Re: Re: Implementation certification procedure

Stefano Bagnara <[email protected]> Wed, 20 Aug 2008 14:26:11 +0200
Newsgroups gmane.mail.spam.spf.devel
Message-ID <[email protected]>
Julian Mehnle ha scritto:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> Stefano Bagnara wrote:
>> Julian Mehnle ha scritto:
>>> More precisely, you have to convince someone who is able to edit the
>>> SPF project's website that the implementation passes the test suite.
>> That's unfair. IMHO.
>> You didn't have to convince anyone that pyspf and Mail::SPF pass the
>> test suite.
> 
> Right, I didn't _have_ to.  I could have easily faked my test suite
> results or simply edited the website and declared Mail::SPF as compliant
> without any proof whatsoever.
> 
> However, it should be trivial for anyone who doesn't trust me to follow
>   http://search.cpan.org/src/JMEHNLE/Mail-SPF-v2.006/INSTALL
> and check whether I have lied.
> 
> The same was, and still is, NOT true for jSPF, however:
> 
> I just gave it another try.  I installed Sun Java 6 and Maven 2 on my
> Debian machine, downloaded and unpacked apache-jspf-0.9.6-src.tar.gz, and
> tried following BUILDING.txt in the hope that it would automatically run
> tests during building (there was no other documentation to be found in the
> tarball on how to run the tests).
> 
> I do not have the James source installed, so I edited pom.xml as per
> BUILDING.txt (see the attached diff), and then ran `mvn -Plocal package`.
> It downloaded a lot of packages, and then this is what I got:
> 
> - ----------8<----------8<----------8<----------8<----------8<----------
> ...
> [INFO] Setting property: classpath.resource.loader.class 
> => 'org.codehaus.plexus.velocity.ContextClassLoaderResourceLoader'.
> [INFO] Setting property: velocimacro.messages.on => 'false'.
> [INFO] Setting property: resource.loader => 'classpath'.
> [INFO] Setting property: resource.manager.logwhenfound => 'false'.
> [INFO] [remote-resources:process {execution: default}]
> Downloading: 
> file:///home/julian/tmp/apache-jspf-0.9.6/stage/org.apache/jars/apache-jar-resource-bundle-1.2.jar
> [INFO] ------------------------------------------------------------------------
> [ERROR] BUILD ERROR
> [INFO] ------------------------------------------------------------------------
> [INFO] Resources JAR cannot be found.
> 
> Embedded error: Requested download does not exist.
> Unable to download the artifact from any repository
> - ---------->8---------->8---------->8---------->8---------->8----------
> 
> Conversely, _I_ can easily convince anyone who wants to know that
> Mail::SPF passes the test suite, and I'm sure I have already convinced at
> least one or two other participants of the SPF project.

Try to remove the "-Plocal" and give it one more try, please.
Here it works with and without "-Plocal", but maybe in your environment 
it is not possible to have an offline build for this.

>>> That's good, but unfortunately it was too difficult (for me, who had
>>> first given it a try) to verify that.  See my previous mail.
>> The same difficulties I find with Mail::SPF testsuites and pyspf
>> testsuites..
> 
> What difficulties?  I'd very much like to make testing Mail::SPF as easy
> as possible.

The difficulties of running a perl environment on my machine.
Not knowing perl is not so easy as you describe it in the INSTALL, but I 
understand this is extremely easy for a perl developer that is the 
target of your product. For a java developer is easy to run tests for 
jspf: there is nothing more standard than our tests in a java world. If 
they don't want to use maven they simply can use any junit tester around 
the world or an IDE supporting junits (I use eclipse).

>> So you simply tell me that to be listed in the certification you want
>> need a perl or python test suite or you won't list it?
> 
> No.

That's good to know.

>> What happens if I create an implementation for an operative system you
>> don't have access to? No certification??
> 
> No, in that case I (or someone else on the project) will first try to get
> access to a platform where it can be run (you could provide such access). 
> If all fails, I (or they) will probably certify compliance based on some
> cursory evidence (such as a log file demonstrating that the implemen-
> tation passed the test suite).
> 
> My personal view aside on the requirements for listing on the website an
> implementation as fully compliant, I am not the owner of the SPF project.
> It's very much possible that someone else capabable of editing the website
> is less demanding.

That's ok. Just put yourself in my shoes and look that the OpenSPF site 
keeps telling the only compliant implementations are pyspf and Mail::SPF 
and you can find in this list I already tried in past to have it listed.

Stefano