ToAruShiroiNeko has quit [Ping timeout: 255 seconds]
ToAruShiroiNeko has joined #openscad
kintel has joined #openscad
kintel has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
kintel has joined #openscad
<kintel>
In terms of scoping and include<> and use<>. Not sure if this is worth discussing, but I've found the current implementation of use<> and include<> a bit annoying to understand and maintain. At some point I wanted to rewrite those to become AST nodes, and in the process encode the logic in a nicer way. I think that's possible without breaking too many existing use-cases. Not sure if this new discussion affects this idea.
<kintel>
Again, just throwing it out there as I haven't looked into that in a long time.
<pca006132>
JordanBrown_: The dependency graph thing is useful for other optimization as well, e.g. cache module calls that have the same parameters
<teepee>
ah, reddit has a post from someone I know ;-)
<teepee>
no comments though
fling has quit [Remote host closed the connection]
fling has joined #openscad
<pca006132>
maybe I can ask emmett if he wants to contribute, he has some nice designs but he is not using the mailing list or reddit iirc
<pca006132>
(those designs are usually pretty mathematic)
<pca006132>
btw, I wonder if there is any interest (or opposition) to implement some sort of an update checker that users can configure, e.g. disable completely, receive news like SCADvent, and whether or not to check for nightly builds
<pca006132>
as a bonus we can probably estimate the number of users from the number of requests (DDOS on the server is a possibility though)
<pca006132>
I know some people are strongly opposed to this kind of telemetry, but not sure how people will perceive this if it only fetch data and can be disabled completely
<Scopeuk>
happy to licence + add if we think it's worth having
<Scopeuk>
hmm looks better with one of the red lines made green
<Scopeuk>
and I think I like it spinning even if it does explode the number of objects https://paste.debian.net/1300342/ a bunch of leavers to pull and it becomes 3d spirograph
mmu_man has joined #openscad
fling has quit [Remote host closed the connection]
snaked has quit [Quit: Leaving]
<J23k12>
Scopeuk if you use a chained hull you need less objects - also there is no license info
<Scopeuk>
yeh, there are a whole bunch of ways to make it more efficient
<Scopeuk>
I had originally planned for a sort of dna double helix then got distracted
<pca006132>
just tried it, the GUI window was frozen for several seconds and say the rendering only took 0.15s :)
<pca006132>
I guess that is a bug?
<Scopeuk>
not really, the preview is a completely sperate render system, it has different performance
<Scopeuk>
I'm guessing from that 0.15 you are using the manifold version?
fling has joined #openscad
<pca006132>
no, I was just clicking the preview without clicking render
<pca006132>
it says "Compile and preview finished. Total rendering time: 0:00:00.138"
<pca006132>
while the render with manifold says "Rendering Polygon Mesh using Manifold... (skipped the stats) Total rendering time: 0:00:04.387 \n Top level object is a 3D object: \n Facets: 1501404\n Rendering finished."
<Scopeuk>
ahh I misread your message, I read it as preview was slow but render was very quick. but it was actually window frozen from a while and then reporting that time was tiny
<pca006132>
oh but with the CLI I do get the report time
<pca006132>
so probably some GUI quirk
fling has quit [Remote host closed the connection]
<Scopeuk>
I wouldn't be surprised to find that the timer thread is depending on the main ui thread starting it so on intial launch it didn't start until the backing data for the preview had been generated
<pca006132>
with VBO disabled it is much faster, without the window frozen issue
<pca006132>
probably something with VBO
fling has joined #openscad
<pca006132>
(I do wish that manifold can render this in 0.14s, but this seems too hard)
<Scopeuk>
yeh, that model is basically everything done wrong as far as performance is concerned
<pca006132>
I feel that the performance is fine, as long as you don't do minkowski
teepee_ has joined #openscad
<pca006132>
I really wish we can fix the performance with minkowski, but the problem is just too hard, intrinsically
<teepee>
as for update-checks, we currently only have Sparkle which is macOS only
fling has quit [Remote host closed the connection]
<pca006132>
it seems that Sparkle is also performing the update
<pca006132>
the update-check that I am thinking about will just remind the user about there is an update, which should be easier for cross-platform
fling has joined #openscad
<teepee>
a more limited cross platform way might be nice. or we just keep Sparkle and have something simpler for windows/linux
misterfish has quit [Ping timeout: 264 seconds]
greenbigfrog has quit [Ping timeout: 256 seconds]
fling has quit [Remote host closed the connection]
<Scopeuk>
I guess simplist is a page on the website that has a single json/xml/single string, with latest version information and the application does a http get and then checks it against it's internal revision
fling has joined #openscad
<teepee>
yes, we have that for the printservices already
<teepee>
website being files.openscad.org as the main website is hosted via the github repo so that would be a bit annoying to update
<teepee>
well or some service on that machine :)
greenbigfrog has joined #openscad
mmu_man has quit [Ping timeout: 252 seconds]
teepee has quit [Remote host closed the connection]