I like this proposal.  It's consistent with the current platform support 
guidelines,
and it adds a way to clearly delineate experimental APIs.   I also think it's 
complementary
to Mats's suggestions of how to annotate experimental APIs.

I have no preference between the term "experimental" vs "preview" though.
I say that because there may be some people that prefer the term "preview" 
instead.

Dave

> -----Original Message-----
> From: Daniel Mihai
> Sent: Thursday, April 6, 2017 12:09 PM
> To: Mats Wichmann <mats at wichmann.us>; Dave Thaler
> <dthaler at microsoft.com>; uzchoi at samsung.com
> Cc: iotivity-dev at lists.iotivity.org; 'Wouter van der Beek (wovander)'
> <wovander at cisco.com>
> Subject: Public and Experimental Public C APIs
> 
> Should we start with the following definitions?
> 
> 1. All C functions included under out/<path_to_IoTivity_SDK>/ are Public
> APIs
> 2. All C functions included under out/<path_to_IoTivity_SDK>/experimental/
> are Experimental Public APIs
> 3. When an Experimental Public API gets promoted to non-Experimental
> Public API, that API SHALL be moved out of
> out/<path_to_IoTivity_SDK>/experimental/ header files
> 4. Apps should expect potential breaking changes to Experimental Public
> APIs, but no breaking changes to non-Experimental Public APIs
> 
> BTW, a couple of days ago I said that we were not even close to ready to lock
> down the Public APIs. Since then, I've learned that all or most of the
> problematic APIs that I had in mind are not yet under
> out/<path_to_IoTivity_SDK>/. Given that additional information, and if we
> apply the definitions I proposed above, locking down the current non-
> Experimental Public API sounds a lot more reasonable to me. We still get the
> chance to improve the problematic APIs, before they become non-
> Experimental Public APIs.
> 
> Thanks,
> Dan
> 
> -----Original Message-----
> From: iotivity-dev-bounces at lists.iotivity.org [mailto:iotivity-dev-
> bounces at lists.iotivity.org] On Behalf Of Mats Wichmann
> Sent: Tuesday, April 4, 2017 5:00 PM
> To: Dave Thaler <dthaler at microsoft.com>
> Cc: iotivity-dev at lists.iotivity.org
> Subject: Re: [dev] Don't make breaking changes
> 
> On 04/04/2017 03:25 PM, Mats Wichmann wrote:
> 
> > It's not that hard to add your own definitions to the Doxyfile.
> 
> I ended up using some html, but cleaner approaches might be possible:
> 
> https://na01.safelinks.protection.outlook.com/?url=http:%2F%2Fwww.stack
> .nl%2F~dimitri%2Fdoxygen%2Fmanual%2Fcustcmd.html&data=02%7C01%7C
> Daniel.Mihai%40microsoft.com%7C038dfeaf6cd844628d2c08d47bb6bbc2%7C
> 72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C636269472151475201&sda
> ta=xpxhfRX60H7ZGau1iXxjHqZ8zVM0qkFFSflrxtri5So%3D&reserved=0
> 
> anyway, I agreed with the comment that it's not great for a developer if
> some stuff in a header is fine and other stuff is experimental. Better
> if we could section it, though depending on the scale of the feature
> that may not be easy.
> 
> Indeed, none of this is terribly easy, it's a maturity thing and much as
> we'd like to think otherwise, from the app-developer viewpoint, iotivity
> isn't really that mature.  To be balanced, it's way better on the device
> side, as there what's interesting is "does it respond to the OIC/OCF
> protocol correctly", and we can actually answer and track that question
> with some degree of confidence through the CTT and other tests.
> _______________________________________________
> iotivity-dev mailing list
> iotivity-dev at lists.iotivity.org
> https://na01.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.io
> tivity.org%2Fmailman%2Flistinfo%2Fiotivity-
> dev&data=02%7C01%7CDaniel.Mihai%40microsoft.com%7C038dfeaf6cd8446
> 28d2c08d47bb6bbc2%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C6
> 36269472151475201&sdata=LPFrqVu5nyw2xby%2FkNxSZ7Z7ad9vsnL%2BdyU
> bP7Pvwy8%3D&reserved=0

Reply via email to