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
<cr1901_modern> whitequark: When you get the chance, could you please test the three.js viewer in a browser on a _non mobile device_, and tell me if panning works?
<whitequark> the one in the master branch?
<whitequark> yes, panning, rotation and scaling works
<whitequark> via mouse
<whitequark> via touch, only rotation and scaling works
<whitequark> but it's all screwed up if devicePixelRatio != 1
<cr1901_modern> whitequark: Ack, I was planning on looking at the devicePixelRatio tonight, but something on my end is screwy; panning isn't working at all.
<cr1901_modern> Great! So why the hell did it ever work at all in the first place o.0;?!
<cr1901_modern> Bug is this line: https://github.com/solvespace/solvespace/blob/master/res/threejs/SolveSpaceControls.js#L198 According to W3C spec, mousemove returns "0" unconditionally in the "button" field.
<cr1901_modern> whitequark: This is a stretch, but what Chromium version are you using?
<cr1901_modern> I wonder if mousemove was implemented improperly in older Chrome versions and was fixed recently...
<whitequark> 59.0.3071.86
<cr1901_modern> I'm using version 60, and right-click, which should _pan_ the model, is _rotating_ the model in my browser, due to aforementioned bug. Wonder what's different about Chromium 60.0.3112.101 on Linux
<cr1901_modern> But I can't do anything about it tonight. I'm in no frame of mind to work on anything
* cr1901_modern takes a snooze
Thaolia has quit [*.net *.split]
Thaolia has joined #solvespace
f8l has quit [Remote host closed the connection]
f8l has joined #solvespace
tristero has quit [Quit: WeeChat 1.9]
<cr1901_modern> Of course I still can't build solvespace, not even cross compiling on Linux. Ultra-shiny new compiler version required for goddamn ID3D11Debug. Why is this so hard?
<cr1901_modern> whitequark: What's your mingw version? I've upgraded mingw-w64-gcc on ubuntu to the one provided by artful, and angle still refuses to compile "ID3D11Debug" does not exist. I simply cannot build solvespace for Windows with the resources I have at home.
<cr1901_modern> whitequark: Ignore previous q please
<cr1901_modern> whitequark: Changed my mind again: by any change what version of mingw distro do you use (mxe? debian packages?). Problem is bigger than ID3D11Debug not existing
<GitHub> [solvespace] cr1901 opened pull request #284: Update README.md with caveats building Solvespace on 16.04. (master...cr1901-patch-1) https://github.com/solvespace/solvespace/pull/284
<whitequark> cr1901_modern: -DOPENGL=1
<whitequark> and then build with mingw
<whitequark> or even vs
<cr1901_modern> whitequark: You said the VS build was broken and pointed me to the issue where it was broken. And anyways, I got it to compile; see pull request
<cr1901_modern> whitequark: I'll try OPENGL=1 in a bit
<whitequark> cr1901_modern: oh
<whitequark> I should just apply the patch from Qt
<cr1901_modern> whitequark: I agree that is preferable to my method, but it applying that patch as-is didn't fix my issues (at least, a bunch of D3D11 typedefs were still missing, and versionheaders.h was completely nonexistent).
<whitequark> that's because my ANGLE is very old
<whitequark> I need to look up whatever version of ANGLE Qt used
<cr1901_modern> Alright, cool. I will also test a native Windoze build at that point.
<whitequark> why can't you just run it on linux if you're building it on linux?
<cr1901_modern> B/c headless :P
<whitequark> or build it on visual studio?
<cr1901_modern> You said VS build was broken
<whitequark> there's a one-line patch in the issue I gave you
<cr1901_modern> Oh...
<whitequark> it's not applied because [reasons]
<cr1901_modern> I... think I totally missed that ._.
<cr1901_modern> So I went through all this trouble b/c I didn't read closely enough... go figure
<cr1901_modern> I'm fixing js issues here: https://github.com/cr1901/solvespace-1/tree/js-fixes Currently, I fixed the bug I noticed yesterday with mousemove. Would you be willing to test?
<cr1901_modern> I'd be very surprised if it didn't work, just wanted to check
<whitequark> can you upload the html somewhere?
<cr1901_modern> Doing now