Quoting Brad Campbell (lists2009@???):

> Details are sketchy now as I made this change in 2005 after 9 good
> years with LILO. I did try very hard to make it work, and it may
> have been an issue with the BIOS ultimately. I actually had a hard
> copy of that particular howto next to my console as I was constantly
> working with LILO and RAID trying to get something that was reliable
> across the plethora of nasty hardware I had to support, so if there
> was a tip in there that might have applied I certainly would have
> tried it more than once. Thankfully I've managed to repress most of
> those memories.

Yes, I know the feeling. (I'd help if I could, but sadly neither of us
even has a test platform at the moment.)

> The fact remains once bitten, twice shy and it was *much* easier to
> get Grub to work reliably than lilo in this particular instance.

Well, whatever works for you is good. When I get around to testing
extlinux on the planned pair of RAID1ed SSDs for my new server, I'll
make a point of simulating a device failure and ensure everything works
(and remember to report back here) -- but it's unlikely that I'll get
around to also testing GRUB. Here in the latter days of 2016 where I do
basic Web searching and find what _purports_ to be a standard solution
to the exact problem you describe, my hunch is that it works. Perhaps
it's just better documented now than it was then. Or, alternatively,
maybe it still wouldn't work on your hardware that you used back then.
Impossible to say.

Back in 2002 when I wrote that small Zen of LILO piece for _Linux
Gazette_, I had a very strong feeling that GRUB had widely displaced
lilo for overwhelmingly inadequte an unconvincing reasons -- and that
remains my impression.

> I've used Grub now on all x86 based machines since 2005 and I've never
> bumped up against an issue that has taken more than 5 minutes to
> resolve [...]

I'm happy that it meets your needs. From my _own_ perspective, what I
see is needlessly complex software that practically begs to be replaced
by something more sparse and devoid of (say) a gratuitously different
device-addressing scheme and the need & ability to read filesystem
semantics. Accordingly, I look forward to finally getting rid of it,
reversing the lazy system administration error I carelessly permitted
(through inaction) the Debian developers to apply to my systems.

I think the sentiment is widely felt around here that allowing Debian
package maintainers to determine one's system architecture through one's
own inaction has had a bad history, even though various of us might
disagree on particular parts of that.

> I'm not rejecting your experience, I'm just saying *I* found a case
> would reliably break LILO, and after learning how to actually
> configure Grub I've never bumped up against anything similar. That's
> just me however.

Makes sense!

And (your) 'horses for courses' is my second favourite Oz expression,
second only to (of course) 'No worries'.

In 2010, I was walking across one of the pedestrian bridges across the
Yarra River near the Melbourne Convention Centre, and a pair of
Australians walking the other way asked me if I knew where a particular
hotel was. I said sure, and offered to walk them over there, as it
wasn't far out of my way. They politely objected that it was too much
trouble, and I said 'No worries!' Their expression was priceless, a
sort of 'Hey, I'm not sure you're allowed to say that with a California
accent.' ;->