whitequark changed the topic of #solvespace to: SolveSpace--parametric 2d/3d CAD · latest version 2.3 · http://solvespace.com · code at https://github.com/solvespace/solvespace · logs at https://irclog.whitequark.org/solvespace
<GitHub> [solvespace] Evil-Spirit commented on issue #163: @whitequark, did you mean something like this?... https://github.com/solvespace/solvespace/issues/163#issuecomment-273965423
<GitHub> [solvespace] Evil-Spirit commented on issue #161: @whitequark, updated here... https://github.com/solvespace/solvespace/issues/161#issuecomment-273966738
<GitHub> [solvespace] Evil-Spirit commented on issue #162: @whitequark, fixed here https://github.com/solvespace/solvespace/issues/162#issuecomment-273966813
balrog has quit [Ping timeout: 256 seconds]
balrog has joined #solvespace
whitequa3k has joined #solvespace
whitequark has quit [Remote host closed the connection]
<GitHub> [solvespace] whitequark commented on issue #163: @Evil-Spirit yes https://github.com/solvespace/solvespace/issues/163#issuecomment-274050341
whitequa3k is now known as whitequark
<GitHub> [solvespace] whitequark pushed 1 new commit to master: https://github.com/solvespace/solvespace/commit/86f229c5e6e1630ef1946933c6bea20df52bc7da
<GitHub> solvespace/master 86f229c EvilSpirit: Render points corresponding to a DOF in the front layer....
<GitHub> [solvespace] whitequark closed issue #163: "Show Degrees of Freedom" highlights should be drawn over the mesh https://github.com/solvespace/solvespace/issues/163
<GitHub> [solvespace] whitequark commented on issue #161: @Evil-Spirit But not the first bug, no? https://github.com/solvespace/solvespace/issues/161#issuecomment-274054258
<GitHub> [solvespace] Evil-Spirit commented on issue #161: @whitequark, not yet. https://github.com/solvespace/solvespace/issues/161#issuecomment-274054575
<GitHub> [solvespace] whitequark pushed 2 new commits to master: https://github.com/solvespace/solvespace/compare/86f229c5e6e1...669e6f0dcbc2
<GitHub> solvespace/master 669e6f0 EvilSpirit: Select the appropriate entity when dragging a circle or a normal....
<GitHub> solvespace/master 119b17a EvilSpirit: Fix an occasional crash when starting to drag an entity.
<GitHub> [solvespace] whitequark closed issue #162: selection of entity from request for dragging doesn't work for circles https://github.com/solvespace/solvespace/issues/162
<travis-ci> solvespace/solvespace#334 (master - 86f229c : EvilSpirit): The build passed.
<travis-ci> solvespace/solvespace#335 (master - 669e6f0 : EvilSpirit): The build passed.
<wpwrak> is it possible to move items from one sketch to another ?
<whitequark> wpwrak: nope
<wpwrak> ah, found it: copy and paste :)
<wpwrak> but it doesn't preserve constraints :( (not even within the same sketch)
<whitequark> wpwrak: um, of course it does!
<whitequark> try copying and pasting a rect
<whitequark> the thing is it only preserves constraints that are fully within the copied stuff
<wpwrak> i copied and pasted a rect with Q constraints (to another rect)
<whitequark> yes. that's not copied.
<wpwrak> inconvenient for, say, pins of electronical components
<wpwrak> at least if you don't want to clutter the sketch with Ds
<wpwrak> not a bad collection for about a day of work: https://gitlab.com/anelok/mexp/tree/master/alt13/
<wpwrak> and the ability to follow constraints is madly useful
<wpwrak> btw, one day i'll have to round some corners in a fairly complex geometry. are there any plans for a generalized fillet or chamfer operation ?
<wpwrak> oh, and lathes are always full circles, are they ? i.e., one can't lathe an arc
<whitequark> (fillet or chamfer) the current NURBS backend doesn't support it...
<whitequark> and no, you can't lathe an arc. this would probably go towards lofting...
<wpwrak> yes, extrusion along a path would be very nice to have
<whitequark> it's not trivial
<wpwrak> the harder the work, the more glorious its completion :)
<whitequark> not particularly tbqh
<wpwrak> well, i'd say that solvespace has a lot of potential. and such a feature would help quite a bit towards making it usable for more designs.
<wpwrak> and more projects that use it, more glory :)
<whitequark> that's true but someone has to pay for it
<whitequark> it's not cheap maintaining a full-time dev for such a project
<whitequark> for now the license fees do it...
<wpwrak> license fees ?
<whitequark> solvespace is licensed for commercial use
<whitequark> see the bottom of http://solvespace.com/library.pl.
<whitequark> I plan to clean up the source of the main app (the solver is already clean) copyright-wise and start requiring a CLA from contributors
<whitequark> with a clause that prevents us from 'going dark', similar to what the Discourse CLA has
<wpwrak> aah, i see. so people are using it in their own products. interesting.
