Meta-discussion about talking about Linux (was Re: [NBLUG/talk]
Boot Loader problem)
mark at oswizards.com
Mon May 3 11:10:17 PDT 2004
-----BEGIN PGP SIGNED MESSAGE-----
The documentation is out there. I didn't code grub or create the installation
program for Red Hat or Fedora Core.
General Boot Methods
I will post one possible solution under a separate thread.
On Monday 03 May 2004 10:22 am, sms at sonic.net wrote:
> I understand, I think. I just disagree: "...the user should..." just
> makes me very nervous as a sort of "policy." It hearkens back, IMHO,
> to ideas like that ol' "UNIX priesthood" chesnut, or the oldschool
> programmer-ism, "Comment my code? Why would I comment my code? It was
> hard to write, dammit, and it should be hard to read!"
> Instead, I think "the user should" have an expectation that the OS serve
> their needs, that utilities serve their needs; and one of those needs is
> for reasonably-transparent functionality. As the question was asked here,
> it's self-evident that "reasonable transparence" wasn't the experience of
> this user.
Agreed. Which is why I decided to try to invoke more discussion on the part
of the user and the group. We all know that the questioner has posted
homework questions to this list in the past. Not a bad thing, but learning
must take place in the college course arena, there are course mailing lists
that the user can ask questions on.... I am a member of that particular
course mailing list and the questions were asked there.
> I guess it depends. If this is a "sandbox" system for a hobbyist, such
> an approach seems fine. OTOH, if it's a business system, or an only-box
> at home, an "object lesson" may be less appreciated. Even an "impatient"
> end-user may not *care* to muck about debugging stuff, and just want a
> *working* environment... and so long as self-debugging is normative for
> linux, it remains a hobbiest OS, not professional-caliber.
An impatient end user in most instances probably would not install 3 operating
systems onto 2 hard disk drives. ; )
> And before you say "but multi-OS booting is a complex issue that shouldn't
> be done by unqualified people..." I'll suggest that in the Real World with
> MS-dominance, multi-booting should be the default expectation, and any
> "default expectation" should be aimed at the average (unqualified) user.
It is the default expectation of the Red Hat Installer to recognize other
operating systems and configure them accordingly in the boot loader. The
install process for a multiboot (triple+) system is NOT foolproof when it
comes to installing the boot loader... you have to know a little bit about
boot loaders to configure the machine correctly.
We have no idea how this user configured his boot loader during install.... I
have a pretty good idea based on the configuration files for the two versions
of Red Hat Linux he installed and the order in which he installed the
> That rant aside... I agree that thinking is better than not-thinking, and
> people are always better off understanding what's going on with their
> computers, particularly as the complexity of their boxes increases.
> What do you think of the idea that (in a situation like this one) a useful
> answer-format might be something like:
> <ANSWER: If you _need_ your answer now, and/or don't care to learn
Not a bad thing. Don't get me or my actions wrong, I post many
answers/solutions right off the top on this list and never bat an eye. In
this instance I felt something in my gut that made me hold onto the solution
for a bit.
Other lists that I am associated with do not post answers to the list. The
answers or possible solutions are sent directly to the questioner. It is
then the questioner's responsibility comprehend, digest and compile the
merits of each solution and post a summary back to the list of the best
possible solutions to the question. Sometimes they do.... sometimes they
Mark Street, RHCE
Key fingerprint = 3949 39E4 6317 7C3C 023E 2B1F 6FB3 06E7 D109 56C0
GPG key http://www.oswizards.com/pubkey.asc
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
-----END PGP SIGNATURE-----
More information about the talk