00:58 <azonenberg> miek, lain: https://www.antikernel.net/temp/probe-vs-zs1500.png

00:58 <azonenberg> this is S11 of the latest AKL-PT1 variant vs the ZS1500 active probe across an open circuit, so 0 dB of S11 (all incident signal reflected) would be the ideal result

00:58 <azonenberg> From DC to 2 GHz

00:59 <azonenberg> you can see the active probe is actually worse than the 500 ohm passive probe starting at about 570 MHz

01:00 <miek> nice

01:09 <azonenberg> I mean this isn't surprising

01:09 <azonenberg> this effect is literally the reason why transmission line probes exist

01:09 <azonenberg> well, that and they cost much less :p

01:10 <azonenberg> For the review i'm using the PicoConnec 921 for comparison which is a $1K 6 GHz passive probe vs the $2K MSRP 1.5 GHz active probe

01:10 <azonenberg> but the AKL-PT1 used in that screenshot will probably sell for $250 or so

01:10 <azonenberg> and also hit 6 GHz of bandwidth :p

01:53 <azonenberg> Just finished the zs1500 review and sent it in to xdevs for publication, will share a link once it's up

04:06 <azonenberg> Bird|otherbox: ping

04:06 <Bird|otherbox> pong

04:06 <Bird|otherbox> I think I fixed the noreturn warning but I can't tell for sure

04:07 <azonenberg> So i'm almost done cleaning up all of the static analysis warnings in libscopehal + libscopeprotocols (glscopeclient itself is another task i'm only starting)

04:07 <azonenberg> There are two remaining types

04:07 <azonenberg> one is the noreturn one you're looking at

04:07 <azonenberg> The other is a bunch of spots where i have format strings that use %lu for size_t

04:07 <azonenberg> normally i would use %zu which i believe is supposed to be the C99 compliant standard way to do it

04:08 <azonenberg> but i recall somebody saying %zu didn't work right on some platform

04:08 <azonenberg> any idea what that platform might be? or should i just go ahead and %zu everything?

04:12 <azonenberg> And then of course we have the linker file unused issue but that's less important

04:25 <Bird|otherbox> I think it might be Windows that has issues with %zu? let me check MSDN

04:26 <Bird|otherbox> actually, according to MSDN, the current MSVCRT printf() supports %zu

04:30 <azonenberg> does mingw use msvcrt or glibc?

04:31 <azonenberg> and how current is current?

04:33 <Bird|otherbox> mingw uses msvcrt

04:33 <azonenberg> ah ok

04:33 <Bird|otherbox> and current would be as of VS 2019

04:33 <azonenberg> let me poke bvernoux and katharina and see if they have had any issues with %zu

05:20 <azonenberg> also lain miek monochroma https://xdevs.com/review/zs1500/

05:30 <Bird|otherbox> little typo: "a manufacturer refurbished"

05:31 <azonenberg> yeah just noticed that

06:47 <_whitenotifier-f> [scopehal] azonenberg pushed 1 commit to master [+0/-0/±3] https://git.io/JTW4o

06:47 <_whitenotifier-f> [scopehal] azonenberg 01da7b3 - Added default-delete copy constructor and assignment operator to Waterfall. size_t format fixes to Agilent and Rigol drivers

06:48 <_whitenotifier-f> [scopehal-apps] azonenberg pushed 1 commit to master [+0/-0/±11] https://git.io/JTW4y

06:48 <_whitenotifier-f> [scopehal-apps] azonenberg 98dca77 - Const-ified more arguments. Reformatted some constructor initialization. Fixed potential null deref in TriggerPropertiesDialog.

08:29 <azonenberg> Bird|otherbox: so i think with those fixes glscopeclient is close to building without analysis warnings. I added some ignores for things like casting floats to char*s when reading/writing to files

08:29 <azonenberg> the noreturn one is still pending

08:30 <azonenberg> and of course still the linker script one

08:31 <azonenberg> I don't want you to waste toooo much time on that one. So if you don't think you can get it done quickly, probably best to table it for now and work on something else

08:31 <azonenberg> unless you're just really excited about fixing clang-analyzer for some reason, in which case knock yourself out :p

09:00 <electronic_eel> azonenberg: re %zu see https://freenode.irclog.whitequark.org/scopehal/2020-09-21#27964202;

09:23 <electronic_eel> about the zs1500 arcticle: I'd have added a sentence about the 1.5GHz datasheet vs. reality. otherwise I think it is good. didn't find any issues

20:51 <azonenberg> electronic_eel: so apparently the most recent msvcrt does support it

20:52 <azonenberg> i might just have to update? will look into that

21:03 <miek> azonenberg: want another review candidate? :p https://www.ebay.com/itm/TEKTRONIX-TMP9617-MICROWAVE-PROBE-DC-40GHZ-GROUND-SIGNAL-GROUND-BP/401970089000

21:09 <azonenberg> miek: No. I have no way to properly test it

21:09 <azonenberg> that looks like a 2.92mm connector which i have nothing to mate with other than SMA which would probably destroy it

21:09 <azonenberg> my VNA tops out at 6 GHz and the scope at 4

22:01 <azonenberg> miek, lain, Kliment, monochroma: https://www.antikernel.net/temp/akl-pt1-evolution.png

22:01 <azonenberg> This is S21 response of probe revs 0.7 to 1.3

22:59 <_whitenotifier-f> [scopehal-apps] azonenberg pushed 1 commit to master [+0/-0/±3] https://git.io/JT8Ux

22:59 <_whitenotifier-f> [scopehal-apps] azonenberg 628cbf6 - More static analysis cleanup