Paid Development in Open Source

[Dale_K]

dale_k, you seem to be basing your arguments on an assumption that hardware manufacturers are paid directly to support one platform or another, which is rarely the case (ie microsoft does not pay a hardware manufacturer to write drivers for windows). each platform has an associated developement cost for them to support it, and they base their decisions on what platforms have the market penetration to make their time worthwhile (as in number of units sold in that market segment). so it is a matter of market penetration, not direct monetary compensation.
I think sponsored open source developement provdes a great way to get around this limitation even for hardware drivers, but it takes copperation from the manufacturers. for other general software like LinuxMCE it is a proven model that works well in the real world.

Agreed, that's the point I was making with the Nvidia example, but it's a cascading effect.  Open source porjects like LMCE will never be a huge market share because it's not commercial.  There is no entity advertising it on TV and in magazines.  There are very few companies even installing it on a commercial level.  This will sound like a catch 22 but part of the reason it's not commercial is slow development which is due to lack of resources because it's not commercial.  Not saying it's a bad thing, just how it is.

I'll wager that if MS purchased Pluto/LMCE and began marketing/developing it there'd be tons of hardware supported. (That's what money can do for you.)

Again, please don't take any of this as knocking LMCE or the devs in any way at all.  I'm just pointing out the advantages of paid development as per the original topic.  I am a big believer in paid for services and I think this project is no exception.

[tschak909]

There is a certain freedom here, that commercial projects never have, and that is that its developers are free to extend the system in any way they see fit.

My role here on this project is as a researcher. This is my one and only agenda. It is not for commercial gain, and will never be. This system is not something that can be sold en-masse yet, because of a large number of factors, and we understand this. It still doesn't stop us from spending the next 10-20 _YEARS_ making this possible.

Yes,

I'm thinking a LOT MORE LONG TERM than any commercial software developer. You should too. Today's Free Software, unlike the vast majority of today's commercial software, will STILL BE AROUND and BE USED and extended. We sacrifice the glam, and the fame today, for the price of doing it right for tomorrow.

This is what is important to us.

-Thom

[Dale_K]

The simple fact is that there is NO way in HELL that a single commercial entity could produce a piece of software, this large, and monolithically support it.

Pluto tried.

Pluto died.

It has taken a community of people to be able to continue development, to be able to develop the pieces that Pluto simply did not have the ability to, for a variety of reasons, including that certain features got lower priority due to pressure from investors, etc.

And I can tell you've never done any software development. In any project, open or commercial, the first 60% gets implemented quickly. The problem is, the last 20% are on the upper swing of an asympyotic scale.

Read that last sentence again.

The amount of time to fill in what is considered the last finishing details is the absolute killer for any project, because you're also debugging during this time, and having to support others during this point. It makes NO difference whether the project is commercial or not. The only difference lies in the prioritization. You may think things are progressing faster if YOUR features are being worked on, but what about SOMEONE ELSE'S features?

Stop being selfish and think about the bigger picture.

-Thom

Thom,

I think you're seriously misunderstanding what I'm saying.  I think that because your replies are turning slightly aggressive.

Of course I've never done any programming/developing, if I had, why would I have helped pay you for development.

As far as a commercial entity not being able to develop and support LMCE or a project like it, I disagree and we'll probably just forever disagree on that point.

I certainly understand that the last 20% is the most difficult.  But I bet it would be much less difficult if the devs were paid to do only that.  Imagine if LMCE stuff wasn't what you did after work but it's what you do FOR work.  Am I to believe that more wouldn't get done?

Please don't take my posts as any sort of attack on the project or it's devs.  Quite the contrary, I'm making the point that development would go faster and be broader if more people paid for it.

I think of my own profession, if a friend asks me to help him with his network and I know it will be free.  I'll get to him when I have the time.  If, on the other hand, someone needs help and they're paying me $150/hr, I'll MAKE the time.  Paid for stuff happens faster, that's the only point I'm making.

