<J24k8>
imagine the customizer profile name would be added to the default export name. Design-customizedSet1.3mf
mohammad1722 has joined #openscad
<Scopeuk>
I guess we could look towards something like the slicers do with a defined export name in the settings using a bunch of substitutable strings to define the final names
<stealth_>
darn it, openscad just crashed.
<Scopeuk>
stealth_ if it's crashing randomly and works after a reboot I have a slight concern your machine is experiencing a ram issue
<Scopeuk>
If you can repeat the crash of course let use know and raise a GitHub issue so we can chase it down
<stealth_>
it just crashed as i was coding, nothing different then what i been doing normally
<stealth_>
system has over 16GB of ram
<Scopeuk>
I don't mean out of ram I mean there may be an area of ram which is malfunctioning.
<stealth_>
thats too much of stretch! lol
<Scopeuk>
A stretch perhaps but it tends to be the most common cause of random crashes. As I say if you can find a way to reliably trigger it we want to know
<stealth_>
will keep that in mind.
<stealth_>
Scopeuk, by any chance you know of a tool to find out if its actually the ram?
mohammad1722 has quit [Ping timeout: 256 seconds]
<Scopeuk>
The usual tool is memtest64. That runs off a bootable usb stick and it will just sit there in a loop testing the ram, it will flag up any errors it finds
<Scopeuk>
It's not a purchased tool, there is a scam out there, most Linux distros/installers come with them
<stealth_>
ya i know what you are talking about, if i get some time to do that i might run it, way too busy to let system do that atm.
<Scopeuk>
Yeh, generally recommend just leaving it running overnight whilst the machine would otherwise be idle
<stealth_>
Scopeuk, true, thanks for your input :)
ccox has joined #openscad
ccox_ has quit [Ping timeout: 260 seconds]
mtm has quit [Ping timeout: 252 seconds]
mtm has joined #openscad
teepee_ has joined #openscad
teepee has quit [Ping timeout: 260 seconds]
teepee_ is now known as teepee
<teepee>
stealth_: I tried AntiMicroX, it seems to only map from gamepad -> mouse/keyboard
<stealth_>
teepee, o i c, you didn't have to do that... i was just curious, like it would have been nice but its not a big deal. thanks though.
<teepee>
I'm searching for some tool too, mostly for other reasons
<stealth_>
i was using sublimetext so i am used to it, now i am trying out the editor that comes with openscad since the renderprocess is faster for multiple files...
<stealth_>
aww
<teepee>
there is input-mapper which can map mouse buttons to keyboard input but it does not work as well as I'd like
<teepee>
it seems to be confused by 2 mice of the same make (I have 2 logitech trackballs mx-ergo) and it often only maps one :(
<stealth_>
well it seem overkill to have to use such feature, editor should just add mouse click detection
<stealth_>
for normal mouse
<teepee>
yes, agreed, unifying the input handling would be nice
<stealth_>
mhmm
<teepee>
in other news, I'm not going to accept "c++ compiler errors are a mess" anymore
<stealth_>
o, i have visited that site, thought it was neat, not sure if it was maintained
<teepee>
maintained by guso78
<teepee>
some small parts are merged, but the big stuff is still pending
<teepee>
I wish I had like 2 month vacation or something ;-)
<stealth_>
aww... well i wonder if they could have improved the interface where it didn't need to manage the `c = cube(...)` and just had it output the function ast into a temp file and render it, that way it would be close to how openscad code is.
<teepee>
yes, he said it's possible but required some python hack IIRC
<teepee>
but maybe someone else has some more ideas how to do it
<stealth_>
another approach would have been to create a custom system that was closed to python TAB style... i hate the {;} usage.
<teepee>
it's still the best way of integrating things I've seen over the years
<teepee>
all the other ones are ugly wrappers needing a huge amount of boiler plate code
<stealth_>
custom syntax*
<stealth_>
well such features needs to be implemented at the start of the project.
<teepee>
not really, it's just a matter of what that language/library provides
<teepee>
but yes, it *might* be quite a bit more complicated later on
<othx>
teepee linked to YouTube video "SUI23 - Introducing Bidirectional Programming in Constructive Solid Geometry-Based CAD" => 1 IRC mentions
<stealth_>
teepee, i am saying you can do it like this https://bpa.st/44TA look at all the way bottom.
<teepee>
no, I'm saying that I believe I saw code similar to what I have written above
<teepee>
but guso78 can give more insight when he's around
<stealth_>
teepee, i get what you are saying and i agree it can be done.
