[colug-432] CentOS

Sun Dec 18 10:21:01 EST 2011

Ubuntu is starting to piss off a lot of their users, as well.  The most
recent version, 11.10, forces you to use Unity, even if you've been
using Gnome all this time.  If you want to use Gnome again, you have to
install it manually.  I'm not sure what the designers intended for
Unity, but right now it's a featureless pile of crap and not ready for
prime time.

Gnome 3 has its own problems, too.  For example, nautilus has dropped
the --browser option but still has it in the man page.  Plus, nautilus
refuses to start if you use that switch.  The user has to go through and
manually edit all nautilus*.desktop files and remove the switch so that
you can launch it from the menus.  Also, some configuration file
somewhere has randomly assigned the down arrow key and the end key to
the compose key.  I haven't figured out which one does that yet.

I'm not sure what the distribution makers are thinking but apparently it
isn't confined to just one.  You'd think they would want to make Linux
more user friendly, not less, if they want to make inroads onto the desktop.

On 12/18/2011 09:58 AM, Neal Dias wrote:
> In the interest of full disclosure, former Red Hat employee...
> 
> Red Hat has clearly lost it's way in the desktop space. Next time you
> have a Red Hat sales guy or consultant on premises, or go to Summit,
> just notice how many Red Hat folk actually run RHEL as their desktop.
> Then notice the number of Mac's in the room. In fact, while at Red Hat
> I wrote an email to memo-list (Red Hat's internal "say anything"
> mailing list) saying that if you're in front of a customer, you should
> be running RHEL on your desktop, not Fedora or anything else. The
> response was enlightening, I think if I had had a cubicle in Raleigh
> someone would have set it on fire. How can Red Hat expect to be
> relevant on the desktop and yet not care enough about the product to
> actually use it themselves?
> 
> Gnome 3 pushed me over the edge when it came to Fedora. The arrogance
> of the Gnome developers to decide they know better than their user
> base how the desktop should work is beyond annoying, and that
> arrogance extends into Red Hat. In fact, one of my concerns is that
> Red Hat has too much influence on the community as a whole from a
> development perspective. As a small example, while at Red Hat I found
> that a feature had been removed from Gnome. I found a bug was already
> open on it, with multiple user comments requesting the feature be
> returned. Red Hat's arrogant response was that they had determined
> that the feature wasn't needed or used very often, and so they worked
> with the *upstream developer* to have the feature removed, and told
> us, the users, that we would have to work with upstream if we wanted
> the feature back, and closed the bug as "won't fix". So Red Hat will
> make knee jerk UI decisions and then influence upstream to accept
> those decisions leaving the user with no recourse.
> 
> What Apple has done with BSD should be call to arms for the Linux
> desktop, instead, we concede the ground and run to the cloud. Open
> Source is wonderful, and I don't want to detract from the commitment
> Red Hat has to Open Source in general, it is to be commended, but
> there is something to be said for strict product management that isn't
> pulled every which way at the whim of the community.
> 
> My $0.02.
> 
> -nd
> 
> On Sat, Dec 17, 2011 at 10:18 PM, R P Herrold <herrold at owlriver.com> wrote:
>> On Sat, 17 Dec 2011, Steve VanSlyck wrote:
>>
>>> I thought I would ask if anyone has any opinions about
>>> CentOS desktop. It used to be pretty spare and utilitarian
>>> and I want something friendlier
>>
>> (rant alert)
>>
>> And I'd like a good fifty cent cigar and shot of single malt,
>> too, while we are at unlikely wishes
>>
>> In my opinion, Red Hat is letting the UI Gnomes in Fedora
>> drive them right into GUI irrelevance to the server room; As
>> the CTO at Red Hat long ago discussed Red Hat's disinterest in
>> the Desktop (X-top), one has to ask:
>>
>>        What is left?
>>
>> CentOS is 'bound at the hip' to Red Hat's decisions, and will
>> be pulled off that cliff as well, in due course [1].  EPEL, at
>> best, can 'temper the blow' of unusability, and the insanity
>> of 'latest is greatest' disease
>>
>> I speak here of XFCE as a desktop, which was previously
