Re: [ubl] UBL face-to-face meeting about to start

[Levine Naidoo <levine@lxndigital.com>]

Great thanks Kenneth

I am free now so will join

Levine Naidoo

AU +61 476 273 935

On 29 Sep 2023, at 18:11, Kenneth Bengtsson <kbengtsson@efact.pe> wrote:

ï

Hi Levine

 

Thanks for joining us yesterday and for staying up late!

 

To clarify a few points (which I think addresses your questions) then the discussion was entirely around how the current semantic model can evolve to improve global interoperability even further. The API discussion was simply to exemplify how the current syntax binding alone is not sufficient. No one suggested (I think!) that UBL should define an API. However, we often hear from (potential) users who want to use UBL to build APIs, and we donât have a good response to them. Previous work with alternative syntax bindings has taught us valuable lessons and has also exposed some shortcomings in the current model. Yesterdayâs discussion was about improving on the existing work product, and with APIs being one example of why and how.

 

Weâll continue the discussion today (09:30 local Irish time) if you can join us :-)

 

/Kenneth

 

 

From: Levine Naidoo <levine@lxndigital.com>
Date: Thursday, September 28, 2023 at 11:12 PM
To: Kenneth Bengtsson <kbengtsson@efact.pe>
Cc: ubl@lists.oasis-open.org <ubl@lists.oasis-open.org>
Subject: Re: [ubl] UBL face-to-face meeting about to start

Hi All

 

My apologies for missing some of the discussion on the APIs â itâs resulted in an impaired understanding on the scope of the activity.

 

Iâve had a think about the topic this morning and attempted to fill some gaps, which is certainly not a good idea :-). So, whilst I understand and appreciate the potential of standardising APIs am still not clear I understand the driver for the TC to undertake this activity and how is it going to make UBL adoption easier. As a universal business language, I was hoping that we would invest more time on how UBL would serve as a common semantic model to bridge systems that use disparate standards.  A universal semantic model should make a universal API more attractive.

 

I have a few questions to help me understand the business case and the outcomes we want to get to with respect to the API scope: 

1. UBL's Rich Model: Given the comprehensive and detailed nature of the UBL semantic model, which encompasses an extensive range of business scenarios, are we contemplating translating each individual UBL document (and full granularity) into its separate API endpoint, i.e. 93 endpoint definitions?
2. APIs as Resource Interfaces: Considering that APIs act as interfaces to specific resources, which might not always utilise the entirety of a UBL model, how will we decide what portions of a UBL document are essential for a given API endpoint?
3. Namespace Challenges: When converting UBL documents how do we plan to handle and represent these namespaces accurately?
4. Adoption support:  How do we plan to support and guide the developer community in adopting these APIs, given the intricacies and depth of UBL models?
5. User community input: As anyone in the actual user community requested it and shared why/how it would make things easier for them?
6. Supported stakeholders: Given the example I cited last night, on a vendorâs lack of support of the UBL model by any stretch of the imagination are we intending that the APIs would support the interactions between corners 1&2 and 3&4. 

 

Yes, standardised APIs could reduce the complexities and costs associated with multiple custom integrations but only if there is comprehensive adoption. Itâs also not without challenges around versioning and supporting different APIs based on who is supporting what version of UBL at any given point in time, etc. 

 

Kind regards

Levine

 

 

On 28 Sep 2023, at 7:28âpm, Kenneth Bengtsson <kbengtsson@efact.pe> wrote:

 

To join remotely: https://zoom.us/j/547975473?pwd=bjVvdDhDcEZXZ3R3cldRYkpuQU13Zz09

 

/Kenneth