Bug#455769: same problem on wheezy + Thinkpad X220T

[Wouter Verhelst]

On 28-03-13 14:04, Daniel Pocock wrote:
> On 28/03/13 12:32, Wouter Verhelst wrote:
>> On 28-03-13 11:47, Daniel Pocock wrote:
>>> On 28/03/13 11:06, Julien Cristau wrote:
>>>> Control: severity -1 important
>>>>
>>>> I am raising this bug to critical, as it meets the definition "makes
>>>> unrelated software on the system (or the whole system) break"
>>>> No, it does not.  hw will shut itself off before getting damaged.
>>> Would you provide a guarantee to all users of wheezy that you will pay
>>> for their laptop repair if this issue causes damage?
>> Oh please.
>>
>> I've never seen a laptop that will not cause a poweroff in hardware if
>> the operating system isn't doing its job.
> I agree laptops usually power down by themselves (my old X60 did that
> many times even with lid open)  - it is just not something that we
> should fall back on to downgrade a bug severity
No, that's not true.

Your claim that the hardware could break due to the software not doing
what it should be doing is wrong. If the operating system does not cause
a suspend, then the hardware should fall back to powering the system off
when overheating is imminent. If the hardware doesn't do that, that's a
*hardware* bug, not a software bug. I've also never seen that happen,
but that could of course be me.

Yes, the operating system should do a proper suspend when asked. If that
doesn't happen, that's a bug in the operating system. But it's not a
critical one (by Debian's definition of that word); and the fact that
there *might* also be a hardware bug that would then cause hardware
damage does not change that.
>> This is also not a critical issue. If gnome-power-manager were to go
>> berserk and do the equivalent of "rm -rf /", or "kill -9" on all
>> processes on the system, or something similar, that would be a critical
>> bug in the "makes the whole system break" sense. As it is,
>> gnome-power-manager just *doesn't* do something it's supposed to do.
>> It's not even close to being critical in that sense.
> The severity guidelines don't say whether the definition of `system'
> includes hardware. 
> A broad definition of system could include the backpack carrying the
> laptop, and the bottle of wine that heats up because it is in there with
> the laptop.
Can you stop the lawyering, please? They're severity *guidelines*, not laws.

I think it's pretty obvious that "system" means "Debian", here, not
"Debian, the backpack in which it's running, the owner of that backpack,
the car in which the backpack is riding along, the exhaust volume of
that car, and the resulting climate change."

Or were you saying that a failure to reduce the climate change impact
should be a release critical bug?
> Some organisations have multiple metrics for each bug, e.g. an impact
> metric to measure how many users suffer.  Non-technical users are likely
> to suffer and have a bad impression of Debian when their laptop heats up
> in their backpack.
Yes, so it's a bug. I don't think anyone has contested that. It may even
be release-critical, though I'm not convinced of that.

However, it is certainly not a *critical* bug. That's what I'm saying here.
>> (on a more personal note, why oh why would you ever want the system to
>> suspend when you close the lid? That's what a suspend button is for. If
>> my laptop is compiling something, I do *not* want it to suspend when I
>> close the lid, thankyouverymuch. Oh well)
>>
> a) many users expect that - I'm thinking about it from the perspective
> of people who don't compile things and aren't actively involved in this list
When I say "on a more personal note", that should be interpreted as "I
realize that not everyone shares this opinion, but..."

Anyway, EOT for me.

-- 
Copyshops should do vouchers. So that next time some bureaucracy
requires you to mail a form in triplicate, you can mail it just once,
add a voucher, and save on postage.