How do you guys organize the source code?

[Rodgers]

I am curious on how you guys manage the source code. As you can imagine, I would like to hack the code, and keep updated with the new releases, but does not have the check in privilage.

The way I found works like this:

1. download a copy of the OpenSees source code
2. import the souce code into a local repository, with vendor as PEER and release tag as the current release number
3. check out a working copy into a separate folder and start hacking it
3. commit whatever changes I made to the local repository
4. when a new release come out, download it and import it into the local repository, with vendor as PEER and the new release tag
5. merge the difference between the two releases into my working copy

is it too much hassel?

The problem is, I lose the original revision number for each file in my working copy, they all start from 1.1 all over again.

Any suggestion?

[fmk]

the following might work.

you might be able to use both repositories if you keep seperate CVS directories .. when you want to synch with one repository switch the CVS directory to be the one for that repository .. i.e. keep 2 CVS directories CVS and a CVSmine or CVSopensees, which one the other is depends what dir CVS is .. you can write a simple script which would recursivly go through the code and change the directories.

[Rodgers]

I tried but it does not seem to work.

From what I understand, it is very difficult if not impossible to keep the original revision number in my working copy.

This is because CVS relies on some logs inserted to the beginning of each file to track the revision history. But the OpenSees copy checked out of the PEER repository only has the final revision number.