Return to BSD News archive
Xref: sserve comp.unix.aix:48568 comp.unix.bsd:15506 comp.unix.pc-clone.32bit:7670 comp.unix.solaris:28000 comp.unix.unixware:14959 Newsgroups: comp.unix.aix,comp.unix.bsd,comp.unix.pc-clone.32bit,comp.unix.solaris,comp.unix.unixware Path: sserve!newshost.anu.edu.au!harbinger.cc.monash.edu.au!bunyip.cc.uq.oz.au!munnari.oz.au!ihnp4.ucsd.edu!agate!howland.reston.ans.net!EU.net!CERN.ch!news From: danpop@cernapo.cern.ch (Dan Pop) Subject: Re: Unix for PC Message-ID: <D0K3nI.Gty@news.cern.ch> Sender: news@news.cern.ch (USENET News System) Organization: CERN European Lab for Particle Physics References: <199411210319.TAA18133@nic.cerf.net> <CHRISB.94Nov30100302@stork.cssc-syd.tansu.com.au> <3bihms$5pf@decaxp.harvard.edu> <palowodaD04HqA.LAJ@netcom.com> <D05760.C33@news.cern.ch> <pageoneD09Mqx.A3u@netcom.com> <D0CDv6.8v@novell.co.uk> <D0E32G.3x8@news.cern.ch> <D0G3ys.DB6@novell.co.uk> <3c4ndg$oq@vixen.cso.uiuc.edu> <D0HuD1.9MB@novell.co.uk> Distribution: inet Date: Fri, 9 Dec 1994 18:35:42 GMT Lines: 54 In <D0HuD1.9MB@novell.co.uk> msohnius@novell.co.uk (Martin Sohnius) writes: >Rafal Boni (rkb55989@uxa.cso.uiuc.edu) wrote: > >: What Dan was trying to get accross, I believe, is that the 1.1.xx where >: xx is some number that was referred to previously, is just that, a >: *BETA* release. > >Thanks for clarifying this. I had honestly not grasped Dan's point. But >now that you are telling me, I can at least answer it! That's because you did read my article carefully enough and removed my first paragraph from your follow-up :-) > >: You were saying how shitty Linux QC was, and Dan gave >: an example of the exact same process in the corporate world... > >No, he did not (which may well be why the point escaped me!). No-one here >would take the attitude "I can't think of any way of testing this, so let's >put it into a beta release and see what happens." If it makes it into beta, >it has already undergone quite a bit of testing. The point of beta testing >is to expose it to all the conditions the original testers didn't think of! >It's not to make their life easier. That's precisely why I used "test" instead of "beta test". > >: If you want a Linux kernel that is stable, etc., don't run the daily >: development version [ie, the *BETA* version], run a blessed and stable >: release version. If you're hacking on the kernel, you might want that >: beta version, and hence it's out there and available.... > >Actually, I suppose where I was indeed unfair, is that those odd-numbered >releases (or whatever) are actually not beta at all: they are alpha at best; >the code that's shipped to fellow developers to have a look at. This is partly true. There is no team of alpha testers, so some changes are only tested by their author and by Linus (if he has the hardware needed, which is not the case for many device drivers). So, the "beta" versions are more like alpha than beta, this is why calling them simply "test versions" is more appropriate. The advantage is that anybody (developer or not) can test the new kernel. If he's competent enough, he can try to fix the problems detected, otherwise he can report them in comp.os.linux.development. The result is that any change is tested on a very large number of different hardware configurations, and any problems who couldn't be noticed on the developers' machines will be reported before the change is incorporated in the production kernels. Dan -- Dan Pop | The only reason God was able to make the CERN, CN Division | world in 7 days was he didn't have to remain Email: danpop@cernapo.cern.ch | compatible with the previous version. Mail: CERN - PPE, Bat. 31 R-004, CH-1211 Geneve 23, Switzerland