On 03/12/2013 02:24 PM, Brian Wheeler wrote:
On 03/12/2013 02:03 PM, Chris Murphy wrote:
> On Mar 12, 2013, at 10:35 AM, Reindl Harald <h.reindl(a)thelounge.net>
> wrote:
>
>>
>> Am 12.03.2013 17:32, schrieb Chris Murphy:
>>> On Mar 12, 2013, at 6:02 AM, Jiří Eischmann <eischmann(a)redhat.com>
>>> wrote:
>>>
>>>> New kernels bring a lot of
>>>> regressions and we don't have enough test coverage to avoid them.
The
>>>> general solution to those problems is to go back to the last working
>>>> kernel version. But by making it less obvious we make these frequent
>>>> problems more difficult to solve.
>>> This is completely specious. A user who considers falling back to an
>>> older kernel as a troubleshooting step also knows how this selection
>>> is made and where to go look for it
>> THIS IS WRONG
> Oh really?
Yes, it is wrong. We're not talking about just new users here. If
you're going to hide how to select a different kernel, how am I, an
experienced sysadmin supposed to figure it out when things go south?
F18 screwed my computer royally with regards to sleep & restore and I
had to boot older kernels to get the machine stable. As it stands,
there were a list of kernels I chose the upper most one which didn't
have problems...under what people are proposing I'd have to google it
on some other machine or just mash the keyboard and hope I find
something that gives me some options
I don't know why people are so enamored by making it difficult to
troubleshoot problems.
I know it is a simplification, but to me, the two sides of this argument
are:
* remove the hood of the car, and keep it off in case something goes
wrong, or to entice new drivers to look in there and guess what is going on.
* keep the hood of the car on, and if something goes wrong, pop it. If
the driver wants to tweak, or have a look around let them pull the lever
and pop the hood.
--ryanlerch