On Wed, 20 Apr 2005 11:50, Steve Holdoway wrote: > On Wed, April 20, 2005 11:23 am, Christopher Sawtell said: > [snip] > > > That's why the distributions offer testing, unstable, and stable levels. > > The public do the testing at the testing and unstable levels. Fedora is, > > and > > always has been, at the testing to unstable level. > > [snip] > Interesting that you use debian as your yardstick... which as you probably > know is my weapon of choice. There is also an experimental distribution, > which is *still* better tested than this. > > But where do you stick Ubuntu, Unknown quantity at the moment, but probably going to develop into something pretty good. Incomprehensible aesthetics.
> Mandrake, Poor to maddening > SuSE/Novell, Best of the commercial distros _at the moment_. What Novell will do with or to it in the future is an unknown quantity. > Gentoo Gentoo stable is, imho, the distro of choice. The documentation is second to none, the irc channels are alive and kicking 24/7, the forums are a real goldmine of good information. > and others on your scale? Debian is the 'other' distribution of choice. The problem with it seems to be that it has ossified. It seems to be more of a 'Social Statement' than a vibrant computer o/s. I may be wrong because I haven't used it for several years, but that is my impression from what I read. > Unlike most of the readers of this list, this is my > livelihood. I *do* run RHEL ( well White Box! ) because Oracle demand that > I do for support. However, Oracle runs just as well on Fedora Core 3, and > I can use a 2.6 kernel. Whopee - but thanks a bunch Larry. You've > subverted the whole cause of Linux by demanding that you pay for it. Or, > given your viewpoint, has he? Larry is just freeloading on Linux. He cannot subvert the GPL. Don't forget that he has competent competition from both Free and commercial packages. Like all monopolists, his time for downfall will come. That's a certainty. > I think that my past is colouring my impression of what an acceptable > level of testing is - I started life developing software for medical > physics where the penalty for getting it wrong could be rather serious - > but maybe your impressions are skewed too far the other way. You still > download FC from fedora.redhat.com, so their name and reputation is still > linked to the product. I have only used RedHat for a very short time - long before Fedora was even thought of. But I formed an opinion of it very quickly. Your experiences, and those of others, has not altered it one iota. > Maybe we should start a campaign to improve the quality of testing? If > everyone accepts this level of software as the norm, then we're never > going to get anywhere. I run a F/LOSS project, and even taking the initial > software download from the developers and getting it to a state that it > compiles and installs cleanly when extracted from the sourceforge cvs > server took over a weeks work. Should I have just stuck it up there, and > ignored the moans of the interested parties who'll never come back and > never support my project? > > I think the implications of sticking any old cr*p onto sourceforge and > calling it open software are very wide ranging, and need to be carefully > considered. The implications of (in effect) doing this and ten putting > your brand name to it are far, far worse! Yes, it stuffs the brand. Just because something is popular, that does not mean that it is the best of breed, or even technically competent. In fact the reverse is usually true. -- C. S.
