Outsystems Mentor is great at building the first version of an app. However, technical debt still accumulates when human modify AI generated logic. The Future: " Self-Healing "Logic Dashboards. The Need: An AI service that constantly scans ODC environment for " Architectural Drift". If a developer creates a circular dependency or sub -optimal SQL query, the AI shouldn't just "flag" it - it should generate a "refactor merge request" automatically.

In some use cases we want to postpone the loading of (parts of) the screen until all data sources are loaded. E.g. To avoid screen flickering or that there is a dependency with all the fetched data.We now have a CheckIsAllDataLoaded action in all aggregates on after fetch that sets a local variable IsAllDataLoaded= IsAllDataLoaded OR (Aggr1.IsDataFetched AND Aggr2.IsDataFetched)Would be great to have a an event thats triggered when all data is loaded.

Would be nice to have Markdown support in the Comment blocks inside the logic. This would enable better communication of significant implementation logic.Additionally, if the Description attribute could accept Markdown, it would greatly benefit documentation. This way, more complex explanations of how to use certain actions and parameters could be provided.

 Background & Current Behavior On OutSystems 11, we were able to query an aggregate view of Users and their Roles and/or Groups (both directly assigned roles and those inherited via group membership) in a single transaction. In ODC, many of the system entities are obsolete including Role. To replicate some of our page, we must now use the newly introduced User Access Management APIs. Get /users had been promising. It can return the list of users' based on the parameters provided. It could have been the best alternative if when passing "assetKey" (applications id), it can return the users with the roles for that application. However, while it does return the users with directly assigned end-user roles, it omitted users where their end-user roles were only inherited from a group. To work around this limitation, we must now: Call Get User Groups for each user For each group, call Get Group Roles Merge and dedupe results in-memory This multi‐step approach leads to an exponential number of calls when you have many users and groups, causing unacceptable latency and resource consumption. Why This Matters Performance Degradation N users × O groups per user + P roles per group Dramatic increase in response times and network traffic Increased Complexity & Maintenance Custom merging logic Higher risk of bugs, especially around pagination, retries, and deduplication Business Impact Current workaround severely degrades performance Enhancement & Expected Behavior Extend the Get /users endpoint so it returns each user’s full set of roles, direct and inherited in one call. We can also add backward compatibility by adding new parameter such as "includeInherited" to identify whether to continues to return only direct assigned roles.

in case use drag table into the flow , outsystems create list & details page in pages , mendix for example allow user to select the style of create and update in popup or in new page

It would be great to open a _CW module and have a way to see all blocks that have custom CSS on it. This will help code reviews to be quicker. Also to refactor the application theme quickly gathering all CSS scattered throughout blocks. Suggestion, look at the attached image. Between the Your block custom CSS and the application there we could have a dropdown with a list of blocks that have a custom style sheet. If you decide to do it, can we also have it done for screens? Pleeeasse? That would make my life much simpler. lol Thank you OS!

Suppose thatI have a Service module , exposing a service action (under the hood rest endpoint).I have a consumer application that references this service action.I change the name of the service module .This change does not result in an incompatible/outdated consumer in service studio.Now on run time I do get errors when I use the logic that uses the service action. This is understandable ofcourse since in the service action is just a rest endpoint (and changing the name, means the url path is changed).However, since the consumer is known, it would be so nice to have a TrueChange warning during deployment of outdated consumers.

In web (reactive & traditional) we can create screens/ web blocks, mark them public and reuse them in other application/ modules by adding its depedency.In mobile application also we can do the same for blocks but not with screens.So mobile application projects should also have capability to mark screen as public and reuse them in other application/ modules by adding its depedency.

Recently our team was responsible for implementing an Offline Data Sync process. In this project, we had one major challenge, how do we know for certain there is a sync process running in the background? We could try using sync logs and sync statuses, but if the user force closes de app mid sync, the app would be stuck on an In Progress sync log. Investigating the script OutSystemsUI.OfflineDataSync, we noticed that script contains many useful variables that are not made public to the app, most importantly, the var isSyncing. In our project, we were able to override the OfflineDataSync js object with our version of this script that includes an action to return the value of the isSyncing variable. This enabled us to implement a much more robust and user friendly background offline sync process, as it enabled us to know when should we allow the user to trigger syncs and by giving, with certain, to the user, a visual indication of the sync status. Our suggestion is that this action is made available by default on OutSystemsUI. It's a simple change to the script. Simply add the getIsSyncing function after the getSyncOnResume function.

Hi, I recently lost a lot of time when implementing an API as the top level node does not highlight an error when you include a / at the end of your base url. At the method level when you do not include a / at the start of the url path it tells you that a / is required at the start. When the baseurl and method are concatenated you end up with an extra // which results in the incorrect URL. It's quite difficult to spot. Please can you add some validation to the API node where a baseURL is added to flag if a user incorrectly adds a forward slash / at the end. I can't be the only one that has been caught out by this. Thanks John