Re: [PATCH] ltmain.in: don't suppress output for PIC compilations

--=-=-=
Content-Type: text/plain

Nick Bowler <[email protected]> writes:

> On 2024-08-15 13:37, Sam James wrote:
>> More annoying to have an error suppressed that is not obvious to debug,
>> especially if one isn't aware of this behaviour.
>
> By default (unless you pass --silent), libtool prints the compiler commands
> it is executing, which includes the output redirection to /dev/null on the
> second command.
>
> So if the second invocation crashes, you can just copy+paste the last command
> line libtool printed without the redirection to /dev/null to see the output,
> and now you are aware of this behaviour, right?

This isn't so useful if doing automated builds and given a log.

It's also clearly not obvious to people who aren't inexperienced, given the
reference I gave, so it's not working that well.

What you described is, of course, what I did with the original xz case I
had, after being initially baffled as I had no idea the behaviour
existed.

I'm not really seeing any sort of justification for doing this other
than "duplicate output maybe being annoying", but the duplicate output
isn't even assured, and it appears based on a misnomer anyway.

--=-=-=
Content-Type: application/pgp-signature; name="signature.asc"

-----BEGIN PGP SIGNATURE-----

iOUEARYKAI0WIQQlpruI3Zt2TGtVQcJzhAn1IN+RkAUCZr5ABF8UgAAAAAAuAChp
c3N1ZXItZnByQG5vdGF0aW9ucy5vcGVucGdwLmZpZnRoaG9yc2VtYW4ubmV0MjVB
NkJCODhERDlCNzY0QzZCNTU0MUMyNzM4NDA5RjUyMERGOTE5MA8cc2FtQGdlbnRv
by5vcmcACgkQc4QJ9SDfkZAsegEAmy6Q/HxusxfVRG5ik1OaCK+zj4eefsJyPVba
acsAZSMA/RX2jXwl1BP1gKt0jj1gXaHfBQAr6UHXshgziF6Bbe0N
=5qDk
-----END PGP SIGNATURE-----
--=-=-=--