<teepee>
ah, ok. it sounded like the comment was for openscad as application :)
<teepee>
I'm not sure F5 is not just page refresh?
<J25K70>
just tested with firefox .. seems there F5 is reloding the page
<J25K70>
with chrome those are used internaly in WASM
<teepee>
so either firefox does not allow for f5 to be captured, or it needs a different way to do it
<J25K70>
it also works in chrome not everytime ..
gunnbr has joined #openscad
mmu_man has quit [Ping timeout: 258 seconds]
othx has quit [Remote host closed the connection]
othx has joined #openscad
<pca006132>
btw, regarding the recent mailing list discussions on the object stuff
<pca006132>
I wonder if any of that will be included in the upcoming release, hopefully soon...
mmu_man has joined #openscad
<teepee>
pca006132: my guess would be that object() will be in, all the other stuff is not even merged or even fully specified, so probably no
<pca006132>
I see
<teepee>
that said, I think the "this/self" stuff should be in too, but it looks like there's very different viewpoints on that
<pca006132>
too much for me to read in details
<pca006132>
but tbh, OO related features typically make people very vocal...
<teepee>
yeah, I did not have the time to follow all the stuff :(
<teepee>
well, I think we don't need full OO, and it's not even all that useful
<pca006132>
I think if we want to add OO features to a functional language, we should probably look at how ocaml was designed, and its pitfall
<teepee>
very true, yes
<pca006132>
though a lot of the stuff are related to static typing, dynamic typing makes things a lot more flexible (and chaotic)
<teepee>
my personal takeway of OO so far is, that it's quite ok/useful for GUI libs. for everything else it's just a bad idea
<pca006132>
probably
<pca006132>
I rarely do OO designs, and I rarely write GUI code
<pca006132>
probably something to do with this
mmu_man has quit [Ping timeout: 248 seconds]
J25K70 has quit [Quit: Client closed]
J25K70 has joined #openscad
mmu_man has joined #openscad
JordanBrown has joined #openscad
<JordanBrown>
teepee, pca006132 seems like OO is good for any case where you want to have N similar things share implementation, especially when all A* are similar to one another and different from B*, and AA* are similar to each other and different from AB*, et cetera.
<teepee>
it's not so much the sharing, it's the extremely bad properties in regard to maintainance and extensions
JordanBrown1 has joined #openscad
JordanBrown2 has quit [Ping timeout: 260 seconds]
sculptor has quit [Ping timeout: 260 seconds]
mmu_man has quit [Ping timeout: 248 seconds]
mmu_man has joined #openscad
mmu_man has quit [Ping timeout: 248 seconds]
mmu_man has joined #openscad
JordanBrown1 has quit [Read error: Connection reset by peer]
JordanBrown1 has joined #openscad
UltraGreen has joined #openscad
UltraGreen has quit [Quit: Client closed]
UltraGreen has joined #openscad
<UltraGreen>
Thought I'd share an update on my tactile signs :) Some progress pictures with the lowest being the most recent: https://imgur.com/a/ebndfkT
<JordanBrown1>
Am I reading that right, that you have a positive offset at the bottom, and a negative offset at the top?
UltraGreen has quit [Quit: Client closed]
<JordanBrown1>
(I see that you've quit, but maybe you'll look at the logs...)
<JordanBrown1>
Did you just arrange that the characters were large enough that the negative offsets didn't destroy any features?
<JordanBrown1>
teepee I can't claim any real experience there. For legacy, technical, and social reasons the majority of my work over the last 30 years has been in C (not C++) or in JavaScript or Python written by C programmers.
<JordanBrown1>
I don't have a lot of experience with large-scale OO projects, or with maintaining OO infrastructures with external consumers.
juri_ has quit [Ping timeout: 256 seconds]
<JordanBrown1>
We have some fairly large (tens to small hundreds of KLOC, I think) of JS and Python, but since they were written by people who were steeped in C they don't tend to be all that OO-flavored.
<JordanBrown1>
fairly large +projects+, of course.
UltraGreen has joined #openscad
pie_ has quit [Quit: No Ping reply in 180 seconds.]
