This might be true if you compare to old versions of SVN, but is no longer true. Even if it were, I’d take SVN’s branches over perforce’s any day, as branching in perforce is a major thing that is easy to get wrong, and thus you can’t use for anything but major projects. With SVN you can do all kinds of nice things (mostly which approximate a DVCS) like make a branch for every non-trivial set of changes. In perforce, no way in hell.

Perforce is so stupid! It’s company should pay me a lot of money for using it!

Thanks a lot for this post!

The problem with the sentiment of “learn how to use it” is that Perforce does such a horrible job of making it easy to learn. The commands aren’t intuitive, you never have a clue what they’re going to call something or where to look for instructions on how to do the thing you need to do (because you can’t figure what it’s called) and you pretty much have to read a book to really get a decent understanding of it. This is total failure of their human interface. Simple example: most of the time “…” is a file path wildcard (totally atypical) but sometimes it’s “*” and when you use “…” instead of “*” it tells you you should use “*” instead, when it could just @#$% accept the “…” since it knows what you wanted!

You may think that having to forcefully tell Perforce what files you want to edit is a drawback, but it actually enables Perforce to not have to scan your entire tree to see what files are opened/modified in order to know what you’ve changed. When you have tens of thousands or hundreds of thousands of files in your Perforce tree, this is a MAJOR performance win.

Sorry to have to say it, because generally I agree with rants about useability but … in this case, you’re mostly just retarded. The tool is fine, learn how to use it (which really isn’t that hard at all) and you will not have any problems.

We got a file deleted this June. Another file with the same name happened to come up last week. And got automatically deleted by perforce again at the integration stage. Outrageous at its mildest.

Perforce has been around forever and a half, and only maybe now they’ll start making it usable, due to truly serious competitors out there (not just svn, but hg, bzr, git, …), mostly in the distributed realm. It’s easier to imitate than innovate. But I’m sure they’ll fuck it up just to make the tool look complicated enough to bewilder the corporate-types who buy this shit.

It wasn’ t mentioned, but just try refactoring c++ or java code in an IDE using perforce. There’s no (or little) support for p4 in IDE’s since…well, who’d want to map this interface onto logical SCM actions. To rename a Java class I (1) do it in the IDE, just as a “preview” (2) get a list of filenames changed, (3) UNDO the change in the ide (cancel the refactoring), (4) write an insane shell script that does a rename in perforce, (5) do the refactoring, (6) copy all the old files back (script), (7) do the p4 changes.

Perforce is very, very manually intensive; I’ve got p4 shell scripts I can use on un*x, but on windows p4 acts differently & you’ve got to use p4v. I submit & switch platforms just to do certain p4 tasks. Perforce is the bane of my professional existence. I do personal projects at work just using svn (and/or hg) and merge it in the “official” p4 repo just for posterity, AFTER the work’s done. (I’d like to look into tailor, as a possibility to help out with this…..)

Perforce, f*ck you from me, too. I’m just insulted that my company thinks this is a good thing. No, I don’t need to read a book; the people who use & like perforce need to use & learn another system to compare it to, or need to actually use perforce themselves.

(Note: this should happen on UNIX)

My company pays dearly for this overpriced piece of crap. An absolute waste of money and productivity. I look forward to the day I have finally migrated all of our users to Subversion.

FUCK YOU Perforce!!!!