Re: Revert coreutils 9.1 to test 8.32 to curr stable

[Brian Inglis <Brian.Inglis@SystematicSw.ab.ca>]

On 2022-06-07 03:39, Jon Turney wrote:
> On 07/06/2022 00:33, Brian Inglis wrote:
>> Verified below issue with coreutils 9.1 promoted last night to stable!
>> Problem does not occur in 8.32, so can we please revert coreutils 9.1 
>> to test and restore 8.32 to stable.
>> Or is there some process by which I can do so?
> 
> You seem to have figured this out, but FTR:
> * Remove [1] or re-upload [2] the package with test: added to hints
> * Upload an override.hint with replace-versions: [3]
> 
> Note that due to the way the depsolver works, replace-versions: doesn't 
> do anything if the package would still be chosen to install.
> 
> (calm tries to warn about that, but it seems you've tripped over a bug 
> where it doesn't take test: into consideration and warns when it shouldn't)
> 
> [1] https://cygwin.com/package-upload.html#deleting
> [2] Re-uploading a package with the same (or missing!) archives, but 
> modified hints is currently allowed for historical reasons
> [3] https://cygwin.com/packaging-hint-files.html#override.hint

Thanks Jon, fo looking into this, but package-test rebuild and upload 
[2] resulted in ERROR and discards:

ERROR: discarding, different /sourceware/cygwin-staging/home/Brian 
Inglis/x86/release/coreutils/coreutils-9.1-1-src.tar.xz is already in 
release area (perhaps you should rebuild with a different 
version-release identifier?)
INFO: discarding, identical /sourceware/cygwin-staging/home/Brian 
Inglis/x86/release/coreutils/coreutils-9.1-1.tar.xz is already in 
release area
INFO: discarding, identical /sourceware/cygwin-staging/home/Brian 
Inglis/x86/release/coreutils/coreutils-debuginfo/coreutils-debuginfo-9.1-1.tar.xz 
is already in release area
ERROR: error while reading uploaded arch x86 packages from maintainer 
Brian Inglis
SUMMARY: 2 ERROR(s), 2 INFO(s)

and same for other arch.

-- 
Take care. Thanks, Brian Inglis, Calgary, Alberta, Canada

This email may be disturbing to some readers as it contains
too much technical detail. Reader discretion is advised.
[Data in binary units and prefixes, physical quantities in SI.]