<GitHub>
[solvespace] whitequark commented on issue #91: @wpwrak I don't really see how this would help. Especially, it is not an improvement over, in your case, clicking on the constraint and seeing what exactly is it attached to--which I believe you can now do with the "explorability" patch. https://github.com/solvespace/solvespace/issues/91#issuecomment-276286189
<GitHub>
[solvespace] wpwrak commented on issue #91: The explorability SolveSpace has at the moment is great, and I expect it to be even greater with Evil-Spirit's pending further enhancement, but the issue is often that I'm not sure if SolveSpace even performed the operation. I.e., sometimes there is no visual feedback, or maybe there is, but not where or how I expect it. So a quick way to check what happened last should be useful. https://github.com/s
<GitHub>
[solvespace] wpwrak commented on issue #176: But why not place it where I asked, instead of making me search for where it may have ended up ? Especially if there are a lot of points (like in the example above), that can get quite tricky. Given that SolveSpace recognized the automatic point-on-line constraint, there is a valid position for that point in the immediate vicinity of the click location. https://github.com/solvespace/solvespace/issues
<GitHub>
[solvespace] Evil-Spirit commented on issue #176: @whitequark, We need to implement ModifyToSatisfy for pt-on-line and call it when we are creating like for distances (since pt-on-line has hidden parameter which needs to be guessed when we just create constraint) https://github.com/solvespace/solvespace/issues/176#issuecomment-276295408
<wpwrak>
funny. sometimes, when activating a workplane, then invoking Sketch > In Workplane, 1) produces a complaint that no workplane is selected, 2) activates the desired workplane
<wpwrak>
alas, so far i hit this only when deep in an-top, so i have no easy way to reproduce it (yet)
<whitequark>
activating a workplane?
<whitequark>
you can activate a group. or you can select a workplane
<wpwrak>
from the fine manual: "To activate a workplane, select it, and then choose Sketch → In Workplane." ;-)
<wpwrak>
i would have used "select", too, but then it would get a little ambiguous
<whitequark>
oh, hm.
<whitequark>
I'm confused by the way you've phrased your first message
<whitequark>
are you activating a workplane twice?
<wpwrak>
no, i previously "lost" it (i end up with "anywhere in 3d" for some reason, which slvs eventually complains about when i try to place a H or V). then i switch back to the workplane. and then this oddity happens.
<whitequark>
ah, I think I see what happens
<whitequark>
so "Sketch -> In Workplane" wants *some* workplane. if you have one selected, it'll switch to that. if you don't but the current group is a sketch-in-plane group, it'll complain and then switch to the group's workplane
<whitequark>
you can reproduce the complaint every time. just press 2 without selecting anything, in a new file.
<whitequark>
it'll complain even if you are already in workplane, and even if you're switching to the same one.
<wpwrak>
hmm. but i did select it. it's all nice and red.
<whitequark>
if you've selected a workplane *and* something else, it doesn't count
<wpwrak>
yeah. maybe it was that. i think it wasn't, but it's something that's easy to miss. i'll check that when it happens again
<wpwrak>
may have been multiple selection. just "lost" the workplane again, and the workflow leading to it makes multiple selection likely: 1) select items for H/V, 2) press H/V, 3) slvs complains about being lost in 3D space, 4) i select the workplane (without realizing the points for H/V are still selected), 5) as you described above
<wpwrak>
yeah, must have been that. sorry for the false alarm.
<wpwrak>
hmm, analysis says all is fine, exporting STL produces naked edges :-(
<whitequark>
different chord tolerance for analysis (which uses display tolerance) and export
<wpwrak>
hmm, they're different indeed
<wpwrak>
alas, the only way to get rid of the naked edges seems to be to keep the chord tolerance high :-(
<wpwrak>
slvs seems to be really really really reluctant to let me have smooth curves :)
<whitequark>
wpwrak: generally, this is a known problem...
<whitequark>
... but somehow you've managed to trip on it even more than usual
<cr1901_modern>
wpwrak seems to be good at finding edge cases :D
<wpwrak>
maybe anelok is more "curvy" than the usual design. of course, slvs makes is rather tempting to do smooth curves all over the place. when scripting in freecad, and often enough having to calculate the trigonometry myself, i didn't do half of them.