Date: Sun, 03 Aug 1997 12:00:37 -0700 From: "Jordan K. Hubbard" <jkh@time.cdrom.com> To: asami@cs.berkeley.edu (Satoshi Asami) Cc: andreas@klemm.gtn.com, current@FreeBSD.ORG Subject: Re: Make this a relese coordinator decision (was Re: ports-current/packages-current discontinued) Message-ID: <2920.870634837@time.cdrom.com> In-Reply-To: Your message of "Sun, 03 Aug 1997 06:39:20 PDT." <199708031339.GAA01694@blimp.mimi.com>
next in thread | previous in thread | raw e-mail | index | archive | help
> * Figure out, if ports are only made for stable ... If -current > * does more and more incompatible changes to -stable ... Who > * should done the fine work of porting perhaps 1250 ports > * to -current, which will become 3.0-RELEASE and 3.0-STABLE after > * that ??? > > That is a very good point, one that I should have mentioned in my > first post. That would be done during the lengthy BETA period, or perhaps sometime before that even starts. Again, to reference my previous posting on this, trying to track -current on an ongoing basis is just SILLY and it's only bound to get sillier as -current moves faster and faster. To make a poor analogy, if I've missed a train at the station do I run after the thing like a fool, yelling for the next 15 miles that it stop, or do I walk back to my car and drive to the next station in order to await the train's arrival there? I don't think that is a difficult question to answer, and similarly I see no point in trying to track -current as an ongoing thing. I'm not saving *any* work by doing so, I'm simply ensuring that each and every bump in the road is something I'm going to hit. By waiting until -current has gone through whatever unheavals we have planned for it and *then* going through the ports collection to see which ports are still relevant and which have build problems in whatever brave new world we've built, that strikes me as a far more sensible policy and one which will not leave the ports team stuck in some insane asylum once we start seriously beating the crap out of -current and shuffling the components around into new configurations. To put it another way, if something as minor as a new version of TCL going in screws up the ports collection this badly then we've already established that it's a frail mechanism in the face of changes in the base OS and we should no longer try and bend that mechanism to such an extent. TCL is just the tip of the iceberg where changes to -current are concerned, and trying to track multiple architectures (each of which needing its own package collection and potentially some ports changes) is only going to add to the misery. Let's stop trying to chase the moving train here, folks - Satoshi's decision was absolutely the right thing to do, even if for the "wrong" reasons! ;-) Jordan
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?2920.870634837>