From owner-freebsd-current Mon Sep 2 23:19:24 1996 Return-Path: owner-current Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id XAA07799 for current-outgoing; Mon, 2 Sep 1996 23:19:24 -0700 (PDT) Received: from eel.dataplex.net (eel.dataplex.net [208.2.87.2]) by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id XAA07794 for ; Mon, 2 Sep 1996 23:19:22 -0700 (PDT) Received: from [208.2.87.4] (cod [208.2.87.4]) by eel.dataplex.net (8.6.11/8.6.9) with SMTP id BAA08836; Tue, 3 Sep 1996 01:19:17 -0500 X-Sender: rkw@shark.dataplex.net Message-Id: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Tue, 3 Sep 1996 01:19:17 -0500 To: "Jordan K. Hubbard" From: rkw@dataplex.net (Richard Wackerbarth) Subject: Re: Latest Current build failure Cc: current@FreeBSD.org Sender: owner-current@FreeBSD.org X-Loop: FreeBSD.org Precedence: bulk >> Perhaps it's just me, but I really don't expect current to build all the >> time. It's just a fact of life... That's why it's called -current. > >Well, while I'd be the first to agree with you, it still appears to be >the case that a goodly percentage of folks want, for lack of a better >and less oxymoronic term, a "stable-current." I also think that we can >give them this and without sacrificing developer access to the "real" >-current (HEAD). > >What would help would be a concerted effort on the part of the >-current developers to move over to CVSup or CTM'd, grabbing the CVS >tree in cases where the latter is used. I also know that people have >traditionally complained that a CVS tree takes up lots of space, but >c'mon guys - disk space is disgustingly cheap now. I think we can >stop citing that as a blocking factor, just as we stopped letting 4Mb >machines or 256K VGA cards stand in the way of previous development. >Let's get all the developers moved over to the proper services they >need for doing their jobs, then we can take CTM-current and >sup-current and reshape them for the needs of those who like to stay >up to date but not *that* up to date. Just judging by the amount of >push-back I've received over the years, I think that group of folks is >larger (and louder) than you think. It's probably time we sorted them >out, if for no other reason than to eliminate the 47 copies of >"-current is broken again!" messages we currently get on -current >(or -hackers, or both) whenever something breaks. :-) In defense of those who complain, I do not think it reasonable to expect to have to do a detective job every tine they get an update simply to get the thing to compile. This is particularly true when you consider that the build potentially "trashes" the working system. However, I also agree that "up-to-the-minute" is beyond the level that most people really need. When you ask that people run "current" so that you get some testing, they expect to test the workings of the driver, etc.; not the make tree. One advantage in going to this scheme is that you can eliminate the "sup" as a separate product target. The "up-to-the-minute" folks should use CVSup. The next level, including the sup servers, can use the ctm delineated "snapshots". The users can then either subscribe to ctm directly or sup the equivalent product. The next level would be the "Jordan's snapshot" which you have given a bit more testing than just "compiles". - - - I guess I see that you and I have a different viewpoint of the "stability" of things. In your model, "current" seems to be just some arbitrary collection of code. whereas "stable" has been tested enough to make sure it compiles and runs. You seem to leave out the "production" level which is supported. IMHO, it you want to build a following for the FreeBSD OS, you need to put greater emphasis on supported stability. I think that it is this market factor that you are hearing complain.