on 10-06-2015 11:00 AM
Hey everyone!
I'm searching for a decent way to handle Powerbuilder Git integration, but so far I'm not quite pleased with my findings.
In my perfect world I would like to handle the version control using only Sourcetree, while Powerbuilder will only be used to alter the code.
However, since Powerbuilder use binary files (.pbl) it becomes a problem, as I cannot see what's been changed in the code except for the whole .pbl file.
Got any suggestions on how to handle Powerbuilder code in Git?
Best regards,
Martin
You can use a Microsoft SCC compatible plug-in like the one from PushOK that integrates into the PowerBuilder IDE.
PushOk Git plug-in for SCC API enabled IDE
hth
Arnd
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Yes, but say there's a good .pbl merger tool. Then there would be no need for Powerbuilder to handle the version controls as we could use Sourcetree and open the files with said tool for merging and to see what the difference is, thus we can get past the check-in/check-out.
I'm trying to make powerbuilder unaware of any version control as the options for Powerbuilder are few and old. That's the sort of solution that probably would be the closest to my ideal.
Any ideas?
Adding PBLs to source control will prevent you and your team from tracking changes and revert changes on object base. Any change on a PBL will cause a new version of ANY object in the library. This will pump up your Source Control system.
In a perfect world the IDE would work on plain source files (like PB .NET does) and PBLs are only needed to keep a pre-compiled version in background.
Until this, PushOK is the best option.
>>Yes, but say there's a good .pbl merger tool.
Well, you don't put the PBLs under source control. You put the files within them under source control. That's why PB needs to handle it. You have to get the modified objects into the PBL and get any changes you make in PB to the object out of the PBL and to the source controlled individual file.
There is a big problem with what you want to do. If another developer checks out an object, changes it, and checks it in, the version you have in your PBL is out of date. Lets say you use Sourcetree to get the latest version of the .sru file. You still need to import that into the PBL before you can start working on it in the PB IDE.
All you are doing is adding a bunch of manual steps that can easily go wrong. Why not do it like every other developer using source control with PB does it? Sure, it will be different from how your non-PB development is done but at least it will be simple and error free.
User | Count |
---|---|
89 | |
10 | |
9 | |
9 | |
9 | |
6 | |
6 | |
5 | |
5 | |
4 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.