On Fri, Mar 29, 2002 at 06:54:32PM -0800, David Talkington wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Dan Stromberg wrote: > > >Thank you for recognizing that admins have more useful things to do with > >their time than a steady stream of redhat upgrades. > > As Ed points out, that really shouldn't be an issue for you. There's no > compelling reason for you, the admin, to chase release numbers. Red > Hat's decision sounds like it has more to do with getting software > partners on board ... developers of enterprise software are the ones > with probably the biggest axe to grind there, since they have a lot of > work to do every time the OS is upgraded:
I'm not sure you read my message very carefully. Not only do we have tons of software to recompile each time there's a new release (not that different from the ISV's really - actually, it's probably more of a problem for us than for them), our users just hate starting out with something older than the latest. So we can try to stick to 7.1 until 8, but our users will often complain if we don't install 7.2 until the minute 8 is released. So it would seem you too can probably infer that skipping every other release doesn't make for the best PR with our clients. Also, most likely 7.1 won't be supported with new patches as long as 6.2 will, but if we always wait for the x.2 release, our users will REALLY howl. Not only that, but there's no guarantee I've heard from redhat, that x.2 will be the last of the x.y's, so if we wait for x.s and support there, there could be a x.3 later that makes x.2 get orphaned sooner as far as patch availability goes. IE, many admins (and even endusers) do have a problem due to the too-frequent releases. It's far from being just an ISV problem. I know doing a steady stream of upgrades may seem like a job security thing to some admins (why complain? At least you're working!), but to others, it just steals a disproportionate fraction of our workload away from our other unixes. IE, we'll be plenty busy even if Redhat does The Right Thing on this issue. I'm in no way worried about not having enough work to do. On Sun, we haven't had to do a major round of recompiles since Solaris 2.x was released - and remember the fierce shreaking over that? Redhat makes us do a Periodic recompile. It's not a good thing. In even starker contrast, we haven't had to do a major round of recompiles on IRIX or Tru64 since I started using them about 9 years ago. These *ix's have much better binary compatibility. And that matters to this workgroup's bottom line in a big way. -- Dan Stromberg UCI/NACS/DCS
msg75304/pgp00000.pgp
Description: PGP signature
