openstack team mailing list archive

[Mark Nottingham <mnot@xxxxxxxx>]

On 28/10/2011, at 12:31 AM, Sandy Walsh wrote:

> I think that would be ideal.
> 
> The key to API adoption will come not only from ease-of-use but from availability of language bindings. Manually creating and maintaining the Python binding is well and good, but having to create and maintain bindings for Java, C#, C++, C, Ruby, PHP, etc is going to be a full time job for many parties. Not to mention versioning support.
> 
> The beauty of having a machine consumable format is to automate the changes in the bindings as soon as the WADL (or whatever) changes. All clients move lock-step with the server changes. As I mentioned before, it's even better when the server WADL is auto-generated from the implementation.
> 
> Yes, there are technical issues around this. The greatest being the "sugar" on the user interface. For example, knowing that something is a password or calendar field. Or using the output from one call as input to another. But for basic <function> <parameters> -> <response> type calls, I know this is do-able. At the very least we should be able to create a nice object model for clients to use. 
> 
> Is WADL rich enough to support this? That's what we need to investigate.

The problem is that people are used to using WSDL and WADL as ways to do code generation, not using them dynamically at runtime. This makes everything very brittle.

I'm totally on board with having the interface being machine-consumable at runtime -- see the previous discussion on versioning and extensibility -- but WADL isn't really designed for this. I'm sketching up something more appropriate, and will be able to talk about it soon (hopefully).

Cheers,


> 
> -S
> 
> ________________________________________
> From: Mark Nottingham [mnot@xxxxxxxx]
> Sent: Thursday, October 27, 2011 10:16 AM
> To: Sandy Walsh
> Cc: Mellquist, Peter; Joseph Heck; openstack@xxxxxxxxxxxxxxxxxxx
> Subject: Re: [Openstack] +1,  All services should have WADLs
> 
> On 26/10/2011, at 11:17 PM, Sandy Walsh wrote:
> 
>> As discussed at the summit, I agree there should be some form of IDL (WADL being the likely candidate for REST), I think manually crafting/maintaining a WADL (or XML in general) is a fools errand. This stuff is made for machine consumption and should be machine generated. Whatever solution we adopt, we should keep that requirement in mind.
> 
> I've stayed out of the discussions about WADL mostly, because up till now they've been centred on creating documentation and running tests. As far as that goes, these are reasonable things to do with it.
> 
> What do you mean by "machine consumption" -- are you saying that you want clients to automatically generate bindings?
> 
> Cheers,
> 
> 
> --
> Mark Nottingham   http://www.mnot.net/
> 
> 
> 

--
Mark Nottingham   http://www.mnot.net/