[PATCH] Turn Powerdevil suspend notification into a dialog

[Aaron J. Seigo]

On September 23, 2009, David Nadlinger wrote:
> Furthermore, the battery level/suspension notification is not exactly
> something _unexpected_ – the user was notified at least once about the
> remaining battery power reaching a low level.

not to mention you have the battery icon down there. i actually asked various 
non-technical users about this early this year to get a feel for how the 
battery notification system works for people (or doesn't). the  #1 thing 
people said was they look at the battery icon after they've been uplugged for 
a while and see what it says.

notification of power issues, it seems, is useful in two primary cases:

* unexpected power shortage (e.g. you thought you were plugged in, but you 
aren't, something that can happen if you don't plug it in properly for 
instance; or your battery's life is diminishing through regular wear and tear 
over time and you haven't noticed yet)

* someone gets very distracted and loses track of time (very possible)

otherwise, the battery indicator itself is enough for nearly everyone it 
seems. so these notifications are needed and are useful but do not need to be 
IN YOUR FACE. they are in the general case relatively unuseful. they are 
emergency measures.

btw, the most common response to "you have no battery left" seems to be to 
plug your computer in if there is power nearby or to shut down the system at 
that point. so if power devil doesn't stop it's "i'm going to suspend" when 
the power is plugged in, that would be quite unexpected for the user and a 
common failure scenario in that case. which is to say, that in the small % of 
cases where this notification is needed one of the two most common operations 
will prevent any interaction from being needed if power devil does it "right". 
that means giving enough time to get a power cord and plug it in and auto-
aborting when power returns.

increasing the time out from 10s to 30s or maybe even 60s should give enough 
time to plug the computer in; and hopefully power devil already auto-aborts 
the suspend action in that case. (Dario, can you confirm that?)

(note that i'm very well aware that my observations as above are not 
scientific studies. i'd love to have such resources at my disposal, but i 
don't. :) the observations are, however, designed to gather relevant data to 
compare with what we are actually doing and often result in changes in how we 
implement things. 

in my experience, this is more than most people who work on these issues can 
say. i experience some exasperation when dealing with these topics because 
people too often weigh in before researching published materials or doing any 
field work of their own.)

> I am, however, strongly in favour of extending the timeout or making
> the notification more »noticeable« (whatever this turns out to mean)
> if this helps some users –

agreed.

> it is just that it has never been a problem for me.

i honestly doubt it's a problem for nearly anyone. i keep asking the question, 
nobody answers except with "well, on an old version of KDE" or "with broken 
kernel support". 

> As for the (K)Ubuntu issue: I must admit that I have never dealt with
> the notification system from a developer point of view, and neither do
> I know notify-osd, but I really don't understand the fuss about this:

when you look at the patches to dozens of applications that are necessary to 
make this all work properly and how it makes Ubuntu systems very different in 
behavior from other F/OSS offerings resulting in a fracturing of the user 
experience in an already small market (the phrase that springs to my mind is 
"dividing one's own house so that others may more easily do the conquering"), 
the reason for the fuss becomes more apparent.

i'd be honestly happy to ignore this whole situation and see, as a curious 
onlooker, how it pans out for Ubuntu users over the next couple years. but i 
keep getting sent patches related to it that i'm not interested in. and yet 
they keep coming. this patch and the resulting thread is a nice example.

i'm happy to see Ayatana experiments if only because trying new ideas is 
something i think is a generally good thing, but i'm not pleased to spend my 
time on things i don't even agree with due to it spilling repeatedly and 
unwanted over the fence and into my yard.

> If there is a (de facto-?)standard which covers actions in
> notifications, frankly, why should we care about a downstream
> application which does not adhere to this? If not, we probably should
> really try to get others to support this useful feature, rather than

the F/OSS implementations *do* support actions in notifications. that's 
because the upstreams all agree that actions are useful and should be in the 
visible user interface. this is a point that Canonical departs from their 
upstreams on.  this is something the freedom in our software allows for. :)

-- 
Aaron J. Seigo
humru othro a kohnu se
GPG Fingerprint: 8B8B 2209 0C6F 7C47 B1EA  EE75 D6B7 2EB1 A7F1 DB43

KDE core developer sponsored by Qt Development Frameworks
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: This is a digitally signed message part.
URL: <http://mail.kde.org/pipermail/kde-core-devel/attachments/20090923/71ed1a29/attachment.sig>