00:44 <whitequark[cis]> <zerobytesecurity> "So if you want some I will..." <- i actually do want some

00:45 <zerobytesecurity> I have a tray of 50 I can send you. They look like the picture and are not complete clips. These are just the tops that you can place on nand chips, just to be clear.

00:47 <zerobytesecurity> Let me know where to send them or to bitsquared office and off they will go this week.

00:51 <whitequark[cis]> so this goes on top of a flex cable to hold a chip on it, right?

00:51 <whitequark[cis]> or does it have to be rigid flex?

00:58 <zerobytesecurity> one second let me get you a link for easier understanding

00:59 <zerobytesecurity> they click right on top of tsop 48

01:00 <zerobytesecurity> https://www.datasheetarchive.com/datasheet/b61a000424f2fff7?term=IC197-4807&type=P

01:01 <whitequark[cis]> am still confused, do you only have one half of them? how do you intend to use them?

01:02 <zerobytesecurity> These do not have lids, I.e bracket.

01:03 <zerobytesecurity> This is what I am building with out the top piece.

01:04 <zerobytesecurity> so its the bottom, and then you have to solder the ffc onto the top, which usually has a cover but good luck i haven't been able to source a whole clip at all

01:04 <zerobytesecurity> better pic https://www.ebay.com/itm/123712159342

01:07 <zerobytesecurity> let me show you the clip I am duplicating because i refuse to pay 100 bucks to someone in china for it.

01:12 <zerobytesecurity> if you want a clip that is fully complete and it will help you on your dev work the brown one in the picture is 48 pin. if you dont want to build yourself or go down crazy route like me. i can send you one full clip.

01:20 <jn> i'm adding tests to an applet, but they don't seem to get picked up (neither locally with python3 -m unittest, nor in GitHub Actions) -- i wonder if i missed some mistake: https://github.com/GlasgowEmbedded/glasgow/compare/main...neuschaefer:glasgow:mpc83xx#diff-1e74f22fbaf5ade4292e70adbf08719c0da8ff2df7676d98f1524ee1b89b7e52R568 (i haven't opened a PR yet, because the code is still quite chaotic and incomplete)

01:59 <whitequark[cis]> it's not automatic, you need a method in the applet class, i think `def tests` or sth

02:23 <jn> i have that, but python behaves really strangely: when i, for example, edit the 'tests' function in software/glasgow/applet/debug/arm/arm7/__init__.py to return None (or remove the tests method entirely), the arm7 applet tests are still run. i'm on an editable install and placing a "raise RuntimeError('foo')" into the global scope of arm7/__init__.py shows that the file is indeed loaded. i'm confused

08:19 <galibert[m]> Catherine: before I try, do you think the glasgow has enough usb bandwidth for two 44100 32-bits stereo synchronous streams? That's 705600 bytes per second uninterrupted

08:23 <zyp[m]> throughput-wise that should be trivial, the only issue you might have is the tradeoff between how much buffering you need to make it reliably uninterrupted and how much latency you can accept

08:24 <galibert[m]> latency is not an issue, not losing data is important

08:24 <galibert[m]> (in-depth analysis of an old expander [synth without a keyboard])

09:07 <Attie[m]> iirc Glasgow can achieve ~42 MB/s, so ~0.7 MB/s should be trivial

12:03 <whitequark[cis]> <jn> "i have that, but python behaves..." <- how are you running tests?

12:03 <whitequark[cis]> <Attie[m]> "iirc Glasgow can achieve ~42 MB..." <- this is correct

12:03 <jn> i found the problem with my tests not being run: my applet is in software/glasgow/applet/debug/powerpc/e300, but i forgot to add an empty software/glasgow/applet/debug/powerpc/__init__.py

12:04 <jn> cd software && exec ../venv/bin/python3 -m unittest -v -k debug

12:04 <whitequark[cis]> oh, yeah, that would do it

12:04 <jn> so, discovery broke but the explicit entry points in pyproject.yaml still worked

12:05 <whitequark[cis]> `glasgow test powerpc-e300-whatever` is how i would run the tests

12:05 <jn> oh! TIL about glasgow test

12:05 <whitequark[cis]> or `pdm test -k ...`

12:07 <jn> with this out of the way, i can actually write tests now :)

12:07 <whitequark[cis]> excellemt

12:07 <whitequark[cis]> * excellent