Microservice Resource's create and delete

[br2490]

I think this is more of a 'we want higher-level services to be responsible' question, but:

Using the resource service I can create: many objects using the same text/turtle (read: nfo:uuid) and still get a 200 reply.

In the above case fcrepo creates one "valid" resource with the nfo:uuid and for each subsequent you get junkers (no nfo:uuid).

Further, if you then try to delete said resources at least as it is coded now you will get an exception, even though the valid resource was deleted.

This came up because I'm working on #150. I think this is a known issue but I thought better to post; please close if it is already known.

[ruebot]

@br2490 what is the exception that you get?

[br2490]

$app->abort(503, '"Chullo says Fedora4 Repository is Not available"');

[br2490]

I think we might chalk this one up to noob. Hang tight.

[DiegoPino]

@br2490 we had this discussion at the beginning of the project and i think i even put a comment in the first versions. We should not allow same UUID to be used twice, but it should not fail right now(?). I think there a lot edge cases we need to handle now and this very interesting conversation for tomorrows call. The questions is indeed this: "we want higher-level services to be responsible?" I would say yes: even Proxy's and PCDM stuff should really be handled at a higher level. But UUID is resourceServices tasks and if this is failing we need to fix it.

[DiegoPino]

$app->abort(503, '"Chullo says Fedora4 Repository is Not available"')== means fedora4 is down or chullo is giving non sense back...

[br2490]

Fedora isn't down though.

[br2490]

and as it is programmed now will it accept the ttl but not just create a stub?

[DiegoPino]

@br2490, sorry got lost. Is this error happening on your #150 code or in the base, already published one?

[br2490]

I think this is because I'm running a 'dirty' Vagrant after some Windows issues. I'll push to a new branch and retry after a fresh Vagrant.

[whikloj]

This error also can mean, bad configuration. We should be verifying a connection on instantiation or die.

[br2490]

is that the $tx checking code I see you use in CollectionServices?

[DiegoPino]

TX are transactions (url argument).

[br2490]

@DiegoPino thank you :) so tx are args! And I think we can close this and open a new ticket about Vagrant and symlinking on Windows. I've had about enough of Windows...