Just did a new install on a netbook--this time I went for my usual
minimal installation. It failed on bind-libs. This strikes me as ood
for two reasons--why would bind-libs be included in a minimal
installation, and why would it cause the installation to fail?
I thought I had seen, somewhere along the line, that anaconda had been
made more robust and wouldn't die on a failed package, especially when
the package is non-essential. Is that simply my imagination playing
tricks? (Googling doesn't bring me any sign of it.)
Even if it is the case, from what I see on CentOS with whatrequires
bind-libs, it's really not necessary for a minimal installation.
(Ironically, trying a complete install earlier worked perfectly.)
Would someone be able to tell me if I am imagining things about Anaconda
being so fragile? Seems that most other installers are able to recover
from something like that, so it should be doable.
The bind-libs being included in a minimal install sounds like a possible
bug, but the same argument could probably be made for anything not
essential to boot into a shell. I'm guessing that it was just a glitch
in the image downloaded on this install, or a network burp, but the
bigger concern, IMHO, is anaconda being so fragile. (Especially if it's
not supposed to be, but as mentioned above, I can't find anything to
support that, so am beginning to think I misunderstood, or perhaps
scanned over, something in the past.)
I think I'm most annoyed over my failing memory. Running the install a
second time, everything was fine.
Thanks for any feedback, (especially on whether there was or wasn't an
anaconda fix).
--
Scott Robbins
PGP keyID EB3467D6
( 1B48 077D 66F6 9DB0 FDC2 A409 FA54 EB34 67D6 )
gpg --keyserver
pgp.mit.edu --recv-keys EB3467D6
Xander: So, we Bronzin' it tonight?
Willow: Wednesdays, kinda beat...
Xander: Well, we could grind our enemies into talcum powder with
a sledgehammer, but, gosh, we did that last night.