califax has quit [Remote host closed the connection]
califax has joined #openscad
kintel has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
J23k38 has joined #openscad
J23k39 has quit [Ping timeout: 246 seconds]
peepsalot has quit [Quit: Connection reset by peep]
peepsalot has joined #openscad
RichardP_ has quit [Quit: Leaving...]
mmu_man has quit [Ping timeout: 250 seconds]
LordOfBikes has quit [Ping timeout: 260 seconds]
LordOfBikes has joined #openscad
IceFlinger has joined #openscad
IceFlinger is now known as iceflignerss
iceflignerss has quit [Client Quit]
<InPhase>
peepsalot: Ah yes, I remember discussing that RGB version with you. :) Much clearer, and yet still seems very unpleasant to work with.
<InPhase>
Probably something like that is the only reasonable way to do that in 3D though. You just need that extra dimension of visual separation to handle the extra dimension of content.
<InPhase>
Wait, it says forked from rcolyer... I didn't even remember that I ever made a shadertoy account. But it looks like I did it for just playing with that concept.
<joseph_>
kintel: Thank you for the note about the GLAD PR. Last week was mainly a "reading week" focused on some of my knowledge gaps. University projects (and last year's GSoC, in a sense) are narrower tasks inside pre-configured environments. So I'm finding out more about the operation/function of GL libraries and early-stage configuration for an application.
<joseph_>
This knowledge will be most useful in future weeks, when I'm actually testing out builds with different profiles after tracking down legacy API calls. The current tasks, which are mainly non-functional refactoring, depend less directly on this background information. But I at least wanted to get started so I can ask potential questions as they arise.
<peepsalot>
InPhase: yeah i don't remember who changed what between the forks, but you forked my original grid, then i forked yours and there was color added in there by someone, lol
<peepsalot>
it was kind of a pain to tweak, still difficult to wrap head around shader code
<joseph_>
I know about the midterm evaluation next week, so I'll work on adding more to the PR and coordinate with teepee about Google's forms. I can ping once I've gotten new changes ready for review and pushed them
kintel has joined #openscad
dustinm` has quit [Server closed connection]
dustinm` has joined #openscad
<kintel>
peepsalot yeah, for usability, perhaps an easy way to toggle the grid of each plane.
<kintel>
joseph_ I realize that your PR is still a draft, but for the purposes of code review, it would be good to know the intended scope of the PR, so that it can be reviewed and merged relatively rapidly
<kintel>
Also, see the issue opened by guso78 on creating a PolySetIndexed. If he starts working on that, it would probably be very beneficial to sync up with him
<kintel>
Hopefully just a missing dependency for installing PIL. Ideas are welcome :)
<InPhase>
peepsalot: Well it looks like I added putrid colors and then you made them palatable. I had a brief mental flashback of me trying to explain the idea of coloring by axis without much clarity, and you not understanding me, so I tried to edit a draft to illustrate what I meant.
teepee has quit [Remote host closed the connection]
teepee has joined #openscad
LinuxHackerman has quit [Server closed connection]
LinuxHackerman has joined #openscad
kintel has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
Guest3 has joined #openscad
Guest3 has left #openscad [#openscad]
anonymous has joined #openscad
L29Ah has quit [Ping timeout: 250 seconds]
h-erectus has quit [Ping timeout: 240 seconds]
h-erectus has joined #openscad
anonymous has quit [Quit: Client closed]
anonymous has joined #openscad
srk has quit [Remote host closed the connection]
srk has joined #openscad
qeed__ has joined #openscad
qeed_ has quit [Ping timeout: 246 seconds]
foul_owl has quit [Ping timeout: 260 seconds]
TheAssassin has joined #openscad
foul_owl has joined #openscad
anonymous has quit [Quit: Client closed]
TheAssassin has quit [Remote host closed the connection]
TheAssassin has joined #openscad
TheAssass1n has joined #openscad
TheAssassin has quit [Remote host closed the connection]
<Scopeuk>
looks like pillow wants zlib and libjpeg headers https://pillow.readthedocs.io/en/latest/installation.html#external-libraries if we aren't using png we could disable the z lib headers or similar for jpeg. looks like the project heavily recommend using the prebuilt wheel via pip rather than building from source
TheAssass1n has quit [Remote host closed the connection]
<Scopeuk>
https://www.youtube.com/@gatlingtommy/videos so this is fun, but curiously from his instagram page e development of the #motherofallblasters was a hell of a ride, here are some numbers of the development process:
<Scopeuk>
* consumed >1000h of my time (incl. getting experience in 3D printing etc.)
