Bug 240942

Good:
+ Naming seems ok.
+ Rpmlint is quite for source rpm.
+ Rpmlint is quite for binary rpm.
+ Rpmlint is quite for debuginfo rpm.
+ Package contains verbatin copy of the license text.
+ Local install and uninstall works fine.
+ Startup of the application works without a crash.
+ Mock build works fine.

Bad:
- Source0 is not a full qualified URL.

Hi Jochen!

(In reply to comment #1)
> Bad:
> - Source0 is not a full qualified URL.
Per my first message there is a reason behind this.

(In reply to comment #0)
> N.B. I'm sure someone wants to tell me "but the URL is wrong" as the comment
in the spec file says, rpmbuild throws a fit when there is a ? after the /, I'm
going to try and find out if there is a fix or not.

#Real SourceURL is: http://culot.org/cgi-bin/get.cgi?%{name}-%{version}.tar.gz 
#but the ? annoys rpmbuild
Source0:        %{name}-%{version}.tar.gz

The issue that this fixes is that if you do not do that, rpmbuild will try and
extract get.cgi?calcurse-1.8.tar.gz, not a good result.

* Well, I checked the URL and for this case it is not a
  problem. This type of URL problem (i.e. the "correct" URL
  contains some formats which rpmbuild cannot treat correctly)
  became a matter of argument (but I don't remember when
  and at which mailing list for now...) and the conclusion was

  "That is okay. Write the correct URL as a comment and
   use the tarball (if needed, renamed) name to Source."

* By the way
-------------------------------------------------------
%configure
CFLAGS=$RPM_OPT_FLAGS make %{?_smp_mflags}
-------------------------------------------------------
Is "CFLAGS" in the second line needed? (%configure sets
CFLAGS before ./configure is called)

(In reply to comment #3)
> * Well, I checked the URL and for this case it is not a
>   problem. This type of URL problem (i.e. the "correct" URL
>   contains some formats which rpmbuild cannot treat correctly)
>   became a matter of argument (but I don't remember when
>   and at which mailing list for now...) and the conclusion was
> 
>   "That is okay. Write the correct URL as a comment and
>    use the tarball (if needed, renamed) name to Source."
This was the same opinion I got from two maintainers on IRC
> 
> * By the way
> -------------------------------------------------------
> %configure
> CFLAGS=$RPM_OPT_FLAGS make %{?_smp_mflags}
> -------------------------------------------------------
> Is "CFLAGS" in the second line needed? (%configure sets
> CFLAGS before ./configure is called)
I think I found in this case it was needed, I'll double check before the package
is entered into CVS though, but I seem to recall that it was needed.

(In reply to comment #4)
>> %configure
>> CFLAGS=$RPM_OPT_FLAGS make %{?_smp_mflags}
>> -------------------------------------------------------
>> Is "CFLAGS" in the second line needed? (%configure sets
>> CFLAGS before ./configure is called)
>I think I found in this case it was needed, I'll double check before the 
> package is entered into CVS though, but I seem to recall that it was needed.
It is not needed. %configure is enough

Remove it.

Spec URL: http://dev.nigelj.com/SRPMS/calcurse.spec
SRPM URL: http://dev.nigelj.com/SRPMS/calcurse-1.8-2.src.rpm

I'm not actually sure where the CFLAGS make came from, but when I was test
building it was needed, but it builds okay still so here is an updated package.

Good:
+ Sources matches with upstream.
+ Local build works fine.

For the not qualified Source0 expression, I can make an exception, becouse you
wrote, that this couse an issue.

It may be nice, if you can submit a bug report agains the rpm package.

*** APPROVED ***

(In reply to comment #7)
> It may be nice, if you can submit a bug report agains the rpm package.
I agree, I will do that over the weekend

New Package CVS Request
=======================
Package Name: calcurse
Short Description: Text-based personal organizer
Owners: dev
Branches: FC-6 FC-7 EL-4 EL-5
InitialCC: 

cvs done.

Built and done, thanks for the review!