Marcus Müller

Chances are high it's me that changed the argument type. Yes, the docstring is incorrect! And I still think we should change the fastnoise source's long to an uint64_t, but...

@mormj I'll need a bit of guidance on this one. Turns out, at least on main we *never* set `HAVE_CONFIG_H`, so *all* of the includes are moot, and all of...

ahahahahaa! https://github.com/gnuradio/gnuradio/blob/53dd57ca8fb8bc929c2b4cef6384d14342ed49e7/gnuradio-runtime/include/gnuradio/rpcserver_selector.h#L11-L22 by virtue of rpcserver_selector, which gets included through basic_block.h, we *always* include gnuradio/config.h, in almost every source file.

> The GNURadio package in repositories tends to be outdated and compiling by source is a struggle due to many dependencies. right now, we're pretty good! The current Fedora, Ubuntu,...

@Po84210 thanks for weighing in on this! I don't think flatpak addresses device permissions, though: The user executing the flatpak needs to have the same permissions as if the UHD/rtl-sdr/…...

This should be rather straightforward to implement, will do this week, unless someone else wants to! `f"{num:6g}"`

Yeah, fails everywhere: HAVE_CONFIG_H is simply never defined, and all our checks from config.h are always false, no matter whether they've been configured correctly :(

The qa_window issue is strange to appear only on an i686 build: I can see how we introduce a small error, as the window::build method actually calculates a `double` error,...

@jdemel ah, thanks! On one hand, a bit sad, on the other, I'm having a hard time finding a 32bit x86 system these days; hm.

@unspecd hey, I've been digging a tiny bit, and debian seems to build things well on 32 bit: https://buildd.debian.org/status/fetch.php?pkg=gnuradio&arch=i386&ver=3.10.3.0%7Erc1-1&stamp=1656071459&raw=0 bit stumped on how to reproduce; will probably need your input...