Alex Clemente wrote:

> In more than 5 years, I met only one guy who compiled a kernel due to
> limited memory in 32bit arch.
>

Ahhh, the infamous Red Hat 4G/4G patch that never made it upstream (
kernel.org accepted)?  Or something else?

I compiled it myself just for the sake of study for the LPI test and
> knowledge myself.
>

Which is where a survey, JTA, etc... would be beneficial using the
existing, LPIC professionals and members.

Who is doing what, how often, etc...?

I.e., while it's cool that LPI has such broad objectives so people learn so
many things, those objectives should still be ...

A)  Relevant to capture 1 (68%/32%), 2 (95%/5%) or even 3 (99.7%/0.3%)
sigma of sysadmins ...

B)  Weighted as appropriate against commonality, difficulty, etc...

If not even 0.3% of sysadmins are not even doing any kernel _module_
building, then it might be worth seriously dropping it.  Maybe kernel
instrumentation, kprobes, performance counters, etc... would be better?
Lots of 'payloads' are going that route now, to avoid modules.

Or ...

If at least 5% are building kernel _modules_ at some point in their career
(even if infrequent), but less than 0.3% are prepping any kernel tree, then
maybe the objective should be re-written and updated to reflect common
kernel _module_ building, and not so much kernel building.

One of the not-so-few, but proud'n ignorant, opinionated Americans ... ;)

- bjs
_______________________________________________
lpi-examdev mailing list
[email protected]
https://list.lpi.org/mailman/listinfo/lpi-examdev

Reply via email to