<_whitenotifier-3>
[scopehal-apps] azonenberg f3835e9 - FilterGraphEditor: double clicking a node now spawns a dockable properties dialog
<_whitenotifier-3>
[scopehal-apps] azonenberg a90882f - Filter graph editor: always show instrument names in channel title bars iff there's >1 instrument in the session
<_whitenotifier-3>
[scopehal] azonenberg 27ba2ca - Added initial visibility control flag to hide/show channels in the filter graph. Not serialized yet.
<d1b2>
<vipqualitypost> hmmm... i tried to do a clean install of ngscopeclient since it's been a while since i've built it, but i'm getting flagged on missing openmp
<d1b2>
<vipqualitypost> ``` -- Could NOT find OpenMP_C (missing: OpenMP_C_FLAGS OpenMP_C_LIB_NAMES) -- Could NOT find OpenMP_CXX (missing: OpenMP_CXX_FLAGS OpenMP_CXX_LIB_NAMES) -- Could NOT find OpenMP (missing: OpenMP_C_FOUND OpenMP_CXX_FOUND) CMake Error at CMakeLists.txt:148 (message): ngscopeclient requires OpenMP but your C++ compiler does not appear to support it. ````
<d1b2>
<vipqualitypost> i'm not totally sure why it used to be able to find openmp before, though.
<d1b2>
<vipqualitypost> how do you delete nodes from the filter graph editor? selecting and then delete key does not work (the manual does not mention this for deleting nodes, but it does for groups, although it does specifically say that it doesn't delete nodes inside)
<d1b2>
<vipqualitypost> awesome! this seems to have worked pretty well! why is the jitter like crazy if I don't clip the start of the trace? is that the PLL stabilizing? also, should I be thresholding the data and comparing that to the recovered clock, or comparing the raw data to the recovered clock? i would imagine thresholded, right?
<d1b2>
<azonenberg> So right now there is no way to delete nodes from the graph editor. This is due to the fact that right now nodes are reference counted and self-delete when the last reference is removed
<d1b2>
<azonenberg> and we have no way to track where the refs are
<d1b2>
<azonenberg> so there's no way to delete them on request. you have to just find every user of the block (filter input, waveform view, etc) and remove them one by one
<d1b2>
<azonenberg> this will change, but it's a major refactoring to do right
<d1b2>
<azonenberg> There is going to be some jitter at the start of a measurement due to the PLL stabilizing unless you get the nominal frequency of the PLL exactly perfect (seems like you start a little low)
<d1b2>
<azonenberg> There is no need to threshold before feeding the clock to the CDR filter or, i think, the TIE filter
<d1b2>
<azonenberg> you will actually get better results without doing so, since the current threshold filter does not interpolate zero crossings to sub-sample precision while the PLL does
<d1b2>
<azonenberg> But yes, the reason the TIE filter has a parameter to trim the beginning of the waveform off is to get rid of the PLL-stabilizing window and only show you the jitter once you're locked
ALTracer has quit [Read error: Connection reset by peer]
<_whitenotifier-3>
[scopehal-apps] azonenberg opened issue #733: Move SCPI console window to an "advanced" or "debug" menu so it's less likely for new users to stumble into it - https://github.com/ngscopeclient/scopehal-apps/issues/733
<_whitenotifier-3>
[scopehal-apps] azonenberg 8d1a8b4 - WaveformGroup: don't attempt to autoscale on middle click if we have no waveform data in the plot