120 messages in com.perforce.perforce-user[p4] Perforce vs. Clear Case
FromSent OnAttachments
George Van Treeck19 Jan 2001 09:55 
Ines Heinz19 Jan 2001 11:00 
Meyer, Maurice20 Jan 2001 08:53 
Hoff, Todd22 Jan 2001 06:29 
John Bridges22 Jan 2001 06:49 
Hoff, Todd22 Jan 2001 07:15 
Richard Geiger22 Jan 2001 08:47 
William22 Jan 2001 09:31 
John Bridges22 Jan 2001 09:33 
Hoff, Todd22 Jan 2001 10:04 
William22 Jan 2001 10:24 
William22 Jan 2001 10:37 
Richard Geiger22 Jan 2001 10:38 
John Bridges22 Jan 2001 10:57 
Arnt Gulbrandsen22 Jan 2001 11:30 
Steve Bennett22 Jan 2001 12:16 
Michael Go22 Jan 2001 12:54 
William22 Jan 2001 13:28 
John Bridges22 Jan 2001 15:52 
Lee Marzke25 Aug 2005 21:31 
Jeff Jensen26 Aug 2005 05:24 
Stephen Vance26 Aug 2005 05:47 
Monica Sanchez26 Aug 2005 06:27 
Monica Sanchez26 Aug 2005 07:19 
Jason Swager26 Aug 2005 08:18 
Christian Goetze26 Aug 2005 11:29 
Jeff A. Bowles26 Aug 2005 15:19 
Peter Buckley26 Aug 2005 15:31 
Lee Marzke26 Aug 2005 15:51 
Richard Willis26 Aug 2005 16:36 
Robert Cowham26 Aug 2005 23:36 
Ivey, William28 Aug 2005 09:34 
Jeff Grills29 Aug 2005 06:26 
Robert Smallshire29 Aug 2005 06:31 
Robert MacMunn29 Aug 2005 07:33 
Robert Smallshire29 Aug 2005 07:52 
Jeff Grills29 Aug 2005 08:05 
Robert MacMunn29 Aug 2005 08:10 
Alexandre Fabre29 Aug 2005 08:14 
Jeff Grills29 Aug 2005 08:14 
Robert MacMunn29 Aug 2005 08:20 
Jeff Grills29 Aug 2005 08:24 
Smith, Jeremy R (CACI)29 Aug 2005 08:26 
Robert Smallshire29 Aug 2005 08:35 
Robert MacMunn29 Aug 2005 08:44 
Jeff Grills29 Aug 2005 09:05 
Raja Aluri29 Aug 2005 14:20 
Arnt Gulbrandsen29 Aug 2005 14:41 
Stephen Vance29 Aug 2005 14:48 
Stephen Vance29 Aug 2005 14:54 
Smith, Jeremy R (CACI)29 Aug 2005 15:11 
Stephen Vance29 Aug 2005 15:25 
Christian Goetze29 Aug 2005 17:37 
Jonathan Biggar29 Aug 2005 18:40 
Jeff Jensen29 Aug 2005 20:00 
Jason Williams29 Aug 2005 20:49 
Jeff Grills29 Aug 2005 21:33 
Richard Willis30 Aug 2005 00:10 
Robert Cowham30 Aug 2005 01:04 
Stephen Vance30 Aug 2005 05:22 
Jeff Grills30 Aug 2005 05:51 
Smith, Jeff30 Aug 2005 06:00 
Stephen Vance30 Aug 2005 06:10 
Jeff Grills30 Aug 2005 06:27 
Jeff Grills30 Aug 2005 06:41 
Rene Medellin30 Aug 2005 07:43 
Stephen Vance30 Aug 2005 08:12 
Jeff Grills30 Aug 2005 08:17 
Ivey, William30 Aug 2005 08:36 
Stephen Vance30 Aug 2005 08:48 
Jeff A. Bowles30 Aug 2005 09:18 
Jeremy Russell30 Aug 2005 09:35 
wol...@datos.de30 Aug 2005 10:10 
Monica Sanchez30 Aug 2005 10:25 
Monica Sanchez30 Aug 2005 12:08 
Smith, Jeff30 Aug 2005 12:57 
Smith, Jeff30 Aug 2005 13:13 
Smith, Jeff30 Aug 2005 13:18 
Smith, Jeff30 Aug 2005 13:21 
Christian Goetze30 Aug 2005 13:26 
40 later messages
Subject:[p4] Perforce vs. Clear Case
From:Richard Willis (rich@gmail.com)
Date:08/26/2005 04:36:08 PM
List:com.perforce.perforce-user

