<kintel_>
peeps[zen] hard to say. Did you observe memory usage over time? I didn't log it, but I saw it jumping from baseline to a very large memory usage and back again a bunch of times
<teepee>
it went to 32gb for me and then crashed with oom
<peeps[zen]>
kintel_: i saw some minor bumps up then back down, but the overarching trend was still always increasing, and mem usage doesn't drop significantly at the end of render
<peeps[zen]>
also tried your suggestion of no intermediate export, and it succeeded but only because i'm now on 32GB RAM computer this time. the openscad process alone showed 23+GB
<kintel_>
huh, why would it differ - that should be evaluated once and cached..
<teepee>
there is a rands()
<kintel_>
oh, and it's possible that we cache ManifoldGeometry not PolySet. Need to revisit that
<kintel_>
oh rands
<kintel_>
that would kill the cace
<kintel_>
*cache
<kintel_>
..so render() just collapses the CSG tree (and memory usage) every now and then
<teepee>
yep
<teepee>
that did not seem to make a huge difference for me surprisingly
<teepee>
but I just ran both cases without further investigation
<peeps[zen]>
that's with the outer code block being: union() for(a=[0:36:359]) rotate(a) render() union() {
<kintel_>
My memory usage was crazy; it would jump from 1GB up to 6GB, then back to 1
<peeps[zen]>
and for (phi=[0:1:180], theta=[0:1:35])
<peeps[zen]>
hmm, maybe differing openscad cache limits? does manifold pay attention to those at all?
<peeps[zen]>
mine are at 10GB each
<teepee>
oh, good point, and oops, I have it set to 100mb only
<peeps[zen]>
well, the thing is, even when I "Flush Caches" i don't see actual memory usage reduced as reported by system monitor. but i just tried it at 1GB each and the overall usage didn't look significantly different
<peeps[zen]>
i thought ochafik had already done some kind of work where multiple operations were batched in some way. but maybe i'm thinking of the fast-csg code?
teepee_ has joined #openscad
<peeps[zen]>
does a large loop of unions end up doing only pairwise with the current code?
teepee has quit [Ping timeout: 260 seconds]
teepee_ is now known as teepee
<kintel_>
I believe so
<kintel_>
With by batching code, it renders in 13 seconds
<kintel_>
..but spends 5 minutes doing _something_ before it shows up (VBO generation perhaps)
califax_ has joined #openscad
califax has quit [Remote host closed the connection]
califax_ is now known as califax
Smeef has quit [Read error: Connection reset by peer]
mmu_man has quit [Ping timeout: 256 seconds]
LordOfBikes has quit [Ping timeout: 245 seconds]
LordOfBikes has joined #openscad
L29Ah has quit [Read error: Connection reset by peer]
J24k48 has joined #openscad
J24k31 has quit [Ping timeout: 250 seconds]
califax has quit [Read error: Connection reset by peer]
califax has joined #openscad
kintel_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
Experimentalist has joined #openscad
R2robot has quit [Quit: Pull the lever, Kronk. Wrong leverrrrrrr!]
R2robot has joined #openscad
Experimentalist has quit [Quit: Client closed]
GNUmoon2 has quit [Ping timeout: 260 seconds]
R2 has joined #openscad
GNUmoon2 has joined #openscad
R2robot has quit [Ping timeout: 240 seconds]
GNUmoon2 has quit [Remote host closed the connection]
GNUmoon2 has joined #openscad
Experimentalist has joined #openscad
cart_ has joined #openscad
Experimentalist has quit [Quit: Client closed]
mmu_man has joined #openscad
mmu_man has quit [Ping timeout: 256 seconds]
L29Ah has joined #openscad
mmu_man has joined #openscad
cart_ has quit [Ping timeout: 260 seconds]
snaked has quit [Remote host closed the connection]
cart_ has joined #openscad
teepee_ has joined #openscad
teepee has quit [Ping timeout: 260 seconds]
teepee_ is now known as teepee
teepee has quit [Remote host closed the connection]
teepee has joined #openscad
kintel has joined #openscad
kintel has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]