In packaging we need to know the uncompressed size of an archive, as that information is passed into package metadata for installers.

The packaging pipeline could do this per artifact, but this may require extracting the archive first. This would be very CPU and disk intensive, as this step may be needed for every integration (Qt6, and other products). Packaging is using the machines from the Open Nebula. One integration can be tens of gigabytes.

Instead, pass the "uncompressed size" from the build agent per build artifact, as this information is readily available on the build agent before it uploads the artifact to storage.

The new packaging pipeline uses the Coin's Rest API, so the "uncompressed size" information should be then included there in the json documents that the API returns.