<Scopeuk>
* in that time >50kg of filament was used
<Scopeuk>
* with the resulting blaster having >70 different 3D printed parts
<Scopeuk>
* printing >800 iterations of part changes
<Scopeuk>
* the design is described via >8000 lines of OpenSCAD code
<Scopeuk>
appologies for multi line spam
<Scopeuk>
not an open design unfortunately
misterfish has quit [Ping timeout: 240 seconds]
mmu_man has quit [Ping timeout: 240 seconds]
kintel has joined #openscad
<kintel>
About pillow: It looks like we used to be able to install the wheels, but that stopped working at some point in the very recent past. Perhaps GitHub changed the Python installation in the CI images
kintel has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
mmu_man has joined #openscad
teepee_ has joined #openscad
teepee has quit [Ping timeout: 240 seconds]
teepee_ is now known as teepee
misterfish has joined #openscad
<Scopeuk>
hmm not entirely sure what is going on with the pillow python and pip, we are using different versions from the runner and they look to be in the msys64 installation directory, I guess this "fun" could be hit if we used runner pip with msys python
Scopeuk has quit [Quit: Scopeuk]
Scopeuk has joined #openscad
<Scopeuk>
hmm it looks like we are on a VERY recent pillow release too 2023/07/01
<Scopeuk>
so maybe the introduction of the issue was the wheel version bump from top of 9 series to start of 10 series and we can just pin the version for a bit
mmu_man has quit [Ping timeout: 240 seconds]
<InPhase>
Scopeuk: We're actually not on any release. I set it up initially to float with most recent, as there's no need to lock down and have epochs of upgrading as long as we're not having problems.
<InPhase>
But, I would certainly concur with someone version-locking it for a time if that bypasses a temporary issue that arises in the recent releases. That's always a delay of resolution sort of thing, but sometimes it is the right answer.
<Scopeuk>
InPhase yeh, I can see it followed latest and we now have a break, the time frame is right. there are no bugs on mysys's package or the upstream that seam relevant and it could be very specific to our enviroment. I guess pinning, trying a run and if it is fixed the issue is at least isolated
<InPhase>
I assume we're talking about a windows-only issue yet again?
<InPhase>
10 has installed just fine for me locally on Linux.
<Scopeuk>
yes, it looks like on the windows ci build it is failing to find the zlib headers (I suspect if it gets past those it will do the same with libjpeg headers) during a build of the package
<InPhase>
I see the CI is marked as failing, but it seems to fail at the end of tests on windows with "Terminate batch job (Y/N)? ^C Error: The operation was canceled." Like it just froze at the end of the testing after passing everything until it was terminated.
<Scopeuk>
I believe pillow is used as part of the test checks
<InPhase>
Oh, well that's a more troubling sequence of failures.
kintel has joined #openscad
<Scopeuk>
welcome back kintel
<kintel>
I see what's happening now
<InPhase>
Let me try actually running the full sequence of tests locally. You were talking about wheels, so I thought you meant yet another installation issue of the packages.
<kintel>
We install pillow without specifying version
<kintel>
..so it pulls Pillow-10.0.0, which apparently doesn't have Windows wheels (yet?)
<Scopeuk>
and it went from 9.x.x to 10.0.0 3 days ago
<InPhase>
kintel: It really should be pulling the most recent available by platform...
<kintel>
I don't think we can query "most recent by platform though"
<InPhase>
Okay, yeah, locally I still get 100% tests passed, 0 tests failed out of 1616
<InPhase>
kintel: I mean, I thought that was the pip default behavior.
<kintel>
no, pip isn't that smart
<kintel>
-> it will try to build from source if wheels aren't there
<kintel>
..and wheel release comes after package release
<kintel>
I've run into that in the past, and the solution was to tie packages to specific versions, and update only after testing the upgrade on each platform
<Scopeuk>
it should be fairly easy to lock to the last of the 9 series and see if that fixes it (and it should) to confirm and isolate the issue
<InPhase>
kintel: Frustrating. I see people talking about --only-binary --prefer-binary and --use-wheels options for pip... And I cannot find any of these. Are they just randomly adding and removing these preferences?
<kintel>
they ship a bunch of pythons - for some reason
<Scopeuk>
yeh, just found the additional python revisions
<kintel>
Anyway, I think the Windows image upgrade is a red herring, as we've successfully build on that image version just before pillow 10 was released
<kintel>
I gotta run. Will check back later. My hunch is that we're pulling wheels from a mirror which is slow to update, and that the safest fix is to tie pillow to 9.5
kintel has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<teepee>
kintel ^ 2mio € for data protection violation via google analytics
<teepee>
heh, no 1mio, still first big one for that topic the article claims
snaked has joined #openscad
kintel has joined #openscad
<kintel>
ouch, EU is getting serious.
<kintel>
It seems like this is mostly about transferring customer data from EU to US though. Once Analytics offers regional storage, would that change you think?