Reduce loaded workspace projects by new "EGL Archive" file. Merged

We try to align ourselves as much as possible to a 3-tier architecture.
This results in an architecture where a main program calls many Called programs, that can other Called programs, that can ?

In our EGL Architecture we've chosen to group the programs (both called as main programs) that belong to a logical subsystem.

The public interfaces of the called programs are stored in a project like: COM.KBC.ITF.PUB
The public called programs are stored in a project like: COM.KBC.FTN.PUB
(there are also about 10 other projects per subsystem but they are not relevant for the RFE, though they do influence the number of projects in the workspace and thus the number of EGL files)

Developers only change files of one subsystem, but in his code he will typically make several calls to publicly available called programs of other subystems.
Due to the dependencies with the Interfaces project, workspaces with over 50 projects are common for us.
Such a workspace can easily contain over 1500 EGL files.

Large workspaces have a negative impact on the performance RBD: long RBD building times, longer start-up times while debugging, ..
Whenever a full rebuild of the workspace is required we need our developers to respect a timeframe for rebuilding their workspace (25 developers per hour), otherwise our clearcase server has performance issues since 1000's of IR need to be updated per user that's rebuilding his/her workspace.


Inspired by adding JAR-files to reduce the number of sourcefiles for java development, we would like IBM to consider the option of creating something like an ?EGL Archive file? such that the EGL developer does no longer have to load projects (which he only needs in read-only mode) in his workspace when (eclipse) building, debugging or generating.

That would improve the performance of the tool, since these EGL Archives will probably not have to be built, and the workspace might be reduced to about 10 projects, or maybe 300 egl files (versus 1500).

The EGL archive could consist of a number of ir-files.
As we see it, it would be our responsibility to manage the content of these archives.
We might want to setup a batch processing that runs daily to update the content IR-files with the latest versions.




If you require further contact with us to explain certain parts, I would like to suggest to have a conference call about this.