[CS-FSLUG] Linux Today - Is desktop Linux too fragmented to succeed? A friend tells it like it is!!

Eduardo Sánchez lists at sombragris.org
Sun May 3 00:00:26 CDT 2009

I don't think McAllister is right, and while worthy of respect, many of 
Tim's objections are not as weighty as one might think.

Let's go over them.

On Friday 01 May 2009 15.57.16 Timothy Butler wrote:
> I think Jerry is wrong on this one.
> Simply put, McAllister hits on points I've been saying (or being told
> by users) for years. It was interesting he mentioned tax software.
> How many people do I know that struggled with WINE or had to boot up
> VMware to do their stupid taxes? It's those sorts of little things
> that help to disqualify Linux by the death of a thousand
> qualifications.

Huh? In my country I can do all my taxes on GNU/Linux. Usage of Free 
Software and fair access to all browsers for the tax agencies is a 
matter of national policy. The point is too U.S. dependent.

> My parents can't switch to all Linux, because of things like tax
> software. I moved my mom over to Linux for a year or so. I
> "converted" to Mac OS X during that time and she ended up following
> because she was constantly frustrated at having to have me come look
> at her machine to do "one more tweak" to make it do stuff.
> She is now a Mac evangelist. She loves the "just works" photo
> downloading, easy music management, the compatibility, etc. Despite
> my best efforts (and at the time, still a belief that Linux could
> work for almost any desktop user), I couldn't replicate that
> experience using Linux.

Mac OSX is not an option due to cost. I know that there are good price-
conscious choices on OSX on the U.S., but in most Latin American 
countries, and glaringly so in mine, AFAIK, Macs are horribly expensive, 
overpriced machines. Perhaps before, when they had a great RISC cpu, 
they were worth the price; but with a good PC you can get several times 
the bang for the buck. Paying top dollar for a great UI doesn't make 
economic sense here.

As for GNU/Linux, I have a problem replicating the common user 
experience, but this is because I'm just getting old. Things such as 
video, webcams, digital camera and photo management, management of 
massive music collections with playlists, etc., are things that do not 
interest me. But when I carried a digital camera to take pictures of my 
vacation trip, I just plugged the camera, and voilá! it was automounted 
and I could manage and process the pictures. I didn't use wireless in my 
laptop, because I assumed it wouldn't work in Linux, because it didn't 
work on XP. But after installing the proper firmware (a one time step), 
it worked flawlessly. So yes, GNU/Linux works, and works great, as an 
end-user desktop.

> IP constraints are real. I can boot up a new Mac, and not only will
> it be (by most people's admission) the most beautiful looking desktop
> available, it will also play virtually any media file I throw at it
> without having to download or configure anything. This is powerful.

I can do the same in Slackware. No downloads ever.

> People don't want even easy downloads. They just want the system to
> work. They want to download an AAC file from the iTunes store and
> sync it with their iPod Touch.

But we don't even use the iTunes store, it won't even accept our credit 

> They want to install TaxCut directly
> from the CD with no tweaking (unless the system does it
> automatically).

It won't work, but it also won't work in different Windows versions in 
too many occasions. Besides that, I think it's acceptable that different 
systems should have different software.

> They want to open the computer up from sleep mode and
> have it work in less than five seconds.

Works on mine.

> I think McAllister's other point is valid too. One person uses
> Ubuntu, another uses Kubuntu, another insists on openSUSE. You can't
> just pick up a single book and learn how to do everything in each of
> them, because each one works differently. Fragmentation has been a
> major issue for years.

This is a non-issue. Each distro should be well documented and that's 

> That most distros have centered on GNOME (and
> KDE marginalized itself with version 4; in the words of a friend,
> managed to do the impossible feat of making Vista look good) doesn't
> solve it.

I think you are not fair with that. You should really try KDE 4.2.2. 
There is nothing marginalized with that version.

> Which I say with disappointment. I thought (and still think) Linux is
> a great solution, but as long as most of the developers are so busy
> ignoring what the users want, Apple will continue to be the only
> provider of a successful desktop *nix.

That's not true. KDE and GNU/Linux are already providing desktops in 
Brazilian schools in the order of *millions*, and people are using 
GNU/Linux desktops everywhere, in both corporate and personal 
environments. This is due not only out of technical excellence, but 
thanks to Linux's best desktop apologists: The mobsters and racketeers 
of the Business Software Alliance and their wretched "licensing audits".

> A few observations on what Linux needs:
> 	1.) It needs someone to pour millions of dollars into WINE and bring
> it up to the point that it can run current versions of Photoshop,
> TaxCut, Quicken, etc. It then needs to be tied into the DE so closely
> that the end user is blissfully unaware of the compatibility layer.
> Focus on apps that really cannot easily be replaced with something
> already on Linux and only worry about currently supported versions --
> no one cares if Office 2002 runs on Linux any longer.

