Packages with missing %check

[Stanislav Ochotnicky]

On Wed 05 Mar 2014 11:23:23 AM CET Alexander Todorov wrote:

> На  4.03.2014 20:36, Mat Booth написа:
>> On 25 February 2014 11:19, Mikolaj Izdebski <mizdebsk at redhat.com> wrote:
>>
>>> On 02/25/2014 11:45 AM, Alexander Todorov wrote:
>>>> 3) Another proposal (sorry don't remember who proposed it) was to have
>>>> %check with a comment why the test suite is not executed (e.g. requires
>>>> network) or why it is executed in %build.
>>>
>>> Commenting why tests are skipped is a very good thing, but I don't like
>>> the idea of adding empty %check sections to my 250+ packages just for
>>> the sake of documenting that tests are ran in %build "because that's
>>> what we do in Java world".
>>>
>>>
>> Agreed, it seems like busy work to me that adds very little value to anyone
>> familiar with Java packages.
>
> You are forgetting everyone that is not so familiar with Java.

Why are you filing bugs (with patches) you don't understand then? Have
you asked *anybody* to help you out with exclusions? Have you gotten in
touch with Java|Perl|PythonX|Y|Z SIG and ask for their assistance in better
identification of testsuites and if they are being run?

You are putting cart before the horse rushing the whole mass bug filing
process without understanding consequences. You are going to be
ignored. Your bugs are going to get closed/wontfix. You are going to get
annoyed. You are not going to achieve much (if anything).

> Also I didn't ask you (as a package owner) to do it explicitly, I've asked you
> to accept a patch which should be much more easier.

Patch which contains text which you haven't verified is
correct. Quoting:

+%check
+# tests are executed during %build
+

How do *you* *know* they are executed during build? Do you even know how
to recognize a difference between following:

ant dist test javadoc
ant dist javadoc
ant jar
%mvn_build
%mvn_build -f
%mvn_build -Dtestnotmatchpattern=\*
%mvn_build -- -Dmaven.test.skip=true
mvn-rpmbuild -Dmaven.test.failure.ignore=true
...

Hint: Except "%mvn_build" and "ant dist test javadoc" these executions
most likely don't run tests or if they do they ignore the failures

>> Wouldn't it be easier to change the whatever
>> tool is generating this report to accommodate for this? "If package invokes
>> %mvn_build then don't expect there to be a %check section" seems like a
>> reasonable heuristic to me.
>>
>>
>
> See https://bugzilla.redhat.com/show_bug.cgi?id=1072417#c4 to avoid repeating
> myself.
>
>
> Even if the tool uses heuristics to exclude some groups of packages it will not
> be obvious why there's no %check section. It could be because tests are executed
> in %build, because they need root or network access and are disabled, because
> the test framework used is not available (see DHCP) or anything else.

Right, so your patches are basically worthless and you are planning to file
possibly hundreds of bugs where maintainers will most likely manually
close as WONTFIX/NOTABUG because it's clear to anyone working in their
respective SIG what's happening there.

> A small comment makes it much more clear and straight forward.

I want my packages to run tests. If there are tests upstream but I am
not running them, sure there should be a comment or even better a
FutureFeature bug to fix the testsuite. We are mostly doing all of this
in Java SIG packages. Adding useless comment to spec files is not the
way to improve things.

If you want to improve stuff then focus on identifying packages which
should run tests but don't.

--
Stanislav Ochotnicky <sochotnicky at redhat.com>
Software Engineer - Developer Experience

PGP: 7B087241
Red Hat Inc.                               http://cz.redhat.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 835 bytes
Desc: not available
URL: <http://lists.fedoraproject.org/pipermail/devel/attachments/20140305/be87ebf5/attachment.sig>