Closed chalin closed 3 years ago
Nice catch, sorry about this! I believe I've fixed it, pushing out v0.88.1-patch1 right now (kind of an ugly version but I still think matching this package's version with Hugo's is convenient). This also includes some refactoring I've done in the meantime so definitely let me know if you see any other issues!
Also let me know if you need a Hugo version other than v0.88.1 patched.
Thanks for the quick fix @jakejarvis!
... pushing out v0.88.1-patch1 right now (kind of an ugly version but I still think matching this package's version with Hugo's is convenient).
I totally agree: I prefer that the package versions match. As I mentioned in https://github.com/fenneclab/hugo-bin/issues/105#issuecomment-942572355, you can use +1
, +2
, etc for hugo-extended
patches. I believe that semver rules (relative to npm's ^
notation) will then allow those patched versions to be picked up automatically.
~Edit: maybe -patch1
works just as well.~
FYI, X.Y.Z-patch1
makes it a prerelease version of X.Y.Z
, which isn't really what we want. IMHO, it would be better to use a +
suffix (which is used for build metadata) instead. For details, see https://semver.org/#spec-item-10. Cheers
When our website contains an error and
hugo
reports the error as expected, hugo's 255 exit status code isn't being propagated:Notice the 0 exit code being reported.
This wasn't the case for v0.83.x or earlier:
Notice the proper 255 exit code is being propagated.
This issue caused a build failure to go unnoticed, and a broken website to be deployed to our production server 😞. For details see, https://github.com/open-telemetry/opentelemetry.io/issues/823.
Thanks for the
hugo-extended
package @jakejarvis, it's super convenient. I hope that this issue can be easily and quickly resolved./cc @austinlparker @nate-double-u @celestehorgan