> > 1. Run "p4 changes -m1 -s submitted //depot/codelinename/..." > > to get the changelist number for the most recent change to that > > subtree as of now. > > Remember it in your script and version.h/version.java file. > > 2. Then run the command "p4 sync //depot/codelinename/... at THATCHANGE" > > to pull the code down. > > I understand why this works. My question is, does the following work? > > > First I do: > > p4 sync > > > > Then I do: > > p4 changes -m1 @myclient > > and write down the changelist number (call it X). > > > > Later on, I can do p4 sync @X to get the same set of files. > > Why would I want to do this, when I could just use your procedure? > Maybe no good reaason. Maybe I just want to have a deeper understanding > of how Perforce works.

the "p4 changes -m1 @myclient" is very expensive to do if a client is very large. For some of ours, it came out to about 50 seconds of server time vs. about 1/2 (probably much less)sec for a simple p4 changes -s submitted -m1

In point of fact, a simple p4 changes -s submitted -m1 will work just fine to determine a changenumber that can be used to reproduce your build. The changenumber it returns does not even have to apply to the set of files in your client! It *is* nicer to have the exact changenumber for your client. I ended up doing this by doing a p4 files on the client and finding the largest changenumber.

Its possible that a lot of this has been speeded up in recent releases...

dave