01:57 <whitequark> ohsix: noooope

01:57 <whitequark> that's a really hard problem

01:57 <whitequark> and we're Working On It but like I said it's hard

03:20 <Jonimus> whitequark: you would accept patches for some sort of GCODE or APT import?

03:23 <Jonimus> part of why I ask is your camera controls in OGL window are actually usable and so far this is the only FOSS program I've found that fits that bill.

03:24 <Jonimus> that said I guess I should get things building first.

03:26 <whitequark> Jonimus: g-code *import*?

03:26 <whitequark> also, what's APT?

03:27 <Jonimus> APT is what most CAM systems use as an IL/IR for generating G-Code

03:28 <whitequark> most? hm, interesting, I've never heard of APT so far

03:28 <Jonimus> https://en.wikipedia.org/wiki/APT_(programming_language)

03:29 <whitequark> that's a pretty neat system, I actually nearly invented something like that

03:29 <whitequark> yes

03:29 <whitequark> I googled it :)

03:29 <Jonimus> ahh lol

03:29 <Jonimus> yeah IBM is way ahead of you :P

03:29 <whitequark> so my plan for generating gcode from solvespace involved doing HSM toolpaths with libarea

03:29 <whitequark> (does APT even do beziers?)

03:29 <Jonimus> the python postp project can already do APT to G-Code.

03:29 <whitequark> but anyway, you've said "import"

03:30 <Jonimus> I said import because I want to build a viewer that doesn't stink.

03:30 <whitequark> have you seen camotics?

03:31 <whitequark> this one: http://www.camotics.org/

03:31 <Jonimus> yeah googling.

03:31 <whitequark> I wrote a bunch of g-code manually and found it instrumental in debugging

03:32 <Jonimus> I'll have to look at that

03:32 <whitequark> I think G-code import into solvespace would be a really strange feature

03:32 <Jonimus> fair enough

03:32 <whitequark> I can't imagine how to integrate it well with the rest

03:33 <whitequark> maybe you could just grab the camera controls code...

03:33 <whitequark> (then put it into camotics? though camotics' camera control is already ok ime)

03:34 <Jonimus> I was also thinking that, but I thought writing it as an importer would be a good first attempt because my C is super rusty, I spend most of my time with VB.NET (eww) C# and Python.

03:34 <whitequark> solvespace is in C++, camotics too

03:35 <Jonimus> yeah I meants C/C++, though my only C++ experiences were a few MAME/MESS patches

03:36 <Jonimus> Oh and just to be clear on the APT thing, CAM packages generally use a subset of APT often called APT CLDATA or similar.

03:36 <whitequark> I see

03:39 <Jonimus> its also ISO 4343-2000 but of course you can't find copies of 16 year old standards easily.

03:40 <whitequark> I wish there was sci-hub for ISO

03:40 <Jonimus> yeah

03:40 <whitequark> is there institutional ISO access?

03:41 <Jonimus> not sure, IEEE or ACM might give you some access.

03:44 <Jonimus> I saw IGES export was on your list, are there no FOSS IGES writing projects out there, I was rather surprised that you had STEP and not IGES export.

03:45 <whitequark> solvespace uses a homegrown STEP export library

03:45 <whitequark> used to be a homegrown DXF export library too

03:45 <Jonimus> ahh nice, IGES being the older standard is usually what I see being included

03:53 <whitequark> we should switch to a proper step library for import and export

03:53 <whitequark> (that is not opencascade, which should be obvious if you ever touched opencascade)

03:59 <Jonimus> I've heard things.

04:02 <Jonimus> I wonder if the BRL-CAD project would have anything there.

04:02 <whitequark> I've looked into it

04:02 <Jonimus> really solvespace's 2D with BRL-CAD's 3D tools might be pretty slick

04:02 <whitequark> well, I was specifically looking at NURBS boolean implementation

04:02 <whitequark> which in solvespace, well, sucks a bit

04:03 <whitequark> but unfortunately I don't think integration is viable (I don't remember the exact details)

04:03 <whitequark> the STEP/IGES export/import shouldn't require changes this drastic

04:09 <Jonimus> the main question there would likely be license, I'm not sure what BRL-CAD uses

04:12 <ohsix> whitequark: cool (working on it)

04:13 <ohsix> stuff is happening pretty fast, rebuilt from master recently; little nice things

04:15 <Jonimus> I forgot home complete yet still utterly unusable brl-cad was, its UI/UX is soo terrible.

04:16 <Jonimus> I mean I know I'm coming from a solidworks background but still this is bad.

04:25 <whitequark> \o/

04:30 <ohsix> doesn't brl-cad have parts that go back to the 60s?

04:30 <ohsix> i tried to use it a long time ago, should try again

04:34 <Jonimus> yeah it does, it reminds me of some of the internals of one of the CAD packages I use at work, except the there is a UI for all the command line stuff, where brl-cad the UI is basically useless.

04:37 <ohsix> https://brlcad.org/gallery/var/albums/historic/Mike_Muuss.jpg

04:50 <Jonimus> not quite as old but a good story https://web-beta.archive.org/web/20081012234622/www.powerinspect.com/general/boris.asp

20:53 <ohsix> http://imgur.com/a/MsnXp any idea how i might model the gentle curve on this in solvespace?

