thelinuxlink.net

Yesterday I was reading the latest DistroWatchWeekly newsletter and thought the commentary on what appeared to be happening over at Freespire brought up some interesting thoughts ( Link to article).

I have my own thoughts on this subject, but I'd like to find out what people here think about commercial Linux distributors funding/sponsoring "community" builds of their distributions.

Are they good ideas, or bad? Do you thinks it's a way for the community to help choose a direction in which commercial vendors develop, or is it just a way to get free labour for their product? Are they good for the Linux community or do they just take focus away from distributions that are true community projects (like Slackware, Debian, Gentoo)?

I'd like to read your thoughts, but please respect each others point of view. And I'm not asking for a distro war, either. Keep it as general as possible.

I really don't care that much. If a company is able to make money by supporting a OSS project then more power to them. As long as they contribute to the community and the end result is something that consumers want, not something they try to push their own goals upon. I think Canotical has done some good things with Ubuntu and same with a bunch of the other corperate sponsered distros. That being said the purely community driven distros can be just as good. Overall I think it is great that companies are finding ways to make money with open source and corperate involvement doesn't effect me negatively at all. I just use whatever I think is the best distro for me.

As much as some hate to say , corporate companies are a big reason the community exists, because nobody can build a box without purchasing it from some corporate firm.
Even if you buy it from system 76 , it still has a intel or amd processor.
It would cost too much to design and develop your own.

Of course this goes beyond the CPU, whatever is in your machine has to sell and in turn suit a corporate need.

Someone starts a new backup project, its not unusual to know that the project started because a company needs to store huge amount of business related work and there is nothing out there that fits there needs. That its open source is a bonus to us.

The exception to this rule could be media players, although any corporate linux seller knows that execs like to see a film or produce a company podcast, so a media player has to exist. I think mplayer etc was solely a community product.

A lot of the desktop enhancements by suse is to make the desktop usable in the corporate environment, of course that benefits the community.

There is the assumption for companies to use open source because its a free cost, no need for licenses and hiring someone to develop it or less people because the world can look at the code.

What corporate company releases there open source software using the BSD license, none, because its harmful to them. There are not guaranteed that someone will work on the project outside the company.

The corporate attitude to software is definitely changing, as long as it does the job , who cares what methodology is used to create the software. What customers want is support of the product and that's where corporate companies come in.

You make a piece of open source software that is very good but in order for X and Y to use it there need support, the product gets more known, big corporates here about it, offer customer service for this project and you the developer has less time to work on due to work etc and some company profits from your work.
What could happen here is, that you get hired by the big company or the company just takes over development.

What i've really dragged out is that we cannot really ignore the corporates, whatever project we use may be used by them or supported by them.
As long as its done in a healthy light, all should be fine.

All those lovely drivers we all love dearly, a lot of them were created because a company needed that device to work, not always, but surely often.

I think it's fine as long as the distro doesn't become a testbed for the corporate version. Hell, even then it's fine as long as that's understood by the users (and devs). I guess my take on it is this - say I absolutely love red hat but don't really think RHEL makes a good desktop - I'm kind of stuck, no? Using CentOS won't help in that case. Using fedora does, though, so it's good for me (imagining I was nuts about red hat) that it exists. The same can be said of suse, though I guess SLED is actually a nice desktop, from what I hear.

On the other hand, sometimes I do get concerned about the huge number of distros and therefore duplicated work. I know it's supposed to be a strength of linux, but I have to admit that sometimes I wonder. That being said, most of my favorite distros are *not* the mainstream ones so I'd be SOL if things were to change substantially.

At the end of the day, I think it's hard to place a (good|bad) judgment on the issue, but I think it's probably good for the users who want the option.

Judland wrote:Yesterday I was reading the latest DistroWatchWeekly newsletter and thought the commentary on what appeared to be happening over at Freespire brought up some interesting thoughts ( Link to article).

I have my own thoughts on this subject, but I'd like to find out what people here think about commercial Linux distributors funding/sponsoring "community" builds of their distributions.

To me, it largely depends on whether we're talking about a regular company, or an open source company.

Judland wrote:Are they good ideas, or bad? Do you thinks it's a way for the community to help choose a direction in which commercial vendors develop, or is it just a way to get free labour for their product?

Frankly, I think it's both. In the case of Fedora, I think it's a win-win situation, the community has some control and yeah like you say it, community contributions helps Red Hat build the next Red Hat Enterprise Linux distribution.

