<ohsix>
is there anyone doing something like training materials or tutorials? the ones on the website are great, more of those
<whitequark>
ohsix: nope
<whitequark>
well, there are some university lectures (?) in chinese (??)
<ohsix>
bummer
<whitequark>
the ones on website cover most of the functionality right now, i think
<ohsix>
i keep telling local people about it but i'm not the best explainer-er
<whitequark>
what language would that be?
<ohsix>
english, i mean people at hacker spaces
<ohsix>
everyone talks about autodesk 360 or whatever, i'm like 'go here, download, model'
<ohsix>
explaining what the constraints are or how to figure out what the mistake you've made is tough
<whitequark>
so here's the problem... i've been using solvespace for a pretty long time now
<whitequark>
I have basically zero insight into newbie mistakes
<ohsix>
ohh
<ohsix>
ok i'm new to all the things, want me to keep notes or something?
<ohsix>
i basically have only modeled that thing i put on twitter, it's supposed to fit into a tripod head
<ohsix>
i think i had one red constraint problem i couldn't figure out, so i just deleted the constraints and did it again
<whitequark>
yup sure, put that on the forum i think?
<whitequark>
and attach the problematic files
<ohsix>
k i'll make a note to do so
<whitequark>
i can't guarantee i'll fix everything since many improvements would take a whole lot of time and/or make things worse for once you've learned it
<whitequark>
but i do make an effort to fix papercuts
<ohsix>
yea i don't think they were bugs or anything
<whitequark>
usability bugs are bugs
<ohsix>
the popups for constraint errors or shape errors are tough too
<ohsix>
noted
<whitequark>
what do you mean by tough?
<whitequark>
the popups themselves are bad? or the errors they represent?
<ohsix>
so rather than being able for people to know about the constraints and having to figure out what the error is, you'd have solvespace be smarter about it?
<ohsix>
they're a messagebox on windows (and i think linux) when i was just starting out i was making lots of mistakes
<whitequark>
well you have to learn how the constraints work to model anything efficiently; this is not really different from any other parametric CAD
<whitequark>
but there are probably discoverability improvements to make
<ohsix>
having them in your face every few seconds while you experiment is a real drag
<travis-ci>
solvespace/solvespace#158 (master - 592bea8 : whitequark): The build passed.
<ohsix>
not a serious suggestion, but something like a little error console might suffice
<whitequark>
ok, you aren't the first person to request that
<whitequark>
what was your GH username?
<ohsix>
if you wanted to print more than errors, maybe printing the dimension change or the arc placement details or whatever for each tool, the thing that would be undone by ^z would be handy too
<ohsix>
and re: parametric cad, true; and i sort of know this short of practical experience from doing a lot of it, but i'm mostly introducing it to people who haven't done much of anything
<whitequark>
re: printing the dimension change or the arc placement details or whatever for each tool
<whitequark>
what do you mean by this?
<whitequark>
the popup already says which dimensions are in conflict
<ohsix>
i'm used to, but also haven't used much autocad; when you use a tool it's modal and there's a prompt, you can click to enter stuff or you can type the dimensions, it shows how the part was added in the little log
<ohsix>
probably not important or needed, but if you were going to have an error log, something like that could be put there too
<ohsix>
ignore that tho, not important :p
<ohsix>
looks like newer autocad doesn't even have that log window anymore
<whitequark>
hrm
<whitequark>
I see what you mean
<ohsix>
i do think i have a file i can shre, it worked out but one of the steps was red and i don't think i figured out why
<ohsix>
if that's a modelling error i could have avoided doing it differently, i dunno; that was the second attempt when i figured out i needed another plane in the middle
<whitequark>
configuration -> draw triangle back faces in red
<ohsix>
oh, yea; faces are backward?
<whitequark>
disables that functionality and the bug with it
<ohsix>
k
<whitequark>
yup
<whitequark>
no idea why
<whitequark>
somethings screwy with the renderer
<ohsix>
i got the gl red book, but only 1.5; didn't have shader-y hw until much later huhuhu
<whitequark>
master will soon go gl2+
<ohsix>
oh?
<whitequark>
there's a patchset in queue...
<ohsix>
hm
<whitequark>
needed for GLES
<ohsix>
k i was boasting to people that it ran on my netbook which is basically sw
<ohsix>
is it on a branch? or where can i get it to test it there
<ohsix>
i don't seriously use it on that but if it can keep working that'd be cool
<whitequark>
possibly via llvmpipe
<whitequark>
not really any idea about perf implications of that...
<ohsix>
it is faster than regular mesa sw rendering
<whitequark>
it is a part of mesa sw rendering
<ohsix>
but i've not used it with shader heavy stuff
<whitequark>
it compiles shaders with optimizations and simd
<ohsix>
ya
<whitequark>
as opposed to just interpreting them
<ohsix>
i'll be checking in a bit, my experience with actually _knowing_ abotu sw rendering on that thing is from before llvmpipe
<ohsix>
it's probably using it now
<whitequark>
is it using any new ubuntu?
<whitequark>
16.04?
<ohsix>
fedora 24
<whitequark>
is that gnome3?
