[AIFI] Add new object type AIFI

[D052766]

No description provided.

[cla-assistant]

All committers have signed the CLA.

[cla-assistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[Markus1812]

Hi, thank you for submitting your AFF. Before we start with the review, please address the abaplint issues. They are mostly regarding unknown ABAP types, as you can only use standard ABAP types or pre-defined types from our types interface: zif_aff_types_v1.intf.abap.

For instance, /aif/proxy_outbound (char 30) can be replaced with ty_object_name_30 from zif_aff_types_v1.

Please also double-check that all titles are in title-case and descriptions follow the sentence-case scheme as described in our documentation.

"! <p class="shorttext">Integration Type</p>         <- title in title case
"! Integration type                                  <- description in sentence case
"! $required
integration_type TYPE ty_integration_type,

[GuilhermeSaraiva96]

Commenting to change status to 'Changes Requested'

[D052766]

Hi, thank you for submitting your AFF. Before we start with the review, please address the abaplint issues. They are mostly regarding unknown ABAP types, as you can only use standard ABAP types or pre-defined types from our types interface: zif_aff_types_v1.intf.abap.

For instance, /aif/proxy_outbound (char 30) can be replaced with ty_object_name_30 from zif_aff_types_v1.

Please also double-check that all titles are in title-case and descriptions follow the sentence-case scheme as described in our documentation.

"! <p class="shorttext">Integration Type</p>         <- title in title case
"! Integration type                                  <- description in sentence case
"! $required
integration_type TYPE ty_integration_type,

Hi, I have fixed now all issues that came up from the automated checks. Now that I cannot use the aif data types I think I need to add some enumerations. Also I think I need to double check whether all fields marked as required really need to be. Hope it is fine to have the pull request in place already so you can start the review!?

[D052766]

I have added some enumerations, moved sap and raw structure to General Information and removed several "required" annotations. I think there should be no more changes until your first feedback now. Any chance of an estimation when that may be?

[D052766]

Related to feedback I got from Katharina, I´ve just committed another update on the structure.

[Markus1812]

Hi, thank you for updating the files. We'll perform a review tomorrow.

[Markus1812]

Hi, sorry for the delay. In general, the AFF looks really good, but we have a few questions:

1. Do you want to use prefixes like is_ or uses_ for booleans? We have marked some already and some already use prefixes. Maybe you can take a look again on that.
2. Do you want your comments containing the name of your domains/database tables in a public repository?

[D052766]

I think all comments were answered and adjustments are done.

[Markus1812]

Thank you for adapting our suggestions. We have some more questions.

[Markus1812]

Thank you again for adapting your AFF. We have some small comments again.

[Markus1812]

In general, the AFF looks good to me, but I have one further question for clarification.

[Markus1812]

Thank you for answering all our questions. The AFF looks good to me, you can now continue and request the ux-review when you are ready for that.

[D052766]

I just removed the enumeration for the integration type. We have to supply this as custom value help from the backend. Please confirm the UX-REVIEW readiness is still given.

[Markus1812]

@D052766 Yes sure, thats okay for me. Is there are reason behind this change? Maybe this would be interesting for the ux colleagues as well.

If your concern is the extensibility of the enum, our wiki states that adding values to the enum is compatible without increasing the format version, as long as there is a default value (which is the case).

[D047539]

@Markus1812 We prefer to set and retrieve values in the backend, rather than defining them in the frontend.