unity-design team mailing list archive

[Chad M/ Germann <cgermann@xxxxxxxxx>]

Oh lovely a line by liner



> Android has a Linux Kernel, WebOS runs on Linux, iOS is Unix-based, so
> yeah.
> To be Unix-like in my opinion has more to do with lower level concepts
> and philosophy, it is less GUI related.
> 
> 
> If a button is difficult to hit because it is too small a target area
> for a pointer, that has nothing to do with Unix-like or not, it's
> Fitt's law or simply HCI. And it is irrelevant, whether we are talking
> about a touch-device, a mouse operated device or what ever other
> pointer-based HID have you.

I wont go into Fitt's law as that is so misunderstood and misused on the
Internet it is almost comedy.

The problem is and you don't see it is your trying to cram a touch
interface on a Desktop OS. 

Ubuntu will never see a Tablet or phone becuse it has too much in the
way

for example 

Power management Ubuntu is far from the best of the best distro for
power usage. (my laptop gets 20 more minutes under fedora than it does
under Ubuntu 11.10)  

Unity is a plug in for compiz. to fully understand the problem I invite
you to do a little experiment. Play a video or game full screen in Unity
than do the same in Gnome Shell than KDE, Note that the Frames per
second slightly faster Mutter and even faster under Kwin. (and the gap
between Mutter and Kwin is narrowing as Mutter matures) it this is
notable on heavy Desktop hardware how is it going to be under mobile
hardware. Hardware vendors are a panicky lot and are going to notice
this and think "nobody is going to want a Tablet OS that cant play
smooth full screen video or play Angry Birds with."

Now to break and go on a tangent for a moment I am going to state this
on the record. And i will stake  my reputation as a Computer Scientist
on the following statement. the OS that continues to build a system
designed around a rich desktop client model will be flooded with users
within five years. 

You may be thinking Why do I say this? Simple think back 10 years or so
and the big Industry fad at the time was thin clients. At the time, they
seemed like a good idea move applications and the heavy computational
lifting to the server room and pipe it to the end users desktop to cheep
Dumb terminals  and save a bundle in IT costs from hardware pricing and
support.

Quickly problems arose as corporate intranets (I find it funny that term
is so old that is is not in the spell checker)  were breached. and
Productivity slowed downs as applications moved as the crawl that
100baseT and Intranet connections could serve them. it was a failure and
not long afterwards shiny Full blooded desktops replaced the thin
clients. 

Now we have this "cloud" (are they nimbostratus or stratocunulus?) fad
and Everyone wants to build for it with there trendy "Post PC"
Interfaces to support it. Gnome, Unity and to a marginal extent KDE want
to get on the cool kids train

Lets look forward to 5 years from now (and I am being optimistic) when
End users figure out that "Cloud Computing" is the return of the thin
client only worse as there stuff is being stored further away and
attached to a even more public and insecure network that is slowed down
by massive amounts of traffic and the effects of "activists" and
disgruntled users. A mass exodus will occure back to rich desktop
clients and at that time the system that can conjure up the most
compelling Traditional WIMP Desktop interface will win.


 
Now ato addres your "Rebukes"

>         On Tue, Nov 29, 2011 at 11:50, Chad M/ Germann
>         <cgermann@xxxxxxxxx> wrote:
>         > On Tue, 2011-11-29 at 12:52
>         +0100, frederik.nnaji@xxxxxxxxx wrote:
>         >> how do you define Unix-like in 2011?
>         >
>         > In a nutshell Keep it Simple, Stupid.
>         >
>         > to expand.
>         >
>         >  1. Small is beautiful.
>         
> 
> 
> beautiful is mostly also useful, but not always usable.
> 
Small has a History of being more userfrendly than large and convaluted.

>         >     2. Make each program do one thing well.
>         
> 
> 
> true, ..and name it after that thing, too, so that the user can
> identify the purpose rather than the manufacturer.
>  
>         [...]
>         >     9. Make every program a filter
> 
> 
> hot, interesting! i think this might be the idea behind lenses and
> prisms.

lenses and prisms are not Filters as you can't pipe the output into
other programs to further analyze the output grep is a filter lenses are
not



>         >    15. Programmer time is expensive; conserve it in
>         preference to
>         >        machine time
>         
> 
> 
> what about user time? How expensive is that?

Non Sequitur question considering that the Quote was about Machine time
but, to corporate world a user's time is cheep compared to an engineer's
time as the engineer pulls a bigger paycheck and also Designs the
product.

But yes the Time of the creative person is a more valuable commodity. as
they can spend that time creating.