Review Request 40: basic RESTful API

[Tim Flink]

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
http://reviewboard-tflink.rhcloud.com/r/40/#review62
-----------------------------------------------------------



blockerbugs.spec
<http://reviewboard-tflink.rhcloud.com/r/40/#comment101>

    'Required' is not a valid keyword for spec files, it's 'Requires'. Also, if iso8601 is used during unit tests, you'll have to add a 'BuildRequires'. Have you built an RPM with this spec?



blockerbugs/controllers/api/api.py
<http://reviewboard-tflink.rhcloud.com/r/40/#comment105>

    You're the author here, not me :)
    
    There are several cases of this in this patch. I'm just marking one, though.



blockerbugs/controllers/api/api.py
<http://reviewboard-tflink.rhcloud.com/r/40/#comment108>

    out of curiosity, how come you're using <rel_name> here instead of something like milestone?



blockerbugs/controllers/api/api.py
<http://reviewboard-tflink.rhcloud.com/r/40/#comment106>

    One of the things that I'm wondering about is consistency with the existing pages and whether or not this is important.
    
    Specifically, I'm questioning naming this route 'spin' when the gui uses 'spins' and whether this would count as an inconsistency for API discovery - thoughts?
    
    I'm also wondering if it's useful enough to allow GET on milestone/<rel_num>/<rel_name>/spin/<id> to get one spin instead of all spins for a release/milestone



blockerbugs/controllers/api/api.py
<http://reviewboard-tflink.rhcloud.com/r/40/#comment107>

    it seems to me that this POST might be simpler (ie, less validation) if the URI had an <id> at the end - do you think that would add or take away any utility of the API?



blockerbugs/controllers/api/api.py
<http://reviewboard-tflink.rhcloud.com/r/40/#comment109>

    I think this should be calculated at request time instead of just parsing what the client sends in - the request time should be pretty much when the request is received, anyways


It's looking pretty good - I think you're planning to add a method for spin modification as we talked about on IRC which is one of my biggest concerns right now.

- Tim Flink


On Aug. 5, 2013, 10:15 a.m., Ilgiz Islamgulov wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> http://reviewboard-tflink.rhcloud.com/r/40/
> -----------------------------------------------------------
> 
> (Updated Aug. 5, 2013, 10:15 a.m.)
> 
> 
> Review request for blockerbugs.
> 
> 
> Bugs: 392
>     https://fedorahosted.org/fedora-qa/ticket/392
> 
> 
> Repository: blockerbugs
> 
> 
> Description
> -------
> 
> Add endpoints:
>  - list bugs
>  - list updates
>  - list spins
>  - create spin
> 
> 
> Diffs
> -----
> 
>   testing/test_validators.py PRE-CREATION 
>   testing/test_api.py PRE-CREATION 
>   requirements.txt 09e0318bc189512f5d324bda8879ad74c4763f95 
>   blockerbugs/models/spin.py 99891448e78c7168d540b479ffd7ef00ce1eec1d 
>   blockerbugs/controllers/api/validators.py PRE-CREATION 
>   blockerbugs/controllers/api/errors.py PRE-CREATION 
>   blockerbugs/controllers/api/api.py PRE-CREATION 
>   blockerbugs/controllers/api/__init__.py PRE-CREATION 
>   blockerbugs/__init__.py c4f2c34f5eac713253336d026aa9485259096cb0 
>   blockerbugs.spec 726fa6920c67cfe36a2e544f97c9e0f16537f11e 
> 
> Diff: http://reviewboard-tflink.rhcloud.com/r/40/diff/
> 
> 
> Testing
> -------
> 
> Wrote test suites.
> I've tested on my develop instance.
> 
> 
> Thanks,
> 
> Ilgiz Islamgulov
> 
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.fedoraproject.org/pipermail/qa-devel/attachments/20130806/5a8838c9/attachment.html>