A great suggestion, but wrong on one point. I do care if "Office 2002" 
runs on Linux. There are many installations with Office 97, 2000, and 
the like in my country. Many businesses purchased one, and just one 
version of certain software packages, and their business depend on them. 
WINE should be very friendly to old software (¿maybe forking Wine into 
two releases, one for older software, and other for the state of the 
art?) Visual Basic is a must, in several versions, and so are XBase 
engines, such as Visual FoxPro, and others like Genexus.
> 	2.) It needs one of the two major DEs to be completely marginalized
> so that there is one standard "Linux desktop."

I respectfully disagree. I wouldn't want GNOME to be marginalized. Its 
usage is really marginal outside the Ubuntu/USA realm.

> 	3.) It needs a major OEM to make Linux its flagship OS and focus
> totally on making really slick, desirable systems (think MacBook Pro
> or even Dell Adamo -- systems fully capable of being "cool") that
> come with it pre-installed.


> 	4.) It needs to come with pre-installed, licensed codecs for MPEG-4,
> QuickTime, DVDs w/ CSS, etc.

Or not. This would be true only when you have some law such as the DMCA 
Anti-Circumvention clause. But everywhere else in the world, if you have 
a legal right to play media, it's up to you how you play it. So instead 
of "licensed", I would put the emphasis on "Free Software codecs".

> 	5.) It needs to come with one best of breed app for every job, never
> two sort of OK ones. If one cannot stand by itself, don't bundle any
> (except through a download service, see below).

That would be nice. But offer the alternatives, not on a download 
service, but on a CD. Bandwidth is scarce and expensive in the rest of 
the world.

> 	All of this is actually quite doable. Here's what Ubuntu should do.
> It should create Ubuntu Pro for $49 off each LTS release. Ubuntu Pro
> would be a paid product so as to buy those codecs. For the moment, it
> could also include a license of VMware (but, with Canonical working
> with VMware to make it "rootless," much like VMware and Parallels for
> Mac).
> 	On top of that, it would launch the Ubuntu App Store, which would be
> an attractive, GUI driven app store much like iTunes Store is for the
> iPhone/iPod Touch. Leveraging the existing apt-get architecture,
> Canonical could make waves as being the first to bring the same ease
> of installation that iTunes does for iPhones to a desktop OS. They
> could make it so, like the iTunes Store, app developers could pay $49
> to join and sell apps for a 70-30 revenue split. this would generate
> huge sums of cash for Ubuntu, which could use that to drive
> development, but it would also make it easy to incentivize the same
> sorts of creative developers who helped Apple distribute 1 billion
> iPhone apps in less than a year.
> 	Likewise, Canonical could sign a deal with Amazon to build in the
> Amazon MP3 store (think as it is on Android, which in turn is copying
> iTunes) and perhaps even Amazon Unbox. Integrate shopping directly
> within Rhythmbox so that it is as good or better experience than
> iTunes on Mac/Windows.
The problem is that this product would not be free. (Free as in freedom)

> 	The last step is finding an OEM. Dell needs a new plan. Work with
> them to launch a complete line from netbooks to luxury laptops of
> systems with Ubuntu Pro preloaded. With Ubuntu's subtle, tasteful
> artwork team, it'd look nice, be very functional and could be slotted
> in price maybe $200 less than the equivalent Apple. These systems
> shouldn't be Windows or Linux systems -- a totally separate Linux
> line with its own unique features That would perhaps sell some in a
> bad economy...
This is really interesting.

> 	Here's an idea that would make it killer: find a way to take one of
> those instant boot "light" Linux systems  and find a way to make it
> so that it could seemlessly fade into the full distro that could boot
> in the background while the user was already doing stuff. Instant on
> could be a killer app.


> 	One last thought, while I'm busy musing. Linux needs a killer app.
> Apple has the iPod. Windows had Office. Neither was totally unique,
> but both gained a following that drove people to adopt the platforms
> they worked best on. Linux needs something that people want. Linux
> needs to be aspirational rather than utilitarian. That'll sell.

Oh man... All we have is just Emacs. :^) Nothing else ... That's bat. We 
really need something like that. Perhaps some of the KDE ioslaves magic? 
A real GUI frontend for LateX?

This is gread food for thought, but some of the points are not entirely 
correct. I am especially adamant on insisting that KDE 4.2.2 is a 
terriffic desktop environment, and perhaps the best desktop the world 
has to offer now, be it Free or proprietary. Marginalizing it would be 
an awful strategic failure, IMHO.

Let's keep the conversation open.


Eduardo Sanchez, B. Th.
Traductor Público Inglés-Español
 With Earth's first Clay They did the Last Man knead,
 And there of the Last Harvest sow'd the Seed:
   And the first Morning of Creation wrote
 What the Last Dawn of Reckoning shall read.

	-- The Rubaiyat of Omar Khayyam


More information about the Christiansource mailing list