Re: git: 2ba4ed91621e - main - sysutils/createrepo_c: new port

On Oct 04 2024, 05:57 +0000, "Jason E. Hale" <jhale@freebsd.org> wrote:
>On Wed, Oct 2, 2024 at 3:21 AM Pietro Cerutti <gahr@freebsd.org> wrote:
>>
>> On Oct 01 2024, 17:04 +0000, Daniel Engberg <daniel.engberg.lists@pyret.net> wrote:
>> >
>> >Poudriere also lists quite a few warnings about dependencies and the
>> >REINPLACE_CMD commands.
>> >>https://pdr2.bofh.network/data/141-diizzy/2024-10-01_11h36m29s/logs/createrepo_c-1.1.4.log
>>
>> Nice. I wasn't aware of poudriere's Q/A phase. I used poudriere bulk to
>> test the port but I don't see that in my logs. How do I enable it?
>> https://gahr.ch/poudriere/logs/bulk/14amd64-dev/2024-10-01_12h26m54s/logs/createrepo_c-1.1.4.log
>
>You can add the following to your poudriere.conf:
>PORTTESTING=1
>PORTTESTING_RECURSIVE=1
>PORTTESTING_FATAL=no
>
>This is equivalent to adding the -trk flags to `poudriere bulk` see
>poudriere-bulk(8). It's basically like using `poudriere testport` in
>bulk mode. Downside is sometimes you may encounter build problems with
>deps that send you down a rabbit hole and the build stalls out on a
>select few popular ports like doxygen that ultimately build and
>package, but touch the filesystem in the process. The command just
>needs to be run again in this case.
>
>Setting PORTTESTING_FATAL=no (or using the -k switch) is kind of iffy,
>though. For example, if a port you are working on simply adds new
>files that are not in pkg-plist, a package will still be created and
>the build will keep going, which could be misleading and negatively
>affect consumers. Poudriere will add this build log to the errors
>directory, however. Definitely check the build logs to make sure.

Thanks!

-- 
Pietro Cerutti
I have pledged to give 10% of income to effective charities
and invite you to join me - https://givingwhatwecan.org