Anselm Lingnau wrote:
> IIRC LPI stipulates an Intel-based architecture
Not to be too anal, but I would refine that as IA-32[e] aka x86[-64].
Intel has produced a number of architectures, and more than just IA-64 too.
> for the LPIC-x exams, hence the emphasis on Intel-based boot loaders
> (LPIC-2 covers not only GRUB 2 and GRUB Legacy, but also – on the
> “awareness” level – LILO, ISOLINUX, and PXELINUX).
The problem is that we're starting to see a population move beyond
just x86-64 in the enterprise space. So while it applies now, we have
to be careful for the future.
The "Good News" is that it looks like aarch64 will be almost always
implemented with uEFI, so GRUB2-EFI or a similar solution is going to
be the tool, at least long-term.
Again ... I only provide information for "forward looking" knowledge.
I.e., I've _predicted_ a lot of changes in the past. Please
understand I'm not a swami or anything, just very "on-the-pulse" of
select industries.
E.g., the semi-conductor industry is very supply-side driven, which
results in consumer adoption. So just by reading the SIA reports, you
can stay 2-5 years ahead. ;)
> From an abstract POV it might be interesting to cover non-Intel platforms
> (especially ARM) but the risk is that this will drag in all sorts of baggage
> that, while probably worthwhile for broadening candidates' horizons, is
> irrelevant to many people's daily practice as well as difficult to teach,
> because education providers are unlikely to have the required hardware around.
Again ... just to be _clear_ ...
I _never_ advocated U-Boot. I only said if _anyone_ brings up LILO or
even ELILO, I would say knowledge of U-Boot is more commonplace.
In other words ... it was an "offering" of a "counter-argument."
Please understand that I play "devil's advocate" and _not_ "general
advocate," in addtion to providing _correct_ or more _up-to-date_ RHEL
support information. ;)
This includes asking people to _watch_ what they say about LILO/ELILO,
when it's not technical accurate.
> I'm saying that as somebody who just spent a quarter of an hour this morning
> talking about Wifi (as per LPI-201) in a classroom that didn't have any Wifi
> hardware except my own laptop and the Raspberry-Pi-with-a-USB-Wifi-dongle that
> I brought in as a Linux-based Wifi access point.
And some of us have been doing ARM, PowerPC/PowerQIC, etc... from
before the days of the Pis and other "cheap kits." ;)
The question is ... will it make someone marketable?
The only thing I'm pointing out is that the "Microserver market" is
about to explode. So it's moving _beyond_ just embedded. The Red Hat
Enterprise Linux 7 for ARM Developer Preview doesn't exist in a
vacuum.
And the SIA numbers for the next 3+ years are very eye-opening, even
beyond AMD's moves.
> If anything we will have to increase our coverage of UEFI because that is
> something that people are actually likely to run into if they're not embedded-
> system developers.
Again ... aarch64-based Microservers are _not_ embedded.
But the "Good News" is that most Microservers _will_ be utilizing GRUB2-EFI.
That was my _only_ points. My comment about U-Boot was a
"counter-argument" _against_ LILO/ELILO ... _not_ an advocation for
including U-Boot. ;)
> I recently rewrote large swathes of our LPI-101 training
> manuals to include UEFI and GPT (and bits from the {UEFI, BIOS} × {MBR, GPT}
> Cartesian product; great fun), and I'm pretty sure there will be ramifications
> on LPIC-2, too.
Unfortunately we're going to quickly find ourselves well behind in
storage coverage.
I'm having great difficulty educating people on AHCI v. NVMe, even if
I can get them to understand SCSI concepts too. NVMe is going to
radically change everything.
-- bjs
_______________________________________________
lpi-examdev mailing list
lpi-examdev@lpi.org
http://list.lpi.org/cgi-bin/mailman/listinfo/lpi-examdev