Judland wrote:Are they good for the Linux community or do they just take focus away from distributions that are true community projects (like Slackware, Debian, Gentoo)?

I think only a fool would argue that Red Hat has not been good for the community. They have invested an amazing amount of money into development of all kinds of things, the kernel, X, GNOME, likely GCC and also non-technical stuff such as legal issues etc. Sure they probably do this primarily to make their own product better, but I don't see how Mark Shuttleworth spending 10 million dollars, or however much it is, on hiring developers and creating a Linux distribution which he intends to make money off of by selling support services, is somehow better or purer (In "halo" terms) than Red Hat investing even more cash into hiring developers and creating a Linux distro which they then sell and make money off of, to make their own product better.

Judland wrote:I'd like to read your thoughts, but please respect each others point of view. And I'm not asking for a distro war, either. Keep it as general as possible.

We geeks need our flamewars!

I will have to say yes because I really appreciate Fedora's aggressive dev cycle. This is a community project but Red Hat sets the tone for that aggressiveness.

Of course, that's just what I like so I can't really fight it. If you like something different then cool. I think that limiting the choices we have would be bad for the community eventually.

Snarkout wrote: On the other hand, sometimes I do get concerned about the huge number of distros and therefore duplicated work. I know it's supposed to be a strength of linux, but I have to admit that sometimes I wonder. That being said, most of my favorite distros are *not* the mainstream ones so I'd be SOL if things were to change substantially.

I have the same thoughts, i much prefer the BSD approach

Each BSD is very different, OpenBSD is concerned about security, Netbsd is concerned about size, FreeBSD is concerned about scalability and a little about the desktop, DragonFlyBSD is concerned about clusters.

All four BSDs have very different kernels, there is some overlap between all of them, but there are different and concentrated effort is on of them to ensure each gets features and whatnot.

Having N^N linux distros is not idea, you can shout how good choice is but at the end of the day all linux distros are basically the same, same kernel with different config settings, and some patches, all distros can be turned into each other.

On the BSD front that is very different, you cannot set up yahoo with OpenBSD to do what FreeBSD does. You cannot put FreeBSD on to a router that NetBSD supports.

How many times have you wanted to do something on one linux distro and its been a pain but on another its not.

allix wrote:Having N^N linux distros is not idea, you can shout how good choice is but at the end of the day all linux distros are basically the same, same kernel with different config settings, and some patches, all distros can be turned into each other.

This is exactly what IS ideal. What you end up having is you have the software tested by as many people as possible, on as many hardware combinations as possible, and in as many environments as possible. If I only have hardware with 1 GB of RAM and I write a desktop environment, I'm not necessarily gonna care whether it can run on a system with only 256 MB of RAM, you can be damn sure that somebody in a poorer nation is gonna be concerned about that. Also, there's a lot of Linux distro that specialize in very niche areas. For example there's IPCop which aims to be a router, there's Trixbox which does Asterisk Vo-IP, there's BackTrack which is for cracking into your neighbor's WiFi, errrhhmmmm, security auditing, excuse me. Having a distro that does one particular task, and does it really well (Maybe by doing it in it's own little way) is a good idea, it means more options, and even a better user experience, because you as the user spend less time setting up your machine, and more time using it.

allix wrote:How many times have you wanted to do something on one linux distro and its been a pain but on another its not.

Try installing KDE on Fedora Core 6 and have the end result be just as good as SUSE's default without tweaking the default KDE settings you get.

Tsuroerusu wrote: This is exactly what IS ideal. What you end up having is you have the software tested by as many people as possible, on as many hardware combinations as possible, and in as many environments as possible. If I only have hardware with 1 GB of RAM and I write a desktop environment, I'm not necessarily gonna care whether it can run on a system with only 256 MB of RAM, you can be damn sure that somebody in a poorer nation is gonna be concerned about that. Also, there's a lot of Linux distro that specialize in very niche areas. For example there's IPCop which aims to be a router, there's Trixbox which does Asterisk Vo-IP, there's BackTrack which is for cracking into your neighbor's WiFi, errrhhmmmm, security auditing, excuse me. Having a distro that does one particular task, and does it really well (Maybe by doing it in it's own little way) is a good idea, it means more options, and even a better user experience, because you as the user spend less time setting up your machine, and more time using it.

Specialised distros that you mention is a different kettle of fish, there also exist on the BSD's, I am talking about the difference between damn small linux and say openSuse, its quite possible to transform both distros to each other. There is not anything vastly different in either distros , its just the cosmetics are different.