20:55 <ohsix> it's close to a circle with the radius of 6- 3/4 inch

20:55 <ohsix> from trying to do the screw stuff the other day, the lathe tool doesn't take an angle to rotate around

20:55 <ohsix> that's normally how i'd do it, then add another section with the band around the end

20:58 <ohsix> lul playing with slic3r cuz it's about time to take the training wheels off, set the speed to max the lulzbot 5 is specified for, it's going like crazy; print looks like it will be crap but it is small and doesn't matter ;]

21:05 <whitequark> take a photo orthographically

21:05 <whitequark> then trace with a bezier

21:05 <whitequark> then extrude a few times

21:08 <ohsix> slic3r is not stingy with the filament

21:08 <L29Ah> i'd have fixed this with dichloromethane

21:09 <whitequark> *squints at dichloromethane* you're from ex-USSR aren't you

21:09 <L29Ah> thanks for your vogeeky efforts btw :*

21:10 <whitequark> ah.

21:10 <whitequark> well that pins it down a bit

21:10 <L29Ah> i don't really understand where do you get the free time to work on all those projects

21:11 <ohsix> whitequark: so from the side profile?

21:11 <whitequark> ohsix: ye

21:11 <ohsix> k, today is 3d printer day, i needed to print a spacer for something and 'im still doing that; headphone repair parts are next :p

21:11 <L29Ah> also this looks fixable with a healthy band of fiberglass fabric and epoxy :3

22:04 <wpwrak> solvespace + 3d printer is really a great combination. there are so many little things can make life easier, but it's not worth spending a lot of time on. with slvs, you have a testworthy geometry within an hour, and a 3d printer is much quicker so set up than, say, a mill (and more home-friendly, too)

22:05 <whitequark> *grumble grumble*

22:05 <whitequark> well I should take care of work-related things before setting that up I guess..

22:06 <wpwrak> ah, so you still have that hair spray on the shopping list ? :)

22:06 <whitequark> yes

22:07 <whitequark> been a bit busy. shopping, groceries, having an adverse drug reaction, replying to email

22:08 <wpwrak> ah, speaking of reaction ... it's best to keep the printer in another room. it's a little smelly, and that can be nauseating, especially if you sit right next to it all the time

22:08 <L29Ah> wpwrak: did you try openscad?

22:09 <whitequark> we have active ventilation

22:09 <el3phanten> hello

22:09 <whitequark> (an inline duct fan plus some flexible ductwork does *wonders* for things like dremel work, painting, soaking glass in piranha, smokers in the room)

22:09 <whitequark> el3phanten: hi

22:10 <el3phanten> whitequark: hi, I just started solvespace. I have been using freecad for a while

22:11 <wpwrak> L29Ah: very briefly. i was disappointed by its 3D engines. e.g., when putting parts together, i had to add small overlaps or it wouldn't merge common surfaces

22:12 <wpwrak> L29Ah: never had such issues with systems based on opencascade. solvespace is fragile with NURBS, but it's okay with meshes. eventually starts complaining about naked edges, there too, but that doesn't seem to cause any ill effects

22:14 <wpwrak> whitequark: heh, you're prepared for everything then :)

22:20 <el3phanten> What annoyed me in freecad, was when I have made an construction of some kind, the workshop ofc. needs dwg or dfx.

22:20 <el3phanten> but solvespace seems to handle this pretty well

22:21 <el3phanten> and easy to draw with.. Better than freecads sketcher, if you ask me

22:21 <el3phanten> freecad has some nice features tho

22:34 <whitequark> yeah, solvespace has no lofting or filleting

23:03 <dx> hey, what's up with the CLA thing? saw comments about it in github but no details or issues anywhere as far as i can see

23:16 <whitequark> dx: solvespace is licensed commercially (with half of profit going to the original dev and the other half to the salaries)

23:17 <whitequark> we need to be able to do that

23:17 <whitequark> for now, only the solver was licensed, and it's in a separate branch that doesn't have any doubtfully licensed code

23:17 <whitequark> the master branch does

23:17 <whitequark> that should be fixed, but in the meantime I'm trying to avoid making the problem worse

23:23 <dx> huh, sounds complicated. so if i'm interpreting this right you're aiming for dual licensing with GPLv3 and something else, and you need people to hand over their copyright?

23:25 <dx> 'hand over their copyright' is probably grossly inaccurate, i never understood that part of CLAs

23:40 <whitequark> this is the 10,000ft view of it

23:54 <dx> i'm just curious on how you'll get that sorted. good luck and try to not let lawyers suck too much of your blood

23:55 <whitequark> CLAs

23:55 <whitequark> like, they're fairly widely used

23:55 <whitequark> almost any large FOSS company requires a CLA

23:55 <whitequark> the question is not of "how" but of investing the time needed to fix this

23:57 <el3phanten> whitequark: could I write a plugin for solvespace?

23:58 <el3phanten> like fillet plugin for example. Would that be difficoult?