<linext>
teepee, that reminds of trying to run non-blocking stdin from php on windows
<linext>
i was complaining that they should make all OSes behave the same and they told me windows just doesn't have it
<linext>
so the answer was c)
Guest12 has joined #openscad
Guest12 has left #openscad [#openscad]
<teepee>
on the plus side, it seem the combination of the latest OpenGL changes from kintel with the v24 mesa3d version is back to working on a windows VM
<kintel>
linext Who would ever run php on Windows? ;)
<kintel>
teepee I think I didn't even merge my Windows GL changes yet..
<teepee>
well, at least it would not suffer from that php+glibc bug ;-)
<teepee>
then it's just the mesa3d fixes?
<JordanBrown>
teepee thumbnailer? Does it just capture a thumbnail and stash it away when you F5, or does it run OpenSCAD programs just because you visited the directory they're in with Explorer?
SamantazFox has quit [Killed (NickServ (GHOST command used by SamantazFox_))]
<teepee>
well, there's 2 more lines, but that's the core of it :)
LordOfBikes has quit [Ping timeout: 245 seconds]
<JordanBrown>
Do you have to run it manually, or does some Explorer-like took (Nautilus?) run it automatically?
<JordanBrown>
If it's run automatically, it makes "write" operations even more terrifying; downloading a malicious file and merely visiting the directory might be enough to allow an attack.
<JordanBrown>
s/took/tool/
<teepee>
yep, fully automatic, hence I will always be against openscad scripts to write anything by themselves
LordOfBikes has joined #openscad
<JordanBrown>
Though even infinite loops are a DoS.
<teepee>
not sure if it catches that, it might via system limits
<teepee>
but I always wondered if there should be some --min-fa=5 --min-fs=0.5 or something like that
<teepee>
well and --max-fn=72
<teepee>
not solving dedicated bad scripts but lazy ones :)
J24k16 has joined #openscad
mmu_man has quit [Ping timeout: 268 seconds]
J24k41 has quit [Ping timeout: 250 seconds]
zauberfisch has quit [Ping timeout: 264 seconds]
<JordanBrown>
OpenSCAD itself should perhaps have a timeout parameter. But first (for interactive work) should be a stop button that actually works :-)
<teepee>
that would be a technical prerequisite
<JordanBrown>
I wonder how much of a performance hit it would be to do renders in a separate process, that we could just kill.
<JordanBrown>
We might not be able to do the OpenSCAD execution in a separate process, because of how preview works, but it doesn't seem like it should be *that* hard to have an abort button that works there.
pca006133 has joined #openscad
foul_owl has quit [Ping timeout: 260 seconds]
<JordanBrown>
"Just" run it in a separate thread, have the main UI thread maintain a "need to stop" flag, and have the execution thread check that flag at opportune points - perhaps every loop iteration and every module and function call.
pca006132 has quit [Ping timeout: 252 seconds]
pca006133 is now known as pca006132
<JordanBrown>
But for the render phase, where we don't have as much control, if we could hand the CSG tree to a separate process and have it hand back triangles...
foul_owl has joined #openscad
pca006132 has quit [Ping timeout: 245 seconds]
kintel has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
pca006132 has joined #openscad
pca006133 has joined #openscad
pca006132 has quit [Ping timeout: 245 seconds]
pca006133 is now known as pca006132
pca006133 has joined #openscad
pca006132 has quit [Ping timeout: 252 seconds]
pca006133 is now known as pca006132
pca006133 has joined #openscad
pca006132 has quit [Read error: Connection reset by peer]
pca006132 has joined #openscad
pca006133 has quit [Ping timeout: 256 seconds]
pca006133 has joined #openscad
pca006132 has quit [Ping timeout: 268 seconds]
pca006133 is now known as pca006132
pca006133 has joined #openscad
pca006132 has quit [Remote host closed the connection]
pca006132 has joined #openscad
pca006133 has quit [Ping timeout: 256 seconds]
L29Ah has joined #openscad
marcus has quit [Remote host closed the connection]
marcus has joined #openscad
qeed has quit [Quit: qeed]
Guest6 has joined #openscad
fling has quit [Remote host closed the connection]
fling has joined #openscad
J24k75 has joined #openscad
J24k16 has quit [Ping timeout: 250 seconds]
cart_ has joined #openscad
snaked has quit [Remote host closed the connection]
mmu_man has joined #openscad
erectus has quit [Ping timeout: 260 seconds]
erectus has joined #openscad
Guest25 has joined #openscad
Guest25 has quit [Client Quit]
erectus has quit [Remote host closed the connection]
killjoy has joined #openscad
killjoy has joined #openscad
erectus has joined #openscad
<pca006132>
I think it is perfectly fine to have a separate process to do rendering
<pca006132>
typically people try to avoid using process because the creation overhead (~2ms for Linux? not sure for Windows and Mac), as well as communication overhead
<pca006132>
for the mesh we work with, I think communication overhead isn't really that huge
<pca006132>
and users generally don't care about 2ms delay for smallest model I guess
<teepee>
but manifold can cancel running operations, right?
