One common scenario in various popular REST APIs is objects with dynamic attributes, and also the way empty instances of such objects are sent. With dynamic attributes, I mean something like this: "AvailableLocales" : { "nl-NL" : "Netherlands Dutch", "pt-PT" : "Portugal Portuguese", "en-UK" : "British English" } Currently, you need an On After Response and the ardoJSON extension to interpret this. It would be better if there would be an option to turn this into a key/value pair list directly. Secondly, quite often, an empty list like the above is denoted with the empty array symbol: "AvailableLocales" : [] Even if the dynamic attributes are a closed class, and you'd manage to make all into an attribute, this results in a parsing error. Instead of an error, the Platform should recognize this as an empty array or object, and not give a parse error.

Following the original thread at https://www.outsystems.com/forums/discussion/5478/how-to-optimizing-disk-usage-by-temporary-asp-net-files/ , it would be a piece of mind having this automated cleanup built-in in the Platform Server for .NET.

Microsoft has a service called Microsoft Learn MCP Server (microsoft.docs.mcp). With this service, tools like Claude Desktop, VS Code, or AI agents can get information only from Microsoft’s official documentation. Links: https://learn.microsoft.com/training/support/mcp https://github.com/microsoftdocs/mcp If OutSystems had a similar MCP server, it would be very helpful. Here are some reasons: Tools like Claude Desktop or VS Code could search only OutSystems official documentation Companies could easily add official OutSystems info to their support tools or knowledge systems Developers could build AI agents that use OutSystems official documentation For OutSystems developers, using official info through an MCP server would be very useful. I hope OutSystems will think about making a service like Microsoft Learn MCP Server.

Implement an additional functionality that allows searching for all uses of a CSS class across all consumer modules of that stylesheet. The front-end teams would greatly benefit from this measure, as it streamlines development and enhances maintenance processes.

 TLDR: CSS Layers are a way to define order of precedence in the case of multiple cascade layers. The suggestion is to implement a way to define these layers, to further improve the development of custom themes and layouts. Why?Being able to define cascade layers allows the developers to separate stylesheets into a hierarchy of priority. Quoting the MDN article:"[A cascade layer] enables the creation of simpler CSS selectors because you do not have to ensure that a selector will have high enough specificity to override competing rules; all you need to ensure is that it appears in a later layer."In practical OutSystems terms, this means that the OutSystems UI stylesheet could sit in an early layer while the Custom Theme of an application could be placed in a later layer, hence avoiding overriding conflicts with existing rules. Furthermore, in environments where a Custom Theme exists, but a superset theme is applied to specific applications, conflicts of specificity could also be avoided. How?This layer definition could be placed in a per-CSS file basis: it's just another parameter like "Name", "Description", "Public", etc. The layer order (or precedence list) could be defined in the Theme Editor, or be set in Service Center.

Currently it is possible to upload a binary file by editing the table. It would be interesting to be able to download the binary without having to create a screen with the download functions. Idea in the image below, in adding a download icon and being able to download this binary in any file.

If you're writing expressions where you're comparing a certain attribute which happens to be a Foreign Key to a Static Entity, then Autocomplete (Ctrl+Space) could show a list of the possible values.Example:How it could be:

After yet another bug with swapping an IF... could the True connector be GREEN and False RED? Thought i already posted this idea years ago, but couldn't find it...

Hi,I think it is interesting that Integration Studio saved passwords for connecting to the environment, just like in Service Studio.Or to make it easier, Integration Studio uses the same foundation as environments already saved in Service Studio (environment, user and password).

Currently, the Serialize node only provide an option to exclude attributes that have default values. However, there is no option to explicitly exclude custom-selected attributes. This capability is especially useful for logging scenarios . In many cases, creating an additional Structure just to map and exclude specific attributes is redundant and unnecessary. Most serialized outputs are eventually passed to a JavaScript node , written to logs, or stored in a database purely for logging purposes. Allowing developers to select which attributes should be excluded during serialization would significantly reduce development time, keep the code cleaner, and improve maintainability. It would also eliminate the need for extra mapping structures that add complexity without real value.