Hi Lee,

I've used Clearcase for about 5 years, P4 for 1yr. Some points, a couple of which answer you questions. Pro's and con's are from the Clearcase view.

1) Pro: Clearmake. Fantastic. Clearmake knows _every_file_ that is used in a build since Clearcase is in total control of the file system. If the developers are doing similar builds, then Clearmake will fetch an object file from another developer if it matches the source in the current developers view. Build time and the need for many compiler licenses diminishes.

2) Pro: Branches. This is the BIG difference between P4 and Clearcase. Branches in Clearcase are wonderful. Branches in Clearcase are NOT directories. A view spec cites which version of a file the client is interested in and it's that version that appears in the directory. Clearcase branching is like nothing else and nothing else comes close to it's brilliance. Be prepared for a steep learning curve.

3) Pro: Config spec. In P4 it's awkward to sync to certain versions of some files, and different versions of others, especially in sub-directories. Branching and the config spec in Clearcase is a clear winner.

4) Con: Have list. Clearcase hasn't got one of these and they damn well should. It's wonderful. As a result, working remotely from a VOB is a real pain and awfully slow. We had a special tool to sync a UNIX VOB to a local Windows directory, but the timestamp was the check-in time stamp which made building in Windows troublesome. P4 defaults to the sync time, so the build system detects all changed files easily. With no have list, syncing to a VOB is slow. I'm in Aus and the P4 server is in the US - syncing isn't a problem, even for our sites in Japan, UK etc. We all sync to the server in the US. Each of your 5 sites would need a local VOB, while with P4, I suspect one depot would be enough.

5) Con: License cost: I'm not sure of the exact amounts, but I'm pretty sure a Clearcase seat costs significantly more than a P4 seat. However, you will save compiler license costs though if you use Clearmake).

These are the major issues. If you can afford the cost per seat and VOB at each site, Clearcase with Multisite is the best bar none. If the cost is too high and there are remote workers (ie: you just want one VOB), P4 wins the business.

Cheers, Rich :-)

----- Original Message ----- From: "Lee Marzke" <lee at marzke.net> To: <perforce-user at perforce.com> Sent: Friday, August 26, 2005 1:31 PM Subject: [p4] Perforce vs. Clear Case

I've got a client who is about to choose between Clearcase and Perforce in the next week or so, and I don't have any experience with Clearcase. Since I haven't been able to find any recent comparisons on the list later than 2001 or so, hopefully someone will have more current info.

I'm not interested in migration info, but any (hidden) license costs, administration overhead, expense of the hardware. Estimate 150 users, about 5 sites worldwide with shared development worldwide in a heterogeneous environment. Large code base

50K files with

multiple targets architectures per codeline. System already has a Gmake distributed builds and automated nightly builds.

Does Clearcase have any advantages here, if the make system is already well done ?

Have the Clearcase heterogeneous problems working with Windows clients been fixed ?

Also, developers on Solaris workstations don't have enough local disk space to for all the source - which is one reason they are looking at using VOB's as a backing-store for all the code modules that haven't been changed. Does ClearMake support using local files in preference to a backing source tree out of the box ?.

I know the obvious answer is to add disk space for the Perforce solution, and this would seem to be much, much cheeper than choosing Clearcase.

Thanks,