20 messages in com.perforce.perforce-user[p4] Why does the revision keep rolling?
FromSent OnAttachments
Dan Harmer02 Apr 2002 16:14 
Ken Wilson02 Apr 2002 16:22 
Mike Pagel02 Apr 2002 16:32 
Mich...@diversifiedsoftware.com02 Apr 2002 16:40 
Eric Dew02 Apr 2002 17:02 
Paul Goffin03 Apr 2002 00:17 
Paul Bayless03 Apr 2002 06:54 
Gregg G. Wonderly03 Apr 2002 08:23 
Karl Elvis MacRae03 Apr 2002 10:30 
Dan Harmer03 Apr 2002 10:31 
Karl Elvis MacRae03 Apr 2002 13:46 
Robert Prentice03 Apr 2002 15:19 
ste...@vance.com03 Apr 2002 15:46 
Paul Goffin04 Apr 2002 00:09 
Paul Goffin04 Apr 2002 00:19 
Schipper, Martijn (Contractor)04 Apr 2002 05:30 
ste...@vance.com04 Apr 2002 10:21 
Tyler, Tom04 Apr 2002 10:27 
Gordon Broom04 Apr 2002 13:40 
Jeff A. Bowles04 Apr 2002 14:21 
Subject:[p4] Why does the revision keep rolling?
From:ste...@vance.com (ste@vance.com)
Date:04/03/2002 03:46:31 PM
List:com.perforce.perforce-user

Actually, Perforce is not "built on" RCS. Rather it uses the RCS file format to store file deltas.

As for the behavior under discussion, it's simply a matter of whether the tool chooses to implement that policy decision or leave it up to the user to decide. Choosing not to increment the revision when the file is unchanged is an optimization, not a rule, although it is an optimization that makes lots of sense. Either way, you end up with a time-consistent view of the file's history. Additionally, it is a minor optimization in the Perforce environment as the additional size to the RCS file is negligible and the metadata is minor.

The recently added option in the GUI certainly simplifies implementing that optimization as a policy.

Steve

On Tue, Apr 02, 2002 at 05:03:13PM -0800, Eric Dew spake

thus:

Perforce (or any other SCM tool) cannot read your mind. If it's checked out for edit and then checked in, then it's checked in. Who knows whether you had some intention of checking in unmodified files? No SCM tool will second guess

your commands.

That's just wrong. RCS knows if the file has changed since you checked it out - and perforce is built on top of RCS. So (IMO) perforce simply does this the wrong way.

Yeah, there's a workaround (revert -a), but the way it's implemented (mindless check-in, change or no change) is backwards. There should be a setting or flag to control this.

-Karl

-- Karl Elvis MacRae VLSI CAD Apple Computer kmac at apple.com 408-974-0847 (voice) 408-313-5741 (cell)

_______________________________________________ perforce-user mailing list - perforce-user at perforce.com http://maillist.perforce.com/mailman/listinfo/perforce- user

--------------------------------------------- This message was sent using WebMail by CyberGate. http://www.gate.net/webmail