<GitHub>
[solvespace] whitequark commented on issue #179: > By the way, for this type of debugging, it may be useful to have an option somewhere (could be on the command line), to make SolveSpace draw a cursor (maybe X-shaped), so that it is included in the screenshot.... https://github.com/solvespace/solvespace/issues/179#issuecomment-276836090
<wpwrak>
hmm, xwd here :)
<whitequark>
something newer than 30 years old might help ;p
<wpwrak>
grr. all those forced migrations ...
<whitequark>
or just patch xwd to add the cursor.
<whitequark>
it's really not my responsibility to work around lack of basic features in tools that about 10 people use worldwode
<wpwrak>
sigh. life would be so much easier if everyone could just accept my lead.
<wpwrak>
one thing that is a little unusual on my system is that i have the right mouse button remapped. so if you somehow access the "real button" information (it's possible - there's one game (widelands) that does this), then you could get confusing information
<wpwrak>
btw, clicking the "real" right button does nothing (as expected, since it's remapped)
<whitequark>
I don't think I do anything like that
<whitequark>
if the menu pops it should be fine
<wpwrak>
could it have anything to do with the line getting red (instead of "pale red" (or maybe red + yellow ?)) when i right-click on it, already before releasing the button ?
<whitequark>
red+yellow
<whitequark>
let's see
<whitequark>
ah yes.
<whitequark>
the line stays hovered throughout on my machine.
<whitequark>
so that's what happens.
<whitequark>
you can break on GraphicsWidget::on_button_{press,release}_event
<wpwrak>
that may be the usual crash i get ftieh GDK_DEBUG=ALL
<whitequark>
also crashes on my machine
<wpwrak>
but it happens after i release the right button
<wpwrak>
heh :)
<whitequark>
dunno
<whitequark>
probably a gtk bug
wpwrak has quit [Ping timeout: 264 seconds]
wpwrak has joined #solvespace
<GitHub>
[solvespace] kellpossible commented on issue #178: @Evil-Spirit and yet we would say that NURBS itself is analytic? And the curves themselves are analytic. Perhaps there would be a way to define the intersection path as a function of the two, and evaluate it numerically to get an adaptive analytic approximation of the result? https://github.com/solvespace/solvespace/issues/178#issuecomment-276945489
<GitHub>
[solvespace] kellpossible commented on issue #178: @Evil-Spirit and yet we would say that NURBS itself is analytic? And the curves themselves are analytic. Perhaps there would be a way to define the intersection path as a function of the two, and evaluate it numerically to get an adaptive analytic approximation of the result? https://github.com/solvespace/solvespace/issues/178#issuecomment-276945489
<GitHub>
[solvespace] Evil-Spirit commented on issue #178: @kellpossible, It is common to use rational cubic bezier for both surface and cure representation, and yes, for this case all is analytical and exact. But when we are projecting cubic curve onto cubic surface, we can't get cubic curves anymore. This is the first. The second... We can't even found this intersection in analytical form. But if you doesn't mean cubic on cubic but just project anything o
<GitHub>
[solvespace] Evil-Spirit commented on issue #177: @wpwrak, I have cloned your repo and switched out this revision. But I can't reproduce this. I have no problem for export chord tolerances: 0.1, 0.05, 0.01. Can you switch and check again? Also, what SolveSpace revision do you have? https://github.com/solvespace/solvespace/issues/177#issuecomment-276999373
<GitHub>
[solvespace] Evil-Spirit commented on issue #177: @wpwrak, I have cloned your repo and switched out this revision. But I can't reproduce this. I have no problem for export chord tolerances: 0.1, 0.05, 0.01. Can you switch and check again? Also, what SolveSpace revision do you have? https://github.com/solvespace/solvespace/issues/177#issuecomment-276999373