TheAssassin has quit [Ping timeout: 272 seconds]
pie_ has joined #openscad
TheAssassin has joined #openscad
<J25K70>
ultragreen ↦ you might read the logs
<UltraGreen>
Yes I have offset 3 at the base. Larger offsets hurt readability. And offset -7 at the top. Smaller negative offsets made Walls vanish for my font
<UltraGreen>
When I change the font size or height I play with the numbers (where I have 3.0 and 10.0 in the formula right now)
<UltraGreen>
Mostly with the 10.0 one though. 3.0 as a maximal offset seems to work well
<J25K70>
what layer height are you printing .. should match your step size in the loop for the extrusion height
<UltraGreen>
Maybe I could do that as optimisation. Currently the step size in the loop is 0.05mm and I have a layer height of 0.1mm
<UltraGreen>
So it worked out well coincidentally I guess
<J25K70>
the loop value from your paste is 0.5
<UltraGreen>
Yes, but I scale the model by 0.1 before printing
<UltraGreen>
But wonky but it felt easier to work like that in OpenSCAD ^^
<UltraGreen>
A bit wonky*
<J25K70>
having smaller steps is not a big problem but if they are not a common factor it can lead to visible irregularities.
<UltraGreen>
Ahh ok. Yeah I guess it was lucky that I picked a multiple then
<J25K70>
why not set the correct scale in scad ..
<J25K70>
guess doesn't matter for your project anymore but if you share your files..
<UltraGreen>
I might publish it on github and clean it up a bit before I do
<UltraGreen>
Right now I'll try to print as many signs as possible before our event starts ^^
<JordanBrown1>
Just watch out for the negative offsets causing pieces of your letters to vanish, making the top be irregular.
<JordanBrown1>
I tend to work with a layer height of 0.3mm, where any irregularity is pretty easily seen and felt.
<JordanBrown1>
0.1mm is pretty but *so* slow.
<J25K70>
just need a printer 3× as fast ..
<JordanBrown1>
yes
<JordanBrown1>
I keep thinking about that, and thinking about a printer 3x as fast, versus one about the same speed but with multiple heads...
<JordanBrown1>
(But if I got a printer 3x as fast, I would rapidly just become less patient.)
juri_ has joined #openscad
snaked has joined #openscad
<J25K70>
not sure about the price but my fav at the moment is bondtech INDX
UltraGreen has quit [Ping timeout: 250 seconds]
<JordanBrown1>
Sounds like magic, but it also sounds like all they are building at the moment is a head, not a full printer.
<J25K70>
bambu doing something like it with HDC
<J25K70>
H2C
mmu_man has quit [Ping timeout: 258 seconds]
<JordanBrown1>
I'm not immediately in the market, so I don't need to think about it *too* deeply.
mmu_man has joined #openscad
<JordanBrown1>
I thought I heard something about Bambu trying to lock people into their filament; that would be a deal-killer.
<JordanBrown1>
Prusa is attractive simply from the "not from China" perspective. And from what I've seen I like their attitude.
<J25K70>
Bambu works most convenient with their filaments - but that is with all manufactures using rfid tags
<J25K70>
maybe the coreONE could be equiped with a bondtech system
<J25K70>
most intriguing to have multiple size nozzles to speed up print time while maintaining small features - slicer didn't support that currently
califax has quit [Remote host closed the connection]
TheAssassin has quit [Remote host closed the connection]
teepee has quit [Remote host closed the connection]
TheAssassin has joined #openscad
califax has joined #openscad
teepee has joined #openscad
<JordanBrown1>
PrusaSlicer has the nozzle diameter as a function of the extruder, so while it wouldn't automatically switch to a smaller nozzle for fine work, you could do it manually.
<J25K70>
that manual switching would mean you need to specify parts but it need to be a function of the surface detail .. or at least that you can say outer shells nozzle1 internal walls nozzle2
<J25K70>
this will come as soon we have printer that switch nozzles automatically.
<JordanBrown1>
Switching toolheads is a lot like switching nozzles. (Though you do have to have duplicated spools of filament.)
<JordanBrown1>
That Snapmaker looks awfully darned cool.
<JordanBrown1>
I should note that I do value "it just works" very highly. I don't need to have the latest and greatest. I want to have something that just works without needing me to mess with it.