<ohsix>
it can be, and it does sort of ok; but no it's running the mate session
<whitequark>
gnome3 needs compositing already
<ohsix>
literally waiting for it to boot :>
<whitequark>
so in principle we could keep the fallback gl1
<whitequark>
but that's a drag
<whitequark>
I'm thinking I can just say "if you don't have gl2 then use solvespace 2.2"
<whitequark>
gl2 is very common nowadays...
<ohsix>
yea maybe i was speaking a little too casual, it isn't a huge concern; just thought of it when you mentioned it
<ohsix>
wtf heh, touchpad is workign in absolute mode on netbook
<ohsix>
well, glxinfo says ES2, so f it, wonder wtf is going on with the mouse
<ohsix>
why does it become impossible to use a computer when there are two mice, but not two touchpads
<ohsix>
alright, haha; this is a mess. i'll keep some notes about stuff, i'm gonna try and get someone that knows nothing to try and do something, and watch what they do
<whitequark>
alright!
<ohsix>
even withotu the cad experience i know what a constraint is, but that's a concept that's going to have to be introduced in a lot of the cases i'll be involved with
<ohsix>
they're hackers, not necessarily programmers or know much math ro whatever
<GitHub>
[solvespace] jwesthues commented on issue #88: `POINT_N_ROT_TRANS`, `NORMAL_N_ROT`, etc., exactly analogous to linking a SolveSpace model, no? You put the point from the STEP file in `en.numPoint`, and `PointGetNum()` applies the transform. https://github.com/solvespace/solvespace/issues/88#issuecomment-255516223
<GitHub>
[solvespace] whitequark commented on issue #88: STEP is the de-facto interchange standard for things like connectors. Everything in the McMaster-Carr catalogue has a STEP model. And you can extract features from STEP, so you can constrain everything else to the edges and points from STEP.... https://github.com/solvespace/solvespace/issues/88#issuecomment-255528345
<GitHub>
[solvespace] whitequark commented on issue #75: @jwesthues One more point I'm thinking about... with true hierarchical sketches, STEP/DXF linking, etc it seems useful to be able to use a single file for interchange, instead of having a directory mess. We could always just put everything as a nested structure or an inclusion of raw data into the .slvs file, but I had what perhaps is a better idea: make the SolveSpace 3 file format use a zip outer co
<GitHub>
[solvespace] Evil-Spirit commented on issue #88: There are no needs to define every privitive from STL, only used in expressions. We can draw construction over this and use it in assemblies. For top of cylinder you can just draw circle in 3d sketch, and make 3 pt-on circle to align it with cylinder. You will get ability to calculate its radius and get ability to constrain it center anywhere you want. Looking forward, we can create ability to use s
<GitHub>
[solvespace] Evil-Spirit commented on issue #88: There are no needs to define every privitive from STL, only used in constraints. We can draw construction over this and use it in assemblies. For top of cylinder you can just draw circle in 3d sketch, and make 3 pt-on circle to align it with cylinder. You will get ability to calculate its radius and get ability to constrain it center anywhere you want. Looking forward, we can create ability to use s
<GitHub>
[solvespace] Evil-Spirit commented on issue #88: Actually, STL is bad for manufacturing. Nor mesh or b-rep is good for manufaturing output, only csg formats can be good. But still there are no any good csg format... Am I miss something? https://github.com/solvespace/solvespace/issues/88#issuecomment-255530018
<GitHub>
[solvespace] whitequark commented on issue #92: Yes, this has been a problem with debug builds in the past with other functions... interestingly, VS2013 doesn't have that. Ideally this should be fixed by redoing the way the text window works. https://github.com/solvespace/solvespace/issues/92#issuecomment-255538242
<GitHub>
[solvespace] jwesthues commented on issue #75: @whitequark I see the benefit. The rules for when to e.g. use a part from within the packed assembly file vs. from the filesystem are perhaps the tricky part. Use the filesystem if available, otherwise prompt the user to extract? https://github.com/solvespace/solvespace/issues/75#issuecomment-255541646
<GitHub>
[solvespace] whitequark commented on issue #75: @jwesthues So let's say we have an assembly with two parts. In a packed file, this would be represented as the following structure, `foo.slvz` being our zip-container and the only actual file here, and `top.slvs` being "the thing that gets linked when you link the entire packed file":... https://github.com/solvespace/solvespace/issues/75#issuecomment-255542943
<GitHub>
[solvespace] whitequark commented on issue #75: @jwesthues Oh, UI-wise I think it can be done in a rather simple way--add a "Embed / Assemble" command alongside "Link / Assemble". One makes a reference to the original file, whatever that is, the other copies it into the zip container and makes a reference to the copy. https://github.com/solvespace/solvespace/issues/75#issuecomment-255543170
<GitHub>
[solvespace] whitequark commented on issue #75: I think "Save for distribution" could be useful as a separate command, primarily because it should drop the generated meshes, entities, etc for everything but the toplevel assembly, or possibly even for that too. This is of course pending the ability to regenerate deeply.... https://github.com/solvespace/solvespace/issues/75#issuecomment-255543453