Date: Thu, 23 Apr 1998 08:14:30 -0700 (PDT) From: patl@phoenix.volant.org To: Studded <Studded@san.rr.com> Cc: current@FreeBSD.ORG Subject: Re: Why is xtend in the base system? Message-ID: <ML-3.3.893344470.6561.patl@asimov> In-Reply-To: <353E542E.B3D24D08@san.rr.com>
next in thread | previous in thread | raw e-mail | index | archive | help
I thought we'd settled this. Xtend is in the base system because it uses kernel facilities that are still evolving at a rate that would make it prohibitively difficult to track from a port. We should table the issue and re-visit it once the kernel has settled down. > Actually trying to pretend that percentages don't matter is bogus. If > your argument were correct everything in the ports collection would be > integrated into the base OS since whether the majority of the users use > a given function or not is "bogus" as a criteria for inclusion. Nonsense. Almost all of the ports are strictly user land programs that are completely source-compatible across multiple OS revisions. Percentages -don't- matter either way in this decision because they aren't the issue. It's like arguing whether Ford or Chevy make better cars based on the available colors. You may as well argue that linux, SCO, and BSDi emulation, or the SYSV shared memory facilities should be yanked into ports because a lot of people don't use them. Ok, more people use them than use xtend; but exactly where and how do you draw the line? Or even get solid enough numbers to know how many people would be affected? > The question is "What kind of system do we want to ship?" and that's > been answered by consensus that we want to ship a system that the > majority (read, "unwashed masses") of users can unpack and use, and at > the same time a system that professional users can use and configure > with a minimum of hassle. (Of course there are other considerations such > as POSIX, etc., but I think you get the idea.) And having xtend in the base system degrades this model exactly how? > The sound card argument is particularly inappropriate since the only > "sound card" stuff that comes with the system is something you must > configure and compile if you want it. The xten stuff comes as binaries > in the binary distribution, is compiled with every make world (unless > you disable it like I did) and is not used by even a significant > minority of FreeBSD users. That would include the vast majority of ISP's > and WWW farms. Ah, so it is the extra overhead in 'make world' that you are complaining about. Well, I hate to break it to you; but the majority of FreeBSD users never do a make world. So if the percentages did matter, your complaint would still be irrelevant... Also, your argument didn't make it clear exactly how the xten stuff differs from the sound card stuff. If I want either, I have to tweak my kernel config. That's as much as the majority of FreeBSD users will ever see. > In conclusion I know that the chances of getting xten removed from the > base are very slim, but can we at least keep the discussion about the > merits in the realm of rationality? You mean back to the technical issues involved in keeping it working once it has been pulled from the base distribution? And in producing a port that will build and work across all the various 2.x and -current kernels that will be in use between the time you yank it and the time the kernel settles down? Are you volunteering to own the xten port? Isn't this a pretty minor issue to get all heated up over? -Pat To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?ML-3.3.893344470.6561.patl>