14:27 <wpwrak> whitequark: i'm feeling creative today :) here's an idea for 3D construction. it's for modularity problems related to making holes. example:

14:28 <wpwrak> you have a screw, basically two cylinders in the simplest case. you can make this a building block easily enough. now you want to define the hole for it, too. normally, you would

14:29 <wpwrak> 1) create the hole shape (pair of cylinders or such), 2) subtract it from what the screws is supposed to go into, 3) create the screw, 4) add screw to the result from 2)

14:31 <wpwrak> now the new idea would be to merge 1-2 and 2-3, by introducing "anti-matter". so the screw "module" would be surrounded by a thin layer of AM. when you add it to your construction, AM makes a hole

14:33 <wpwrak> does this sound like fun enough ? :) so in stead of needing two parts (screw and anti-screw), you need just one. even better, if operations preserve the AM property, you can add more things to the model without having to worry about anything blocking that hole

14:34 <wpwrak> so union would work something like this: M = matter, A = antimatter, 0 = empty space. M + M = M, M + 0 = M, 0 + 0 = 0, A + anything = A. needs a new operation to clean out antimatter, so that you can, say, make negatives, e.g., for molds.

14:45 <whitequark> yeah, I do see your point

14:45 <whitequark> this is somewhat similar to "keepoff" in ECAD, right?

14:46 <whitequark> DocScrutinizer05: remember our earlier automated tea making discussion? it seems wildly popular among my twitter followers, *especially* british

14:47 <whitequark> someone brought up https://en.wikipedia.org/wiki/Teasmade, which is a tad too simplistic sadly, and also seems to be wholly out of fashion

14:51 <wpwrak> dunno ecad :)

14:54 <whitequark> electronic

14:54 <whitequark> eagle etc

14:54 <whitequark> as opposed to mcad, mechanical

14:55 <wpwrak> aah. well, kinda. that is, if you interpret DRC trying to keep things out as a boolean operation :)

14:56 <whitequark> oh I was thinking of polygons

14:56 <whitequark> you can define keepoff areas inside components and those will subtract from polygons

14:56 <whitequark> seems quite similar to me

14:57 <wpwrak> well, sort of. but you don't merge components, etc. the effect on merging (union) would be the most interesting feature of AM

14:57 <whitequark> one problem I immediately see is that the drill thickness for threading is generally smaller than the screw size

14:57 <whitequark> moreover, it differs with material in which the drilling is to be done

14:58 <whitequark> e.g. if you have M3 then you drill to 2.5 in SS or 2.75 in Al (iirc)

14:58 <whitequark> er... no, those diameters aren't right. but the concept is. I had to deal with that just a few weeks ago

14:59 <whitequark> this also doesn't work in cases where for example you want a hole to go through the entire thickness whereas a screw is shorter

14:59 <wpwrak> hmm yes, maybe the ability to combine things in the same module isn't the best use

15:00 <whitequark> what I had in mind for this use case is a "drill" operation that extracts a cylinder+cone shape from the base material in one go

15:00 <whitequark> and maybe another "tap" operation that does that and adds thread metadata

15:01 <wpwrak> of course, modules should probably be parametrized anyway. so material dependency and such could be handled at that level. of course, it would be even cooler if the module could "sense" what it's drilling into, but that may be overkill :) (nice feature of that: you could make it adapt if you have a sandwich of different materials :)

15:01 <whitequark> uhhhhhh

15:01 <wpwrak> (-:C

15:01 <whitequark> I do see what you mean, but I'll need to see many more compelling cases for that

15:02 <whitequark> modules *should* be parameterizable, that's coming in SolveSpace as soon as I clean out the last remnants of the horrible C-style global state

15:03 <wpwrak> or go all the way and make an "attach wizard" :) wizzard(A + B) -> finds best way to attach (screws, snap, live hinge, whatever), adds that, adjusts rest of design accordingly, preserving general style :)

15:03 <whitequark> going to be the first major killer feature over freecad, which only supports that by generating models with crappy python

15:03 <whitequark> as for the "wizards" you'll be able to do it using the internal scripting engine

15:03 <whitequark> that will have access to the complete sketch and the internal UI

15:04 <whitequark> you only need to add logic ;p

15:05 <wpwrak> (drill operation) one way to get properties similar to the AM i proposed would be to have some tag for the "drill" volume that could be used to check at the end whether the volume is being invaded by something

15:06 <whitequark> what do you mean?

15:06 <whitequark> clash analysis? it's a thing in large MCADs

15:21 <roh>

15:26 <wpwrak> crash ... but matter vs. "antimatter". i.e., you'd make the screw hole and declare "this hole must always be empty". then you merrily construct the rest of your critter. and at the end you check that all holes of this sort are indeed empty.

15:36 <whitequark> I think this is just regular clash analysis of the complete assembly with all screws

15:36 <whitequark> roh: empty message?

15:36 <whitequark> or some char that doesn't render over here?

15:40 <roh> space when attaching screen anew ;)

15:43 <DocScrutinizer05> whitequark: (teasmade) hehe

15:45 <whitequark> very clever passive "pump" using steam pressure to push water out of teapot

15:45 <DocScrutinizer05> I had this one https://www.willhaben.at/iad/kaufen-und-verkaufen/d/krups-teemaschine-teetime-top-zustand-157946980/ and it worked quite nicely, but of course you needed to clean and fill it after each turn

15:45 <DocScrutinizer05> yeah, this one did same. I think I described it back when we discussed how to build it

15:46 <whitequark> oh, if so I haven't understood you back then

15:47 <DocScrutinizer05> http://www.teemaschinendoktor.de/Alternativen/Krups/krups.html

15:47 <DocScrutinizer05> the water sits in a tank with lid above the tea and has an upside-down U pipe

15:48 <DocScrutinizer05> with a little hole in the top of the bending, to allow smal amounts of expanding air to go out without pressing water out

15:49 <DocScrutinizer05> when water boils, it gets pressed through the U pipe by water vapor pressure

15:49 <DocScrutinizer05> when the tank is empty, the temperature of heating rises >100°C and thus heating off and teatimer start

15:50 <whitequark> oh i see

15:50 <DocScrutinizer05> after timer expires, a magnet opens a valve at bottom of the tea tank

15:58 <DocScrutinizer05> some more inspirations: https://www.otto.de/p/grundig-gourmet-teebereiter-tm-8280-w-white-sense-1-7-liter-1750-watt-weiss-edelstahl-510022859/#variationId=510022959

15:59 <DocScrutinizer05> this one is still available it seems

15:59 <DocScrutinizer05> costs more than gold

16:00 <DocScrutinizer05> but it seems you can adjust the water temperature, so it obviously has a pump to move the hot water into the tea tank

16:00 <DocScrutinizer05> hot = not boiling

17:00 <whitequark> 249eur isn't even that much

17:02 <whitequark> compared to what I think my device would cost

17:02 <whitequark> well, there's no way I can give an estimate within a kilometer of the real cost but it's a manufacturing nightmare for one

20:27 <wpwrak> fun discovery: http://www.sherpals.com/

22:35 <DocScrutinizer05> wpwrak: you invented that? why not?