[openstack-dev] [nova] Call for Help -- OpenStack API XML Support

David Kranz david.kranz at qrclab.com
Fri Aug 10 12:57:55 UTC 2012


I hesitate to wade into this but here are a few more things to consider.

1. As the primary instigator of the recent statement on API stability, 
obviously what is documented and/or implemented in the current API should
     continue to work. Whether XML support should be in the core of 
future versions is a different story (see below).

2. I don't know why George insists that "developer laziness" or 
"boredom" is responsible for the current situation. If these developers' 
employers said XML
    was important they would give it more priority. We can only conclude 
that there are not many employers saying that to their employees. Is 
there any
   company invested in OpenStack putting resources on this?

3. We are talking about nova here. None of the other OpenStack APIs use 
XML but I did see some middleware in keystone to translate back and forth.

4. Most users of the APIs never want to see unparsed JSON *or* XML. To 
them it is
    really more a wire protocol and they will either use some language 
binding, or write their own, or use the cli.

This war goes on. Amazon uses XML. The new Google Compute API uses JSON. 
These companies have VPs of product management and engineering and 
make     choices.   No matter what the nova team ends up doing with XML, 
there are not going to be in-core, spec'd XML bindings for the other 
APIs. New projects are not going to implement both. It is an intriguing 
possibility that there could be some shared middleware for translation 
that could be used by all projects to handle this issue. OpenStack needs 
to make choices too.   I think the PPB really needs to consider this issue.

  -David


On 8/10/2012 8:21 AM, George Reese wrote:
> Just to be clear on what I mean by the truth...
>
> You do realize it is 100% unacceptable in any non-clown development 
> shop to...
>
> * Not write XML support for an API that is documented as JSON and XML...?
> * Not write test cases for both the JSON and XML...?
> * To consider altering the spec of a product in the wild in a way that 
> breaks compatibility because developers are bored with the feature...?
>
> Any one of the above items is clownish. Taken together, they are 
> representative of a pattern of software engineering defended largely 
> by the employees of Nebula that is much more offensive to this project 
> than anything I have ever said.
>
> But you keep on focusing on what I am saying.
>
> -George
>
> On Aug 10, 2012, at 12:28 AM, Mark McLoughlin <markmc at redhat.com 
> <mailto:markmc at redhat.com>> wrote:
>
>> On Thu, 2012-08-09 at 18:20 -0500, George Reese wrote:
>>> It goes to the complete clown show that Nebula team is running with
>>> respect to OpenStack development.
>>
>> Tone it down on this list or we'll have to start actively enforcing our
>> code of conduct[1]. Do you want to be the first person censured under
>> it? We could name the "be respectful" clause after you if you like?
>>
>> These kind of attacks are completely out of line on a mailing list where
>> people are trying to work together and get stuff done.
>>
>> Mark.
>>
>> [1] - 
>> http://wiki.openstack.org/Governance/Foundation/CommunityCodeOfConduct
>>
>
> --
> George Reese (george.reese at imaginary.com 
> <mailto:george.reese at imaginary.com>)
> t: @GeorgeReese               m: +1(207)956-0217               Skype: 
> nspollution
> cal: http://tungle.me/GeorgeReese
>
>
>
>
>
>
> _______________________________________________
> OpenStack-dev mailing list
> OpenStack-dev at lists.openstack.org
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20120810/acebd06d/attachment.html>


More information about the OpenStack-dev mailing list