The HeliOS Project is now.....

The HeliOS Project is now.....
Same mission, same folks...just a different name

Search the Blog of helios and all comments

Loading

Sunday, March 31, 2013

Resistance Isn't Futile


I've spent a lot of time in conversation with various people about the State of Desktop Linux.  While I have my own ideas as to how we (could have) gained a larger market share on the Desktop, my firm belief is that, at this time.....

Anything we do from now on will be too little, too late.

That ship has sailed, it's water under the bridge, however you wish to express it.

Linux, as we know it, is not going to ever become a major player in the Desktop market.

I've made peace with that.  Since 2005, I've chewed my lip over how this can change, but the fact is, there are simply too many people not willing to do the things necessary to make it a reality.  Whether you believe it or not, whether you like it or not...

The absence of A Linux Product that's visible in the consumer market has been our greatest failure.  Now, I will be the first one to admit, I never saw this coming.  Linux is a player on the desktop, or laptop; as you prefer.  It's just not called Linux...

It's called Chromebook.  It's called Android.

Pick all the nits you want, argue that it's only the kernel and not the system.  The fact remains that Linux has found a place in the consumer market.

it's just not the way many of us would have liked it to come about.

Some will blame the fragmentation of the Linuxsphere.  Some will say that too many choices have diluted the product and confused the customer.  I'm not going to disagree with that theory.  Friend and colleague Dietrich Schmitz argues that point strongly.

So if we are to embrace that particular theory, what is to be done?

Let's fragment it some more.

I'm not being flippant.  I mean it and I fully support it.  If the damage is already done, as we watch our train picking up speed without us on it, then I will argue:

Fragment away...what's the downside?

Well, that's just what SolusOS creator Ikey Doherty has done.  But he hasn't simply used another major distro as a base. He's tore the whole thing down and started from the beginning.

The way he thinks it should be done.  A pure fresh start from top to bottom, bottom to top.

Doherty isn't some hack that's taken the remastersys ax and chopped up a distro to claim as his own.  Ikey is a brilliant distro developer and he's had a nagging, almost compulsive belief that his Linux creation could be better.  Constant upstream battles with both Gnome and Debian have brought Ikey to the point where he spends more time patching and shoe-horning than he has in creating.

His Solution?

It's easy PiSi.

Let me turn this over to Ikey Doherty....he makes it abundantly clear about why this change, and why he's doing it now.  Ikey knows there is dissension in the ranks, but he believes his vision for what SolusOS can be, is the best of all choices.

Why Pisi and Why Independence?

SolusOS has always strived to provide the utmost stability to its user base. In a turbulent Linux ecosystem where desktops change almost daily it seems, stability and longevity become core concerns.

Another one of our core aims we developed with Eveline (SolusOS 1) was modern user space software, such as Firefox, Thunderbird, LibreOffice, etc. However, the man hours that went into maintaining the repository were not justified for a 'derivative'

Derivatives (in this case, of Debian) typically provide the user with a custom distribution of their parent distribution. Invariably a repository does not exist, or a small PPA is provided in the case of Ubuntu. In our case, we have over 7k packages in the Eveline repository, and it's become very much apparent that keeping our changes in sync with upstream changes is literally an uphill battle.

So with all that in mind, we realized our core issue within SolusOS was the package management system. Our single biggest task was maintaining a repository, and doing so with the .deb package format made it a rather mundane experience attempting to keep our system in sync with Debian.

For a long time I've been a lover of PiSi. It's an incredibly powerful, yet simple, package manager, originally created by Pardus Linux. With simplified package creation, powerful dependency resolution, roll back of operations, delta packages, a powerful configuration system (COMAR) plus many more features, its an awesome piece of kit. The challenge was to make it fit in with SolusOS.

So, we opted originally to mold Debian to PiSi. This turned out to be an impossible task, with many packages relying on other Debian specific packages, patches and makefiles to even configure, the situation became impossible. We were no longer using either the parent distribution or the package manager to their full extent or showing them justice.

The logical step would have been to re-base on another parent distribution. However the same issue cropped up time and time again, sending us in circles. We could not make our goals and wants merge with any other project. There was no other choice.

A Distro Is Born

After many hundreds of hours I successfully built a new base system from scratch. This was no simple feat, as to run Python (required for PiSi) and D-BUS (for Comar), you need the majority of the base system in place already. Many failed attempts at bootstrapping the GNU toolchain (compiler/glibc/etc) eventually proved fruitful. I had a system I could replicate with 100% accuracy and chroot into.

Then we tried the insane. Let's make PiSi work with this. And we did :) We've spent a long time now preparing package after package, buildsystem after buildsystem. And now it all works. With an independently booting image, utilizing its own init, kernel and ramfs, we had a base. Everything in the core structure has been tested time and time again, leading up till now. We have something that is truly unique. This is not another derivative of an upstream distribution, nor is it a fork of anything currently existing. The entire distribution has been put together by hand, with our own packages sourced from the original codebases, with a vanilla policy not allowing us to clutter up our packages with items we deem not necessary (Hey, it worked for Slackware ;))

So, we have a fully networked desktop-enabled system now. Amazingly 3D acceleration still performs well under Qemu, we have a very solid and stable codebase and are adding packages daily. With over 700 commits to our source repository in less than 2 weeks, there is no worry about missing packages or a lack of stability. We have a vision and direction that's kept us unique from the birth of SolusOS 1 up till now, and with our own base and packages, we'll be able to show the world what the original vision was all that time ago. Change is coming, and its name is SolusOS. We're ready, are you? :)

So, there it is.  While Ikey won't speculate on when SolusOS 2.0 will be ready for prime time, he does plan to release a beta in the near future.  Will Ikey be able to stay the course?  Can a "one man show" stand the test of time?

