Source Code Management when filenames continually change

[verytidyboy]

Hi

Your software interests me strangely.
I work in simulation modelling and, as our models develop or a client asks us to test the effect of different settings, we continually change the name of our model and experiment files. This is an embedded part of our process and very difficult for us to change. Equally we have trouble manging code changes. Is there any way to use fortress in this way? I have tried looking at other source code management products but the changing filenames is a real stopper?

BR
RC

[lbauer]

When the file is renamed in the Vault repository, it is easily tracked, because it has an ID number that stays the same even through renames. However if the file is renamed in the working directory, the Vault client sees this as a new file. It might be possible to continually rename items in the repository to match the contents of the working directory.

Perhaps if we had a better idea of your workflow, and examples of how filenames change, we could offer a better suggestion.

[verytidyboy]

Thanks for the reply Linda.

Well that is better than expected/previous experiences. :=>

We basically keep a model file and an experiment file.
The model file "Name1.mod" contains the operational code and the experiment file "Name1.exp" contains most of the data and some static expressions. These are then compiled and run as a case. eg Name1

Sometimes we grab a .mod and .exp file and modify the .exp file (the data settings). then we rename them as "Name2.mod" and "Name2.exp" and compile and run them to create a case "Name2".

Other times we have to enhance/change/bug fix the logic so we grab a .mod file and a .exp file, modify the .mod file and then rename them "Name3.mod" and "Name3.exp" and compile and run them to create a case Name3.

The results we get are often multiple files of the data that changed over the life of the run. We mostly zip everthing up eg "Name3.zip"

So at any time we can grab an old zip file, open it and re-examine some of the results or use the .mod and .exp file to build from or to compare with a different case to check which settings/code has changed.

Anyway I will download and try your demo personally and see if I can get it to do what I want.

If I stored the ID number in the model and experiment files then could I book in a file that I had renamed externally by telling fortress that these files belong to this ID. EG

I book out Name1.mod and Name1.exp, change them, name them Name2.mod and Name2.exp and then run the case etc.

Can I book Name2.mod and Name2.exp back into fortress by telling it the ID number so it knows that these are directly descended from Name1.mod and Name1.exp?

I realise this is an odd way to use it but we have odd jobs!

Best Regards

Richard

[lbauer]

No, Fortress applies the ID in the database. You'd need to rename before you checked in. However, you could add a comment with the other name, or label the files or the folder so you can easily retrieve them later through a history search or Get label operation. Labels are tags you add to files or folders that mark the state of the item at a particular point in time. Usually, labels are used to mark builds or releases, but you can label for any reason.

[verytidyboy]

Hi Linda.

So I can script it so that when I check in it saves the old name inside the file and renames it to the common name?

And when I check the end of that branch out again it finds the name in the file and renames it?

If I can't script this then I may as well be keeping records in a spreadsheet table manually, which doesn't work well.

BR

RC

[lbauer]

You might be able to. Here's the documentation for the Command Line Client, which has a Rename function.

http://download.sourcegear.com/misc/vau ... nt/clc.htm