<teepee>
I believe it *has* been done, as in actually working
<stealth_>
i am just saying if planned right the project can be done the way i mentioned all the way at the bottom as well, but it needs being very creative.
<teepee>
as in "a different language" just with different syntax?
<teepee>
I suppose that's possible but does not seem very realistic
<stealth_>
teepee, no, should be able to do it in python
<teepee>
in the end it's a matter of someone doing the work
<stealth_>
why i said it needs to be planned from the start
<teepee>
if it's possible with "normal" python, poke guso78, I think he loves experimenting with stuff
<teepee>
so he just might try implementing it :)
<teepee>
as there is no official integration yet, anything is still open for grabs
<stealth_>
ya, that might be cool
<teepee>
only as soon as things are officially released, backward compatibility becomes a big topic
<stealth_>
since it kind of maintains the openscad code style and it would be easy for anyone to migrate for either or
<teepee>
a number of strange openscad behaviors are not really "fixable" as that would cause existing stuff to break
<teepee>
and we don't really want to break all existing designs on thingiverse
<stealth_>
true, true
<t4nk_fn>
teepee, that 'Bidirectional Programming' stuff was directed at me I guess?
<teepee>
yes
<t4nk_fn>
I'll need to have another look and listen later on, couldn't realy follow along here just now.. but it's interesting and I'll definitely have a look!
<teepee>
yep, I'll try to get the mails going again, I asked them if we could just talk here on irc
<teepee>
but it's been quite for a couple of weeks
dbyy has joined #openscad
<teepee>
lol, I'm bitten by the python virus it seems, starting to forget typing the ; for c++
<t4nk_fn>
lol
<dbyy>
hello all
<t4nk_fn>
I think it's optional for python, right? at least I hope so, the other day I decided to do it anyhow when I *shudder* did something in python
<teepee>
yeah, I think it's allowed as statement separator in python, not sure
<teepee>
but I'm fighting with some older c++ code failing to build with newer gcc
<teepee>
hello dbyy
<dbyy>
hey teepee hows going?
<teepee>
good, lazy saturday
<dbyy>
yeah a rainy lazy Saturday
<dbyy>
may I ask a question?
<teepee>
sure
<dbyy>
I want to start a little project and create a box with some mesh on the side walls. Is there somewhere, something I could look at (libs, models) to get some ideas how to approach it?
<teepee>
mesh meaning like a grid structure?
<dbyy>
Yes, to reduce the material to be printed, e.g. a mesh made of hexagons
<dbyy>
I saw some tutorials for FreeCAD but I'd like to try it in OpenSCAD
<teepee>
hmm, not sure if there's any dedicated library for that, but the bigger all-purpose libraries might have something
<J24k8>
dbyy HexGrid(help=1) will arrange children in a hex pattern (find in ub.scad) As it provides a $pos parameter you can conditionally switch on and off
TheAssassin has quit [Remote host closed the connection]
TheAssassin has joined #openscad
L29Ah has left #openscad [#openscad]
kintel has joined #openscad
L29Ah has joined #openscad
<kintel>
pca006132: FYI in case you have ideas. Triangulation saga continues: So my theory is that the triangulation issues happen because we have guaranteed non-overlapping polygons in int64 fixed point object space which become overlapping in double world space.
<kintel>
One random test dropping the fixed point scale, causes the precision loss and hence aliasing in double to become less likely, which drops the crash rate a bit
teepee_ has joined #openscad
teepee has quit [Ping timeout: 260 seconds]
teepee_ is now known as teepee
stealth_ has quit [Quit: Leaving]
<kintel>
teepee I noticed that snapshots are removed after 33 days from the file server using a cron job. I assume this is to avoid runaway disk usage. Do you think we could up that to a year or so? I find it useful for certain regression testing processes to have old binaries available
<teepee>
sure, the disk is not very big, but we can see how far it goes, right now there's not much else running
fling_ has joined #openscad
fling has quit [Ping timeout: 260 seconds]
fling_ is now known as fling
raboof has quit [Ping timeout: 260 seconds]
raboof has joined #openscad
<kintel>
teepee worst case, disk usage for snapshots go from 3GB to 30GB. With 690GB total, it should be fine :)
<kintel>
I'm not familiar any longer how the scripts are maintained though :/
aiyion1 has quit [Ping timeout: 260 seconds]
Gungee has quit [Remote host closed the connection]