fedora-review denied: [Bug 225238] Merge Review: adaptx

[bugzilla at redhat.com]

Bug 225238: Merge Review: adaptx
Product: Fedora Extras
Version: devel
Component: Package Review

Andrew Overholt <overholt at redhat.com> has denied Andrew Overholt
<overholt at redhat.com>'s request for fedora-review:
https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=225238

------- Additional Comments from Andrew Overholt <overholt at redhat.com>
MUST:
X rpmlint on adaptx srpm gives no output

W: adaptx invalid-license Exolab Software License

I've emailed fedora-maintainers about this.  It looks like it'll be okay but I
want to confirm first.

* package is named appropriately
* specfile name matches %{name}
X package meets packaging guidelines.

. BuildRoot incorrect.	As per this:

http://fedoraproject.org/wiki/Packaging/Guidelines#BuildRoot

it should be:

%{_tmppath}/%{name}-%{version}-%{release}-root-%(%{__id_u} -n)

. do we need section free?  I think it can safely be removed.
. can the Summary be expanded a bit?
. the commented-out build-classpath line can probably be removed.  was the
dependency on js removed?
. are the two patches still necessary?
. do we need the explicit Epoch?  I can't find anything in the guidelines about

this but I think it's a bit verbose.  After speaking with others, we've come
to the conclusion that it doesn't violate the guidelines ... but I still don't
like it personally :)

X license field matches the actual license.

. see above

X license is open source-compatible.

. again, see above.

* license text included in package and marked with %doc
* specfile written in American English
* specfile is legible
X source files match upstream
. I assume upstream doesn't provide source drops?  The svn export is fine in
this case (please use an SVN tag), but can you put exact instructions for how
to
generate the tarball?  Preferably something that can be duplicated by just
removing leading #'s.  I'd also like to see the comment be before the Source
entry but that's just personal preference :)  Also, I don't think the RHCLEAN
is
necessary because I think we can re-distribute what binary jars they include
upstream.  As long as we symlink to our built ones, it should be fine.	

* package successfully compiles and builds on at least x86 (it's building on
the other arches in Fedora Core presently)

* BuildRequires are proper
* no locale data so no find_lang necessary
* package is not relocatable
* package owns all directories and files
. I've said in other reviews how I don't like how the javadoc symlinking is
  being done, but this won't hold up the review.
* no %files duplicates
* file permissions are fine; %defattrs present
* %clean present
* macro usage is consistent
* package contains code
* -doc subpackage fine
. javadoc package present also
* %doc files don't affect runtime
* shared libraries are present, but no ldconfig required.
* no pkgconfig or header files
* no -devel package
* no .la files
* no desktop file
* not a web app.
* file ownership fine
? final provides and requires are sane

Do we need a java dependency somewhere?  How about a Requires on the things we
BR like log4j?

$ rpm -qp --provides i386/adaptx-0.9.13-4jpp.1.i386.rpm
adaptx-0.9.13.jar.so  
adaptx = 0:0.9.13-4jpp.1

$ rpm -qp --provides i386/adaptx-doc-0.9.13-4jpp.1.i386.rpm 
adaptx-doc = 0:0.9.13-4jpp.1

$ rpm -qp --provides i386/adaptx-javadoc-0.9.13-4jpp.1.i386.rpm 
adaptx-javadoc = 0:0.9.13-4jpp.1

$ rpm -qp --requires i386/adaptx-0.9.13-4jpp.1.i386.rpm
/bin/sh  
/bin/sh  
java-gcj-compat  
java-gcj-compat  
jpackage-utils	
jpackage-utils	
libc.so.6  
libc.so.6(GLIBC_2.1.3)	
libdl.so.2  
libgcc_s.so.1  
libgcc_s.so.1(GCC_3.0)	
libgcc_s.so.1(GLIBC_2.0)  
libgcj_bc.so.1	
libm.so.6  
libm.so.6(GLIBC_2.0)  
libpthread.so.0  
librt.so.1  
libz.so.1  
rpmlib(CompressedFileNames) <= 3.0.4-1
rpmlib(PayloadFilesHavePrefix) <= 4.0-1
rtld(GNU_HASH)	

$ rpm -qp --requires i386/adaptx-doc-0.9.13-4jpp.1.i386.rpm 
rpmlib(CompressedFileNames) <= 3.0.4-1
rpmlib(PayloadFilesHavePrefix) <= 4.0-1

$ rpm -qp --requires i386/adaptx-javadoc-0.9.13-4jpp.1.i386.rpm 
/bin/ln  
/bin/rm  
/bin/rm  
/bin/sh  
/bin/sh  
rpmlib(CompressedFileNames) <= 3.0.4-1
rpmlib(PayloadFilesHavePrefix) <= 4.0-1

SHOULD:
* package includes license text
* package builds on i386
  ... and others in brew ATM; I don't envision a problem here
X package functions
  . I don't know how to test this package
X package builds in mock
  my mock setup doesn't seem to be working but I don't anticipate any problems
  here as the package currently builds fine in brew