Several problems in "make test"

[farblos]

There are a number of problems in the current test suite (1.66 + some PRs):

• there are tests t/26_symlink.t and t/26_bug_26.t. One of them should be renumbered for clarity.

• test t/26_bug_26.t is not mentioned in MANIFEST and not packaged, hence

• test t/26_bug_26.t fails for yet unknown reasons

• in Travis (e.g. Job #26.1 of pull request 53), common.pm moans that "zip doesn't seem to work". While that does not let fail any tests, it potentially skips essential ones

While look into these ASAP.

[farblos]

Can reproduce failing test t/26_bug_26.t locally but don't have the time to look into more details now. @pmqs, any opinion on that one?

No idea yet what to do about last problem ("zip doesn't seem to work") but to add diagnostic output.

[pmqs]

• there are tests t/26_symlink.t and t/26_bug_26.t. One of them should be renumbered for clarity.
• test t/26_bug_26.t is not mentioned in MANIFEST and not packaged, hence

Created PR #55 to address the issues above.

[pmqs]

Can reproduce failing test t/26_bug_26.t locally but don't have the time to look into more details now. @pmqs, any opinion on that one?

will try to get a look at.

No idea yet what to do about last problem ("zip doesn't seem to work") but to add diagnostic output.

I'm not seeing that error.

[farblos]

I'm not seeing that error.

Agreed. Must have been an intermittent issue, to be seen only in the one log file I've quoted above.

[farblos]

* test `t/26_bug_26.t` fails for yet unknown reasons

@pmqs: Seems that the zip64 PR #45 has thrown out the changes in Member.pm of your bzip compression commit 9122328. (I remember having eliminated parameter $mode to function head because the only value ever passed to that method before commit 9122328 was 1.)

I guess that this could be the root cause, but don't have time to validate it now.

Um, probably you could add a one-liner to test 26_bzip.t in your PR #54 to provide more details what the test should actually do?

[pmqs]

@pmqs: Seems that the zip64 PR #45 has thrown out the changes in Member.pm of your bzip compression commit 9122328. (I remember having eliminated parameter $mode to function head because the only value ever passed to that method before commit 9122328 was 1.)

I guess that this could be the root cause, but don't have time to validate it now.

yes, I that is exactly why the test is failing.

Um, probably you could add a one-liner to test 26_bzip.t in your PR #54 to provide more details what the test should actually do?

Done

[redhotpenguin]

Thanks. Will push 1.67 this weekend.

…

On Fri, Sep 20, 2019, 3:26 PM Paul Marquess ***@***.***> wrote: @pmqs <https://github.com/pmqs>: Seems that the zip64 PR #45 <#45> has thrown out the changes in Member.pm of your bzip compression commit 9122328 <9122328>. (I remember having eliminated parameter $mode to function head because the only value ever passed to that method before commit 9122328 <9122328> was 1.) I guess that this could be the root cause, but don't have time to validate it now. yes, I that is exactly why the test is failing. Um, probably you could add a one-liner to test 26_bzip.t in your PR #54 <#54> to provide more details what the test should actually do? Done — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#54?email_source=notifications&email_token=AAAXRZN4RJKMSCVKFK6WO7LQKVEYDA5CNFSM4IYWJJ52YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD7IBIXQ#issuecomment-533730398>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAAXRZJEZNDZOT2NNLD2FJTQKVEYDANCNFSM4IYWJJ5Q> .

[redhotpenguin]

Sorry - maybe I read too fast here. Looks like there's still some investigation going. LMK when a release is needed.

[farblos]

To summarize, we have the following issues and PRs that I consider worth including in the next release:

• this issue, fixed by pmqs' PR Minor fixes for some issue in #54 #55 and Fix for bzip2 passthrough #57

• issue zip64 extension does not respect versionMadeBy/versionNeededToExtract settings at member level #51, fixed by PR Issue 51 #53

• issue Fix pack issues introduced in zip64 extension in Archive::Zip 1.66 #56

[pmqs]

To summarize, we have the following issues and PRs that I consider worth including in the next release:

• this issue, fixed by pmqs' PR Minor fixes for some issue in #54 #55 and Fix for bzip2 passthrough #57
• issue zip64 extension does not respect versionMadeBy/versionNeededToExtract settings at member level #51, fixed by PR Issue 51 #53
• issue Fix pack issues introduced in zip64 extension in Archive::Zip 1.66 #56

Agree

[farblos]

All problems fixed in version 1.67.