Actually, SolusOS isn't a one-man-show.  Tech guru Justin Krehel  has been working on SolusOS for quite a while now.  With the gaining popularity of SolusOS, it was only a matter of time before Justin became more widely known as a key contributor to the project.  Stay tuned in the next two weeks.  We plan to do a feature on Justin.

He is quite brilliant and I think you'll appreciate knowing who he is and what he does.

All-Righty Then.....





9 comments:

Kevin (Whizard72) said...

GNU/Linux is, always has been and probably always will be, a power user system. It's for people that understand computers and technology and have specific needs that can't be filled by Windows or Macs. Linux is soundly entrenched and rightfully so in the server and supercomputer worlds but as a desktop there's too many people with too many conflicting ideas to allow that to happen. Maybe that's for the best though, do you really want your Linux turning into Mac OSX where it's made specifically for dummies?

The single biggest hope desktop Linux has lies with Canonical and Ubuntu. Unless we get OEMs willing to ship computers with Ubuntu, it'll never get market share. People don't want to install it, they want to buy a computer with it already present!

The Stallmanite freedom spouting is altruistic and I happen to agree wholeheartedly with it, but it's NOT THE WAY OF THE WORLD AND NEVER WILL BE IN OUR LIFETIMES. Especially given the backward steps our Gov't and President are currently taking in terms of Freedom.

I'm okay with desktop Linux never turning into a dumbed down brady bunch OS with no options that Mac OSX is. I'm okay with having computer skills that beats 90% of the general public out there.

To address the future: Nobody knows what that future is. Linux isn't going anywhere. Mac OSX isn't likely to go anywhere because there are plenty of brainwashed Apple fanboys out there that think they're expensive junk was worth every penny they overspent. People still won't want Windows 8, I can promise you that.

Gary Newell said...

I really like SolusOS. It has that "Just Works" feel similar to Linux Mint.

I wrote a review on my blog http://www.everydaylinuxuser.com/2013/02/everyday-linux-user-review-of-solusos.html.

There are lots of distros out there that are just rehashes of a theme.

The really good projects always rise to the top in terms of popularity and SolusOS is a really good project.

IGnatius T Foobar said...

The unification of the Linux desktop was jettisoned almost singlehandedly by Miguel de Icaza, who (probably after being paid by Microsoft to do so) created the GNOME/KDE schism, setting open source desktop adoption back by enough years to let Microsoft and Apple make their next moves.

And also ... +1 to what Kevin (Whizard72) said about the Stallmanites. Really, whenever someone says "GNU" in anything other than a historical context, they need to be repeatedly slapped in the face until they get pragmatic.

As for desktop Linux in general ... it's not going away anytime soon.

Anonymous said...

I am an average computer user. I found Solus and was impressed enough it is now my main OS as I have no desire to use Win8. I mostly do a little browsing, e-mailing and a lot of music stuff. It is pretty easy to use. I have tried a lot of Linux OS's to find one I liked and this one and Zorin are my favorites. I actually think Solus is the easiest to use right off. Yet it still has everything the power user needs.

Mike said...

"Fragment away...what's the downside?"

Amen! I could not agree with you more. Another excellent article, Ken.

Anonymous said...

I don't understand, how replacing one package management with another will fix Linux issues - you then need to maintain 20,000+ applications and a dozen hardware architectures by yourself as Debian community is doing and you create another fragment.
Everybody tries to create super-duper packaging system and at the end of the day Debian wins hands down. This is lost effort.

Ikey Doherty said...

"I don't understand, how replacing one package management with another will fix Linux issues - you then need to maintain 20,000+ applications and a dozen hardware architectures by yourself as Debian community is doing and you create another fragment.
Everybody tries to create super-duper packaging system and at the end of the day Debian wins hands down. This is lost effort."

These are clearly the words of a misinformed individual. I personally prefer to know who I'm replying to, not some Anonymous entity :)

If you believe that Debian indeed contains over 20k source packages you're been quite misled. Debian actually contains much less, as each package can be split into subpackages. The real number would be roughly 6k source packages.

Nobody has said we're trying to fix "Linux problems". We're addressing problems within our own distribution.
Also, PiSi is an already existing Package Manager, created originally for the use of the Pardus Linux Project, before the change of the turkish government.

Had you engaged in a little search before making your comment, you'd know this. In fact you could've just asked me and I'd be quite happy to tell you.

If you have any further questions, just drop me a line: ikey AT solusos DOT com
I'll be happy to explain in my own time anything you're unsure of or uncomfortable with :)

Anonymous said...

I run my business small computer needs on LAMP, our computers run Linux, our home computers run Linux as well as our laptops and our parents computers. Red Hat, KDE and family are part of our everyday lives. We have Android on our phones and on the tablets.

Sorry but we won.

Doesnt mean we should stop or not strive for more but we have made open source and free software 'normal'.
So normal that most people dont even know that that the linux kernel is everywhere.
Cmon, how many Tivo users knew they were using Linux? 1%?

Its ok.
So its not the glamour filled victory some people wanted. the one with press releases and auditorium full of media waiting to hear the gospel being spoken.

Ive seen Linux in action in places like Cuba, Venezuela, Senegal, Serbia, , Spain, Italy, Greece, Vietnam and China, Honk Kng and Australia (i work in heavy machinery sales and support).
Its universal and its impact in countries both rich and poor is immense.

Anonymous said...

I think we should be humble. After Linux conquered almost all of computing, a single locations resisted: The Desktop. I think a research group of legal and economic investigators could spend their lives studying why the Desktop resisted, but that has become only of historical importance.

Personal computing is switching to mobile phones and tablets. It seems Linux is already well entrenched there. This year, more Android phones will be sold than Windows desktops.

Ikey, your work looks great. Success!

Winter