>> packaged in EPEL by Kevin Fenzi, [KRUD -- Kevin's Rad Uber
>> Distribution, pre-dating Fedora, and discontinued in light of
>> it] formerly of tummy.com in Colorado, but who recently hired
>> on at Red Hat
>>
>> EPEL is functionally enslaved and unwilling to do anything
>> that adds a package which updates or conflicts with RHEL.
>> Their project, their choices
>>
>> Nothing I care for much -- Gnome is a bloated waddling pig,
>> KDE in RHEL / CentOS is starved for development love and
>> resources and partially stabilized at best ... Rex Dieter
>> outside and three or four others inside the 'Hat, are about it
>> for the vast majority of RH related KDE commits, and because
>> of the way release cycles break, KDE always lags.  >shrug<
>>
>> It might kill Fedora in their conflicted desires -- to have
>> more live packages than Debian, and to out-kewl Ubuntu -- to
>> abandon their chase of the latest and greatest bike-shed.  It
>> surely would 'better the breed' to run one 6 month release
>> cycle wholly devoted to writing the missing documentation, and
>> handling bugs in a fashion other than auto-closing old and
>> untouched ones. But that's not going to happen, because the
>> Fedora lists resemble so much as a third-grade cafeteria food
>> fight, conducted with all the social maturity of several rival
>> competing cliques of high-school sophomore, teen-age girls
>>
>> As to running a wholly owned subsidiary of crowd-source
>> drones, Red Hat's management of Fedora has done a wonderful
>> job sucking the air out of much of serious development talent
>> from an architected design plan of a future for *nix
>>
>> Perhaps fortunately, the era of the desktop as the principal
>> venue for end users to do compute is winding down in favor of
>> pads and mobile devices.  Tool builder cultures use TUI *nix;
>> the next generation moved to GUI goop like Eclipse and other
>> 'visual' environments that just are not and cannot be as fast
>> in producing crafted code;  a mass of code is pooped out, to
>> be sure, but nothing a machine was not largely responsible for
>> producing the boilerplate body of ... ick
>>
>>
>> All I really want in an X-top is lots of console windows, xpdf
>> and xournal, a tab based browser, and miscellaneous short
>> lived helper apps ... that I can navigate between without
>> touching the mouse most of the time.  OS/X fails here; Windows
>> too.  In taking a count, something over a third of the
>> packages on my principal development box were locally
>> produced, because the offerings of others stunk up the place
>> so much
>>
>> Apple did a great job productizing BSD and bringing
>> graphically pleasing integration to that herd of GUI cats --
>> take a count the next time you are at a conference, and it is
>> pretty clear that Linux is not where the mind-share is.  One
>> sees corporate provided lappies with MSFT's latest, or for
>> contractors who provide their own tools, Apple kit with a VM
>> based Windows box inside to talk to the Exchange Server
>>
>>
>> But the future remains in the datacenter -- the cloud, as we
>> can call it in the trend of the moment with an MVC-ish
>> approach -- presentation and input logic close to the end
>> user; business logic in the middle, and data-farms behind, and
>> at the bottom, stuff is happening either in the command line,
>> or in structured communication in the panels of a browser-like
>> front end sending Remote Process Calls (possibly wrapped in
>> XML, usually via the Universal Firewall Bypass and Transversal
>> Protocol [HTTP/HTTPS]) back to 'machines of loving grace'
>> ensconced in dimly lit rooms
>>
>> Just my $0.02 ... sorry for the rant
>>
>> -- Russ herrold
>>
>>
>> [1] the trade press shows RHEL 5.8, possibly 5.9, certainly
>> 6.3, 6.4, and 6.5 [3 in one year -- 2012 -- to respond to the
>> 'it is too stale' pressure that any Enterprise distribution
>> ends up facing], as well as a new RHEL 7.  And that RHEL 7
>> will need to finish solving stabilization of systemd as a
>> replacement for init, and upstart ... and the author of
>> systemd did SUCH a fine job on PulseAudio, that we trust him
>> ... why?
>> _______________________________________________
>> colug-432 mailing list
>> colug-432 at colug.net
>> http://lists.colug.net/mailman/listinfo/colug-432
> 
> _______________________________________________
> colug-432 mailing list
> colug-432 at colug.net
> http://lists.colug.net/mailman/listinfo/colug-432
> 