<whitequark> yeah, it's surprisingly popular
<whitequark> or at least, more popular than I expected. I of course cannot disclose details
<wpwrak> ;-) let hope this doesn't create a conflict of interest, i.e., resulting in features not getting added to avoid competing with a paying customer
<whitequark> actually, so far it's been exactly the opposite, features are added to the GPL version at request of customers
<wpwrak> wonderful :)
<whitequark> in general I simply don't have the bandwidth to maintain two separate codebases. I'm like, one person
<whitequark> it's not realistic
<wpwrak> yeah, such split development would be a pain
<wpwrak> you should get some folks at google, mozilla, or such to find it useful, and get them to sponsor your work long-term :)
<whitequark> M-Labs does help out a bit
<whitequark> my own wallet does help out a bit, to some dismay
* whitequark shrugs
<whitequark> CERN did express some interest but it never went anywhere
<wpwrak> (own wallet) hmm yes, circuilar dependency :)
<wpwrak> ah, pity
<whitequark> should ping them again...
<wpwrak> btw, regarding assembly, is my understanding correct that, given a design that consists of a number of groups in a single .slsv file, one could basically also put every group or group+extrusion pair into its own .slvs file and assemble them, and things would basically work the same ? or are there limitations ?
<whitequark> uhmmm, to an extent the same
<whitequark> obviously the imported geometry is "fixed"
<whitequark> so you can't easily edit it, you have to re-open
<wpwrak> yes, the workflow gets a bit inconvenient. though i guess you could fix that :)
<d42> i kind of like the multi solvespace workflow it forces :^)
<wpwrak> d42: i can imagine that a workstation setup like in the "swordfish" movie would become handy at some point ;-)
<d42> yup :^)
<wpwrak> whitequark: e.g., with a quick "open for editing" step, an undoable auto-save for the current slvs, and maybe a fast "return to current end" function. so that one doesn't have to go to "open recent" all the time
<whitequark> nope
<whitequark> what should be done is true hierarchical editing, with multiple sketches loaded at the same time
<whitequark> this requires a LOT of refactoring though
<wpwrak> yeah, i'm suggesting low-hanging fruits that are almost as tasty :) plus, it would probably be easy to migrate designs from what i described to fully hierarchical, once implemented
<whitequark> not really, rather this encourages cluttering up the codebase even more
<whitequark> if a change doesn't make solvespace at least not worse off, I don't do it
<whitequark> this is the only way I've been able to keep it actually standing
<wpwrak> heh :)
<whitequark> I understand the frustration, but consider freecad
<whitequark> you know, the CAD with three assembly workbenches, none of which actually work, and that crashes and corrupts your savefiles every few minutes
<whitequark> and a really bad clone of every worst choice in CATIA's UI
<whitequark> I would like to not become freecad. This is basically my #1 goal here
<wpwrak> yeah, that gazillion workbenches is rather irritating
<whitequark> what about STEP import btw? this is relatively low-hanging, surprisingly
<whitequark> stuff from mcmastercarr...
<wpwrak> i've never used STEP
<whitequark> ... because it's not supported properly almost anywhere? :P
<whitequark> btw, kicad has STEP export of boards with components
<wpwrak> (not supported) could be :) but for most things i do, STL works fine.
<whitequark> but with STL there will be no entities to constrain against
<whitequark> whereas with STEP I can import them almost as easily as from an SLVS file
<wpwrak> i guess you could reconstruct some edges and such in STL. anyway, my projects tend to have one "central" CAD system at a time, and the rest of the tools involved only do relatively dump processing
<whitequark> (STL) nnnnot reliably at all
<whitequark> and that would break after updates anyway
<wpwrak> so there usually isn't a scenario where one would pass such diverse information around
<whitequark> whereas STEP has relatively consistent numbering
nathan7 has quit [Ping timeout: 260 seconds]
nathan7 has joined #solvespace
<ohsix> halp build error!!11
<ohsix> make[2]: *** No rule to make target '../res/locales/en_US.po', needed by 'res/locales/en_US.po'. Stop.
<ohsix> don't know enough about cmake and can't get into it now, installed all the junk from the readme
<whitequark> git pull?
<whitequark> like, why is there no ../res/locales/en_US.po ?
<ohsix> yea i just pulled, that's why i'm building it, let me clean and stuff and check it out
<ohsix> f it nevermind it worked after git clean -dfx
<ohsix> something something autotools gets this right something something ;]
<whitequark> wat?
<whitequark> autotools doesn't even run autoreconf without --enable-maintainer-mode
<whitequark> and/or doesn't let you build anything at all without either a) checking in a shitload of stuff into the repo b) requiring everyone to install your exact version of some shitty m4 library no one knows what's even for
<ohsix> heh
<ohsix> --missing will add stuff from the host when you start using the tree (with an ./autogen.sh or similar)
<ohsix> arguably having some of those files in the repo is wrong, but there are exceptions
<ohsix> hrm there's something up with subpixel text stuff
<whitequark> грь
<whitequark> *uhm
<whitequark> there's no subpixel AA in solvespace anywhere because I can't stand subpixel AA
<ohsix> yea it's on lines hang on
<ohsix> probably render target stuff
<ohsix> i'm sort of is/was in a meeting heheheh, and people are still in here
<ohsix> was going to passively show off solvespace