Best practice for VC 7 INT/GNT in Azure Devops (TFS)

We're about to switch from PVCS to Git, and to start build our projects in Azure Devops (TFS). Problem now is that we have some 200 - 300 individual VC 7 INT/GNT programs which we would like to build and deploy individually. And that seems to imply that we would need to put each and every program in a separate project, which would be rather inconvenient for all involved. Is there another solution to this?

  • To what project are you referring, Visual COBOL projects or something else? For VC for Visual Studio there is a project type for .int/.gnt. You can place multiple COBOL sources into these projects and it will automatically build each one to a separate .int or .gnt output. For VC for Eclipse you can do the same only it is a setting in the Properties under Micro Focus-->Build Configurations-->Link Settings-->Target Type set to All .int/gnt Files and then you can specify the type on the COBOL tab. These project types will build all programs at once and copy them to a specified output folder. 

  • Yes, Visual COBOL projects. But what will happen if we have a single project in Git / TFS containing all our ~ 300 COBOL files and someone commits and pushes from his/her local computer to the Git server? If there is a pipeline set up in TFS, will the pushed file be built and deployed automatically by TFS or will all files in the project be built and deployed? The latter would not be desirable.

    Or is it possible to avoid the automatic build and deploy, and start the pipeline manually instead? And in that case, do we need one single project for each of the 300 source files, or is it possible to have a single project and in some way point out which file to build and deploy?

    But maybe we try to use Git and TFS in a way you're not supposed to use it... ?