<pca006132>
no
<pca006132>
at least not yet
<pca006132>
probably not hard, but will require some intrusive changes
qeed has joined #openscad
<teepee>
hmm, mixing it up with something else then... I thought ochafik did look at this and found it's already possible.
<teepee>
are there other callbacks happening? maybe he thought of injecting an exception via some processing callback?
<pca006132>
I don't think so
<pca006132>
I know there are some functions that can take a long time and does not call any callback
guso78k has joined #openscad
teepee_ has joined #openscad
teepee has quit [Ping timeout: 260 seconds]
teepee_ is now known as teepee
<teepee>
the is an unfinished PR using boost serialization for data transfer to a separate process
<Guest6>
tepee: LOL, I spent 4 plus hours last night trying to be a prompt engineer for a .scad file I'm trying to fix, to no avail.
<Guest6>
using chatgpt.
<peeps[work]>
have you tried asking a human intelligence?
Guest6 has quit [Quit: Client closed]
<teepee>
apart from the ] vs. ) typo, the list comprehension is suspiciously wrong ordered, like in python
mtinman has joined #openscad
<mtinman>
Was Guest6, now myself
<mtinman>
So, as I was saying, it appears that chatgpt is not so good at scad coding, or I'm just not good at prompt engineering.
<teepee>
I said: apart from the ] vs. ) typo, the list comprehension is suspiciously wrong ordered, like in python
<teepee>
hence the chatgpt guess
<teepee>
JordanBrown: I did not look very deep into it, but gnome seems to use bwrap to isolate the thumbnailer
<mtinman>
I figured that I would come over here to see if I could get a "correct" answer to my scad question, as I am no programmer, and have put forth a real "college try" in an attempt to get a function added to a scad script I am working with, but not having good luck at it.
<teepee>
well, you could put the script into some pastebin and ask the question
<teepee>
right, so now you need to go through the definitions on top of the file and copy any of the *3 variables and create a *t for those too
<mtinman>
Ok, I'll work on that...
<teepee>
e.g. there's line3 = "+1-234-567-1234";
<teepee>
and we need a linet = "+1-234-567-1234";
<teepee>
as copy
<teepee>
after that F5 should work again
<mtinman>
Yeah, I saw the warnings, LOL.
zauberfisch has joined #openscad
<mtinman>
teepee, for clarity I want title line to look like line 4 here, not line 3. When I made that code change, it makes titlte line look like line 3.
<mtinman>
Hold on, I'm gonna start over, LOL
<teepee>
oh, oops, then copy the other block
<teepee>
the one with line4
<mtinman>
Ok, so I'm going to make the space for the title line first...
<mtinman>
Ok, done.
<mtinman>
Now let's make title line look like line 4.
<teepee>
right, so copy the 4 lines of that last block that creates line 4
<mtinman>
AFAICT, it's line 83, starts with: translate([0,offset_line_4,Plastic_thickness-0.2])
<mtinman>
DOwn toward the EOF.
<teepee>
yep, and 4 lines including that last one just with the }
<teepee>
no wait, 5 lines, two } at the end
<mtinman>
DO you have 93 lines in the original file?
<mtinman>
Mine ends with ;
<teepee>
let me open the original file again
<mtinman>
I don't think that's right, to clodse with ;
<mtinman>
Oh yeah I guess it does.
<teepee>
83-87
<mtinman>
Copy that to 88?
<teepee>
if you place the cursor after the } in line 87 it highlights the matching opening brace in line 83 at the end
<teepee>
yep, copy to 88
<mtinman>
I'll send a code snip for you to look at, just a sec.
TheCoffeMaker has quit [Ping timeout: 256 seconds]
<InPhase>
Brass.
<mtinman>
Yeah, I'm surprised my brass nozzles are holding out fairly well, I just changed the retraction settings a bit, seems to be doing better, but I still need to replace the nozzles on my printers, lord knows I got like 30 spares...
<InPhase>
I just never bothered to buy spares for a while, and now it's too late... And time to replace the whole unit I think. :)
TheCoffeMaker has joined #openscad
<teepee>
never had anything other than the single default 0.4mm brass nozzle until recently
<teepee>
only changed with the prusa-mk4 now
<mtinman>
Yeah, there are a LOT of nice "replacement nozzles" to choose from nowadays, 3d printing has changed significantly over just the last 4 years, since I started... And tepee, I think he's referring to buying a new printer, not sure, LOL.
<teepee>
0.8 is nice, but does suffer a bit from the not too powerful heater of the mk4
<mtinman>
is mk 4 direct extrusion?
<teepee>
probably, like I did with the mk4 which is now also 1.75mm filament which was the biggest problem as the 2 older ones were all 3mm
<teepee>
yes, the extruder is quite neat, with included load-cell for probing, so it does actually probe with the nozzle
<teepee>
will have to see how it goes with the multi-material-unit, that just shipped after half a year delay :)
<mtinman>
Oh wow, that is a COOL feature, no probe offset to worry about!
<teepee>
yep, it's really impressive, I never thought I need things like bed leveling, but seeing that now in action is just astonishing