[GLLUG] Why I left Ubuntu ~ Everyday Linux User

Sat Jul 20 21:11:12 EDT 2013

On Jul 20, 2013, at 1:19 PM, Matt Parrott <parrott.matt at gmail.com> wrote:
> Do I have evidence that the Linux desktop has failed? Pick a metric.
> 
> Saturation? Satisfaction? Performance?
> 
> I was being broad about things, but since we're asking for "evidence", I'll step back and ask you to make (or point to) a case that the Linux desktop has been anything short of an unmitigated disaster. I can't prove a negative, can I?

You stated: "...the open source community has proven itself utterly incapable of creatively and effectively adapting its core principles to the desktop, and has failed spectacularly."  I don't see evidence that the open source community has 1.) shown itself ineffective of adapting its core principles to the desktop and 2.) failed to adapt its core principles to the desktop.

Firstly, I have to wonder what you define as the 'core principles' of the 'open source community'?  Freedom of choice?  Free and open code?  Community support for projects?  Licensing that benefits the community?  There's so many ways this could be interpreted.  I'm just going to choose the 'Free' part, both in choice and in code.  

Also, you don't tell me what you constitute a disaster.  That you don't like it?  It's unpopular?  It fails to compile?  It is just vaporware?  It's aesthetically unpleasing? 

I think the overabundance of desktop environments actually shows that the open source actually is negatively impacted by its own success, but it certainly has been effective at producing desktop environments that live up to the principles of the open source community.  

When you says "A is B", and I say, "provide evidence that "A is B", there's nothing in my question to that asks you to prove a negative.  In fact, both of your statements would be easily proved false by supplying one single instance where a project either 1.) shown itself effective of adapting the open source community's principles to the desktop or 2.) succeeded to adapt said principles to the desktop, which when you think of it, is really the same thing as #1.  I'll just choose KDE (http://community.kde.org/KWin/Wayland) because I personally like it.  Getting KDE to work on top of Wayland has been an open source community effort to adapt its codebase to the new display environment.

Perhaps a better argument to be made would be that the core principles of the Open Source community are not ideal in producing a unified desktop environment that can be consistently applied across all devices.  Part of that is because people who tend to use Open Source products also tend to like things like choice.  Also, not all devices are ideally suited to run the same interface as the other devices, which is probably best illustrated in recent news by the beating Microsoft is taking on its Surface tablets and Windows RT/8.

> For example, much of the Scientific and
> Engineering software I support use extremely complicated interfaces
> and their own windowing toolkits, on both Windows and Linux.  I see
> little incentive for these companies to turn them into browser-based
> applications.
> 
> You say that they all have their own custom toolkits like that's a good thing.

It's a fact of life, I'm not thrilled but I'd rather see vendors providing Linux apps than not.  I support the OS where a lot of extremely complicated (and sadly, expensive) products run, and they are fragile and break even between minor releases of X and related libraries.  If anything, a very common way of producing a common codebase for interface elements is to use Java, and that has its own failings.  

I have a hard time believing that this wouldn't be the case with a browser-based desktop.  The broken dependencies will just be pushed into the browser/server.

> On top of that, most of these browser-based desktops are NOT EVEN OPEN
> SOURCE.  Sure, parts of the Chromebook's OS are based on source that's
> open, but to get it to work, you have to rely on Google's
> infrastructure, which is closed.  Google Chrome itself has parts that
> aren't open.
> 
> It would be trivial for the open source community to emulate and improve upon ChromeBook's model. Believe me, if they did, I would be piddling with it as soon as possible. I consider myself a temporary refugee from the open source desktop until the community stops trying to compete with Apple and Microsoft for the most bloated, cluttered, and blingy experience possible.
> I consider myself dogmatic about what I consider to be open source principles. Unfortunately, ChromeBook's desktop is more closely aligned with them than Unity, and Google Docs is more closely aligned with them than LibreOffice. Unity is an ugly monolithic attempt to ape proprietary software, just as LibreOffice is an ugly monolithic attempt to ape proprietary software. Google Docs has a very flexible and open API that empowers one to quickly and easily stitch some fantastic custom solutions together from the available parts. Meanwhile, LibreOffice is a poor man's Microsoft Office.

I'd like to point out here that Unity is certainly not representative of the Linux Desktop, particularly as an Open Source Linux Desktop -- it was foisted upon Ubuntu at the protest of the Gnome upstream developers.  What you're seeing there is actually less of a community-driven project and rather a corporate-driven project.

I personally find LibreOffice quite usable, and I use Google Docs quite a bit.  My wife prefers LibreOffice over whatever absurd interface Microsoft Office has chosen for their latest incarnation, and she's not alone.  And for what its worth, Google Docs tries to ape the same proprietary software as LibreOffice does (and provides fewer features) -- so I don't really count that as a win for GDocs.

> Code which happens to be open isn't necessarily more open in an overarching context than closed code. Ideally, I would like a completely open stack, but I'll settle for a closed stack which allows for an open workflow than an open stack that emulates a closed workflow.

What's an 'open workflow' and 'closed workflow'?  Is your software open or not?  No, it isn't.  It could be taken away from you at a moment's notice, and you'd have nothing to do but try to find legal means to get it back.  Is your content being shared with the NSA?  Likely.  Can you do anything about it except take your docs out?  No.

The motto of the 'Browser Desktop' should be: Giving up openness, freedom and privacy in the name of convenience and aesthetics.

--
Jonathan Billings <billings at negate.org>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.egr.msu.edu/mailman/public/linux-user/attachments/20130720/6ba2f8a5/attachment.html>