Tsuroerusu wrote: Try installing KDE on Fedora Core 6 and have the end result be just as good as SUSE's default without tweaking the default KDE settings you get.

See my above comment.


For anyone who is interested in a decent FreeBSD vs Linux benchmarks take a look at
http://jeffr-tech.livejournal.com/

As we are finding out with the heating embers at Debian and Gentoo, if you don't have some sort of figurehead leading the distros things can easily fall apart.

As great an idea as pure community distros are, there needs to be firm leadership once it hits a critical mass. A few people can compromise as long as all parties are willing to give something. And if the group becomes large enough and there are two (or more) immovable camps, forks can begin and weaken the whole.

When you have an Canonical or Red Hat leading a community distro it's a good thing. These are companies that have so far listened to their users and responded. They cannot be expected to please everyone, but on the tightropes they walk they have maintained a good balance. Decisions are made with the interest of the target audience for the distro, and not always those of the engineers and developers who have a narrower focus.

That's a good thing in my mind.

Getting back to topic (please start your own BSD/Linux thread if you'd like to discuss further)... because Red Hat / Fedora was mentioned, I'll pose this question:

Worse case scenario: What would happen to Fedora (all the work, time and energy put into it by the community developers) if Red Hat's business came under the strain and decided that too much of its resources was going in to keeping Fedora running?

Corporations have a lower "pain" threshold to losing money than, say, Patrick Volkerding. That is, I'm confident in assuming that Pat would tolerate a much lower return on Slackware than Red Hat would on it's own distro.

My concerns about community builds of corproate distributions is the unpredicability of what a corporation is willing to sacrifice in order to save itself.

Being corporately owned, its not likely that Fedora will break off to become its own distro. if Red Hat decided to stop funding the project. Same goes for SuSE and OpenSuSE.

There's no guarantee that Slackware will be here tomorrow either. But I would think it has a better chance at adaptation (or being championed by someone else if Pat decided to pack it in) than if it were owned by a corporation.

If I was a real believer in Red Hat or SuSE's products, then why not just buy Red Hat or SuSE Linux and put the support (money that is) where it really counts?

Just asking.

chuck wrote:As we are finding out with the heating embers at Debian and Gentoo, if you don't have some sort of figurehead leading the distros things can easily fall apart.

As great an idea as pure community distros are, there needs to be firm leadership once it hits a critical mass. A few people can compromise as long as all parties are willing to give something. And if the group becomes large enough and there are two (or more) immovable camps, forks can begin and weaken the whole.

When you have an Canonical or Red Hat leading a community distro it's a good thing. These are companies that have so far listened to their users and responded. They cannot be expected to please everyone, but on the tightropes they walk they have maintained a good balance. Decisions are made with the interest of the target audience for the distro, and not always those of the engineers and developers who have a narrower focus.

That's a good thing in my mind.

I agree decisive leadership is important, as long as its not a dictatorship.
If those two distros merged then more work would be done on getting the whole deal more polished and working.
The reason i say this is because an application would only need to be compiled once instead of twice, work would be more productive.
In the spear of things, kde is compiled of the amount of linux distros out there.
Its duplicating work IMO.

chuck wrote:As we are finding out with the heating embers at Debian and Gentoo, if you don't have some sort of figurehead leading the distros things can easily fall apart.

Agreed, as well.

Judland wrote: Worse case scenario: What would happen to Fedora (all the work, time and energy put into it by the community developers) if Red Hat's business came under the strain and decided that too much of its resources was going in to keeping Fedora running?
.

You have access to information i do not know

seriously, linux and BSD would still exist without corporate backing, although i cannot actually see that happening, because even when the linux kernel was in its infancy, people were still getting paid to work on it, because its a product there needed.

One thing for sure if big corporations backed out, the speed of development would slow down.

I have a snappy question, why are you so pessimistic about the future of free software?

allix wrote:One thing for sure if big corporations backed out, the speed of development would slow down.

I have a snappy question, why are you so pessimistic about the future of free software?

I think you're missing my question. I'm not concerned about Linux in the least. I'm also not against/concerned about corporate sponsorship of Linux projects (sponsors are not necessarily owners) or them employeeing Linux developers. We all can't be self employed

I'm asking, if you're a Fedora user or contributor, what would your reaction be if Red Hat pulled the plug on their Fedora project? My question is regarding Linux distributions that are owned by corporations but rely on community support/development and the dangers for those of us who use such distros. if said corporation decides "enough with the free stuff!"