The Linus view of GNOME 3.2

Tim ignored_mailbox at yahoo.com.au
Tue Dec 6 00:48:16 UTC 2011


On Mon, 2011-12-05 at 09:20 -0800, Robert M. Witkop wrote:
> I upgraded my hardware about the time that F15 came out, so I decided
> that I would install f15 instead of f14. When I found that gnome did
> not run out of the box, I had to make a decision, to I want to go
> through a learning curve, or do I want to develop systems. 
>  
> If I wanted windoz, I would run windoz. 
> 
Years ago, I also used computers long before we'd even heard of Windows.
My own first real personal computer was the Amiga, and that was chosen
after being thoroughly put off by the other personal computers that I'd
had a play with (that others owned).  I liked it for what it was, how it
worked (much better than the alternatives, at the time), and what could
be done with it.  I could look after it quite easily, and there were few
secrets.

A few years later, I faced the no-choice of having to get a PC, to be
able to do stuff in a PC-only world (or the majority of it being that
way, that it might as well be PC-only).  And I got sick of it real
quick.  It was, and still is, a suicidal OS, plagued with problems, and
probably always will be.  Another thing that really pissed me off, about
it, was how much of it was secret.  I got really really sick of the "see
your admin" error messages.  I was the admin, there was no one to ask.
The built-in help was crap, the manual gave no information about the
issues.  Help on the WWW was far from satisfactory, and even when you
did find an answer that was correct, the solutions were utterly
ludicrous from a computing point of view (the design/philosophy of
Windows is just plain nuts).  And really, the only way to know what
you're doing (if you wanted to make a job out of IT), rather than be
some klutz who might stumble on a few things, was to go on some
expensive training course, again and again.  And, when you look into
what some of these course were about, it was galling.  Networking
training reduced down from knowing about administrating TCP/IP, to
merely learning the crapped out way that Windows does it.  Forget about
learning about DNS, or TCP/IP traffic, or any of that, just learn which
boxes to fill in for this version of Windows.  Then go on another
expensive course when they release a new version.  One of these courses
was little more than what you see in the setting up your network, in the
installation guide for Fedora.  It was that /sparse/ in info.  Fill in
the blanks, don't actually learn about what you're doing.

Then, at long last, Linux came to a point where it was usable as a
personal computer, in the current world.  I dropped Windows in a flash.
Never regretted it.  Enjoyed the lack of secrecy (documentation, things
working in a sensible manner that you could work out the local
implementation, source code if you wanted to peruse it, no hidden files
on your drive from package installations that you can't tell what will
put where), and the return to a sane filing system (non-suicidal filing
system, files stored in sensible places, separation of system's from
application's from user's files).  But over the last few years I keep
seeing one Linux-thing after another going Windows-like (dumb ideas,
cloning Windows, no documentation, go see your admin).  With developers
either copying the worlds worst examples of computing (i.e. Windows),
and not knowing (because they've never used other systems), or not
acknowledging that the Windows way is the horrible way.  Or it's Windows
developers migrating to a different OS (Linux) and just carrying on
doing the same crap, instead of learning how to do thing in a better
way.  I suspect the latter, since it seems to be that it used to be Unix
users going off to Linux, since Unix was expensive but they could afford
to personally use Linux as a similar alternative, so we got a lot of the
Unix mindset (designed by computer boffins).  But now it seems to be
Windows users doing the same thing, migrating from an expensive or
pirated product, to something they can afford to play with, and they're
building stuff (but with all the lack of experience of a teenage hacker,
someone who just cobbles together something that seems to work, but
doesn't integrate well into an established system, and then they mess up
rest of the system to suit their hack, instead of fitting in).


> I know "opinions are like a part of our anatomy in that everyone has
> on", and I usually keep my opinions to myself, but GNOME3 is so
> different from 2 that it should have been forked as a new product, not
> put out as a revision.

I tend to agree.  If some of the Gnome developers (or the same applies
to KDE), wanted to go out in a radically different direction, they
should have started their own new project, and left the current one to
those interested in it.

If the old Gnome, or KDE, or whatever, continued or died a natural
death, that would be it.  The current debacle has been a forced
termination.  It virtually precludes some from continuing on with
working on the older desktop, because it's been deliberately poisoned.

And it's not just the change in direction, it's the huge increase in
computing power that's a big problem.  Some years ago, Compiz came out,
with fancy flashy effects for your desktop (pretty, but unessential).
That required a 3D accelerated card, which wasn't appreciated very much.
Those of us with the hardware had a play, but would notice that Compiz,
by itself, was using almost all the available resources.  The computer
was getting sluggish, the card was getting very hot.  Never mind wanting
to do other things with your computer, than merely run the desktop.  I
want to browse three page, at the same time, while writing a document,
while listening to some sound file...  I could actually do that
pre-bloated desktop era.  And the new Gnome, requiring lots grunt, has
gone down the same path.  *And* by all accounts, the fallback option
still has bloated needs.

Buying new and expensive hardware every few years, for artificially
necessary reasons, is the Windows mindset.  And it doesn't work well for
Linux, because you also need Linux drivers for the hardware, and you
find that your hardware purchase choices are narrowed down to about
three items.

No, I'm not being a stick in the mud, I'm being pragmatic.  My design
background is more electronics than computing, but the same principles
apply.  Designing something with extreme, and difficult to fulfil,
requirements, is not sensible.

-- 
[tim at localhost ~]$ uname -r
2.6.27.25-78.2.56.fc9.i686

Don't send private replies to my address, the mailbox is ignored.  I
read messages from the public lists.





More information about the users mailing list