00:32 <bartmassey[m]> My students and I and issues on GH and stuff have been having huge trouble with `link.x` and the Discovery Book. Some host platforms it gets used twice, some it doesn't at all; depends a lot on hard-to-identify context.

00:32 <bartmassey[m]> A thing that would help is if it could be made idempotent. Anybody know enough linker-script magic to make it OK to use `link.x` a second time and have it not do anything?

00:32 <bartmassey[m]> Thanks for any help you can give.

01:50 <rmsyn[m]> <bartmassey[m]> "My students and I and issues..." <- what crates are having issues? is it pulling the `link.x` from the local project, or a dependency? if there is a `link.x` in a dependency + one in the local project, I've had luck with renaming the local one to something like `memory.x` (picked it up from other BSPs).

01:50 <rmsyn[m]> maybe that's not the issue, though?

01:51 <JamesMunns[m]> rmsyn[m]: Also maybe check that they don't have BOTH of a `build-rs` and the linker script set in `.cargo/config.toml`?

01:56 <JamesMunns[m]> s/the/a/, s/linker script set in//, s/`?/` that are setting the linker directives?/

02:35 <M9names[m]> bartmassey: can you be more specific than "some host platforms"?

02:35 <M9names[m]> it would be nice to reproduce it so it could be understood why it's happening, but that's going to be painful if we have to manually test every possible platform

16:24 <bogdan[m]> I suspect I integrated embedded-test badly. Is this the way the output is supposed to look like?

16:24 <bogdan[m]> seems very noisy

17:24 <t-moe[m]> <bogdan[m]> "clipboard.png" <- At the moment this is just how things are by default. But you can supress stderr (aka redirect to /dev/null) and use --format to control the output format (json, compact, ...)

17:29 <bogdan[m]> thanks! at least now I know I'm on the right track