azonenberg changed the topic of #scopehal to: libscopehal, libscopeprotocols, and glscopeclient development and testing | https://github.com/azonenberg/scopehal-cmake, https://github.com/azonenberg/scopehal-apps, https://github.com/azonenberg/scopehal | Logs: https://freenode.irclog.whitequark.org/scopehal
<azonenberg> bluezinc_: figured it out?
<bluezinc_> azonenberg: I've vaguely figured out cmake now
<bluezinc_> now things are failing for even more esoteric reasons
<bluezinc_> azonenberg: is getting Render error 506 a common issue?
<bluezinc_> looks like there's some incomplete/invalid FBO, which leads to a GL_INVALID_OPERATION, and then a bunch of GL_INVALID_FRAMEBUFFER_OPERATION
<azonenberg> There is a known bug that i have not yet tracked down, which occurs when you move a waveform to a new viewport at unpredictable intervals
<azonenberg> is that what you were doing?
<azonenberg> it happens maybe one in ten times i do it and i havent yet found a reproducible trigger or cause
<azonenberg> In other news, i experiented a bit with the carbon paint
<azonenberg> it can be thinned with acetone, but getting reproducible thickness is nontrivial
<azonenberg> if you put it on full strength with a swab it tends to cake and crack
<azonenberg> It was just kind of a side experiment, the new pcb at fab is the main line of research re probes right now
<bluezinc_> carbon film isn't fun to work with (even for the professional board shops)
<bluezinc_> azonenberg: so it looks like there's a lot of errors happening around resize 4 in WaveformArea::on_resie
<bluezinc_> ,s/resie/resize/
<bluezinc_> and then on_render fails the sanity check at the end
<bluezinc_> and a large portion of the display is black
<bluezinc_> I'm just launching this - not doing anything fancy with moving viewports or any of that
<azonenberg> bluezinc_: yeah i was experimenting with seeing if i could make a resistor with lower parasitics
<azonenberg> re errors, interesting
<azonenberg> so, that doesnt actually look too wrong
<azonenberg> The toolbar at the top should be showing up which is funny
<azonenberg> (above the timeline)
<bluezinc_> if I mouse over parts, they start to show up
<azonenberg> now that is interesting
<azonenberg> so, just for lulz, can you try making the dummy scope report four channels and see what happens?
<bluezinc_> yeah, sure, 1 sec
<azonenberg> Very interesting
<azonenberg> this looks like ubuntu, what gtk version are you on?
<bluezinc_> stock disco, 3.24.8-1ubuntu1
<azonenberg> i wonder if its multisampling related
<azonenberg> can you pastebin a glxinfo?
<azonenberg> (we should probably have a graceful fallback if antialiasing doesnt work - not saying this is the problem but it's a suspect)
<azonenberg> (also heading out for a bit, will be on _mobil in a few)
<bluezinc_> azonenberg: it also appears that this might very well not support antialiasing.
<bluezinc_> but you know better exactly what capabilities you use.
azonenberg_mobil has joined #scopehal
<azonenberg_mobil> back
<monochroma> 20:08 < azonenberg> (also heading out for a bit, will be on _mobil in a few)
<monochroma> 20:11 < bluezinc_> https://paste.debian.net/1089016/
<monochroma> 20:11 < bluezinc_> azonenberg: it also appears that this might very well not support antialiasing.
<monochroma> 20:12 < bluezinc_> but you know better exactly what capabilities you use.
<azonenberg_mobil> ok so make a ticket for poor non-AA support and I'll try to add a fallback mode tomorrow
<azonenberg_mobil> but the dumo mentions GL_ARB_multisample being supported
<azonenberg_mobil> dump*
sorear_ has joined #scopehal
elms_ has joined #scopehal
sorear has quit [*.net *.split]
elms has quit [*.net *.split]
elms_ is now known as elms
sorear_ is now known as sorear
m4ssi has joined #scopehal
azonenberg_mobil has quit [Ping timeout: 276 seconds]
vup has quit [Ping timeout: 258 seconds]
vup has joined #scopehal
m4ssi has quit [Remote host closed the connection]
bvernoux has joined #scopehal
<azonenberg> o/ bvernoux
<bvernoux> hello
<azonenberg> did you see the probe respin i tweeted?
<azonenberg> any thoughts on the multi resistor/compensation design?
<azonenberg> https://i.imgur.com/gI4gizD.png i played around a bit in Sonnet modeling it
<azonenberg> If this works remotely as good as the sims sugget i'll be ready to send it your way for characterization soon when it comes back from fab
<bvernoux> ha nice
<bvernoux> I'm working on VNA Qt tool ;)
<bvernoux> and I can retrieve full S2P data to an official Touchstone 1.1 S2P file ;)=
<bvernoux> so It will help
<bvernoux> I have also received my calibration kit even if it miss 1x Male Load that I shall receive in few days
<bvernoux> I will check your simulation
<bvernoux> with QUCS
<bvernoux> seems veru interesting
<azonenberg> The big missing bit is full 3d field solver modeling of the entire pcb, mmcx, and tip
<azonenberg> sonnet free caps you to pretty small simulations so that isn't an option
<bvernoux> My Tool WIP ;)
<bvernoux> S2P captured by my tool (using PC USB <=> GPIB HP8753D) and displayed with Smith V4.1
<bvernoux> it is just a basic attenuator characterized with a very good VNA
<bvernoux> and my S2P are not far from the reference
<bvernoux> even without the final calibration kit as I have not entered the CAL KIT details and I'm using female load + male-male adapter instead of male load (as I shall receive it soon with some spare RP-SMA adapters ...)
<monochroma> oooo smith charts
<bvernoux> see the ref https://imgur.com/a/2v7r1PS
<bvernoux> I shall obtain that at end ;)
<bvernoux> I plan to release my VNA Qt tool as open source
<bvernoux> when it will be more mature
<bvernoux> as it is a shame that there is nothing working ...
<bvernoux> old projects from HP or other does not work fine with standard GPIB VISA dll ...
<bvernoux> I'm using IVI Foundation VISA components to have maximum possible compatibility with any USB GPIB ...
<bvernoux> it is amazing that more than 30years old VNA or other GPIB instruments can be managed from today PC with USB/GPIB ;)
<bvernoux> azonenberg, about the simulation in 3D OpenEMS could be very nice but there is a learning curve ...
<azonenberg_work> yeah
<azonenberg_work> the big thing is that there isnt any good way to do the design without writing a ton of code
<bvernoux> see you
<bvernoux> bye
bvernoux has quit [Quit: Leaving]