picks up where it left off.
<teepee>
I think it counts CSG tree nodes
<teepee>
normalized against the number of nodes
<nbr0wn>
Yeah....so for example one file I have, the max count is 2573...it'll happily count up to 2566...then jump to 55, then 1292, then 2566 again, then carry on
<nbr0wn>
I'm just wondering what it's smoking during those times :)
<teepee>
hmm, that's odd, does that show up in the progress bar too?
<teepee>
maybe it's just traversing and using the maximum value
<nbr0wn>
It doesn't, no
<nbr0wn>
I could keep track of the max value and just ignore any deviations, but it seemed odd
<nbr0wn>
Yeah, looks like that's what the Qt progress bar is doing
<nbr0wn>
When in Rome I guess...
<teepee>
yep, it has that max compare
<teepee>
it iterates the node tree depth first, so the numbers go down when it goes back up and reaches 1000 once it reaches the last child node from the last branch
<nbr0wn>
I see...do you think it will usually traverse in a more-or-less increasing fashion, or are there going to be cases where the tree is wildly imbalanced, and may hit some of the highest numbered nodes first?
<nbr0wn>
Or early...causing the progress bar to stall
<teepee>
it tends to get stuck at the end usually because it does scale with operation count not operation code and the last operation usually is the most expensive one by far
<nbr0wn>
Hm...maybe I should just count the nodes rather than trust the node id
<teepee>
now I'm asking :P... trust for what? any cool feature happening?
<nbr0wn>
Heh...trust it to have as reliable a progress bar as I can make without doing too much extra work. I made a thing: https://parapart.com that uses openscad-wasm, but the STL export can be slow, so I wanted to add a progress bar for a little feedback
<nbr0wn>
It still needs some polish and a bunch of real data, but it's getting there.
<teepee>
nice
<JordanBrown[m]>
teepee: The French translation PR looks fine to me.
<JordanBrown[m]>
Though I think that L'Académie Français and its allies may be fighting another losing battle.
<teepee>
JordanBrown[m]: haha, I'm not too fuzzed about that, I usually go through the translation PRs via google translate to see if there's anything crazy going on but otherwise hope people will not try to be bad
<JordanBrown[m]>
Then again, I speak a language about which it has been said "We don’t just borrow words; on occasion, English has pursued other languages down alleyways to beat them unconscious and rifle their pockets for new vocabulary."
<JordanBrown[m]>
I was curious because GT said that the two words have the same meaning.
<nbr0wn>
teepee: Thanks, I may reach out
<JordanBrown[m]>
... which tells us that GT is descriptivist, while L'AF is of course prescriptivist.
<teepee>
in totally different news, there's some work going on about a WebGL CSG viewer, maybe that could be combined with the WASM version of OpenSCAD too
* teepee
has no idea what that scriptivist stuff means ;-)
<JordanBrown[m]>
They are often applied to dictionaries.
<nbr0wn>
That would likely streamline things as I'm having to export STL and then load it into a viewer
<JordanBrown[m]>
Descriptive dictionaries tell you what the language *is*, on the street.
<JordanBrown[m]>
Prescriptive dictionaries tell you what the language is *supposed* to be.
<JordanBrown[m]>
Describe versus Prescribe.
<teepee>
ah, I suppose in germany we are too boring and only have Duden as single source
<JordanBrown[m]>
Sorry, I don't know Duden. I know a little French, but even less German.
<JordanBrown[m]>
And of course over here in the US we think that L'AF is a very strange concept.
<teepee>
it's the main known dictionary where the name became a synonym for the whole category
<teepee>
nbr0wn: yeah, a cached CSG tree + client side CSG viewer could be a great thing for performance
<JordanBrown[m]>
And of course, these days in American English we have all these people talking about "crypto" and not referring to cryptography at all.
<teepee>
cyber crypto!
<JordanBrown[m]>
and don't get me started about people who think that "email" is a countable noun.
teepee_ has joined #openscad
peeps[work] has quit [Quit: Leaving]
teepee has quit [Ping timeout: 255 seconds]
teepee_ is now known as teepee
<teepee>
so we'll follow the authoritative answer :)