[los93sol]

Please note this thread is not intended to thwart the way things are currently done.  I wanted to get a discussion going about other ways to do things and constructively discuss this topic to see what good can come out of it.  At this point I'm voting to get the initiative back underway to get an official store going.  We have already seen at least one person here who was able to get around the licensing issues by not branding the gear as LinuxMCE compatible and not selling pre-installed systems.  Instead selling gear that is completely plug and play and known to work with a little effort (as long as it is fully documented and tested in the wiki) is absolutely a viable option as far as I can tell.  Anyone else interested in continuing this discussion?  I know it has been hashed out a few times in the past and maybe the best thing to do is dig up the old thread at this point if people are interested.

[colinjones]

Cool it everyone, please! If the thread turns nasty or goes further off topic, I will have to lock it. I am considering splitting out my comment into a separate thread...

[totallymaxed]

dale_k, you seem to be basing your arguments on an assumption that hardware manufacturers are paid directly to support one platform or another, which is rarely the case (ie microsoft does not pay a hardware manufacturer to write drivers for windows). each platform has an associated developement cost for them to support it, and they base their decisions on what platforms have the market penetration to make their time worthwhile (as in number of units sold in that market segment). so it is a matter of market penetration, not direct monetary compensation.
I think sponsored open source developement provdes a great way to get around this limitation even for hardware drivers, but it takes copperation from the manufacturers. for other general software like LinuxMCE it is a proven model that works well in the real world.

Agreed, that's the point I was making with the Nvidia example, but it's a cascading effect.  Open source porjects like LinuxMCE will never be a huge market share because it's not commercial.  There is no entity advertising it on TV and in magazines.  There are very few companies even installing it on a commercial level.  This will sound like a catch 22 but part of the reason it's not commercial is slow development which is due to lack of resources because it's not commercial.  Not saying it's a bad thing, just how it is.

I'll wager that if MS purchased Pluto/LinuxMCE and began marketing/developing it there'd be tons of hardware supported. (That's what money can do for you.)

Again, please don't take any of this as knocking LinuxMCE or the devs in any way at all.  I'm just pointing out the advantages of paid development as per the original topic.  I am a big believer in paid for services and I think this project is no exception.

Dale_K I think you seriously have the 'wrong end of the stick' here. There is so much OSS out there touching your life...'its just that your blind to it. Its sitting behind almost all the routers, Wifi AP, Switches, PBX's, Printers, Photocopying Machines, PVR's, Satellite Receivers, In-Car-Management systems, the vast majority of the internet is hosted, routed through, switched and packetised by OSS. This software is everywhere...but in general its on few Desktops...more now that ever...but still far less than commercial software.

Also if you look at the software that makes up many of the basic building blocks of LinuxMCE itself you will see that these same packages are used in all manner of software applications...but they hide themselves away behind the scenes and many users are unaware of them at all...and there is nothing wrong with that as the UI is supposed to insulate the user from then. However without these OSS components LinuxMCE could not exist... and many of these same packages also form the underpinning of that big list of products that use OSS...without them the internet and the vast majority of consumer and professional products that connect to it would disappear. Collectively these products and the OSS software that makes them possible are a much bigger market than Windows.

Lastly I think Thom's initial response to your first posting in this thread was a reaction to your comment  about LinuxMCE compatible hardware becoming harder and harder to find. What you have to realise is that there is an enormous effort underway to transition from a LinuxMCE that was born inside a commercial organisation and only partly re-engineered when it was released on Kubuntu ie LinuxMCE-0710. The hardcore software Dev team is doing an incredible job trying to move this enormously complex amalgam of software away from its Pluto derived roots and to take it to the next stage where the whole project is built and managed by the community. This process is an incredibly long and complex one and it is this work that has delayed the release of the 0810 version. However once we have the tansition to 0810 behind this immediately opens up LinuxMCE to much newer hardware platforms and devices and prepares the way to future release on 0910 for example.

So please take all of this into consideration when looking at this project and its development path.

All the best

Andrew

[Che]

Yeh, but OSS is hurried along by donations, and seeing as we are neither requiring or even advertising donations at the moment (other than a very small link on the main page), I do not see how this goes against Libre software ideals. It's merely showing appreciation for (an amazing) piece of development, which will also speed up and improve said development. Mozilla is an example of this. Entirely open, but also very fast development time. Why? Donations.

I truly believe that LMCE could be the Firefox of the Media Centres, as it is just so damn amazing, how can't it be? All we need is more hardware suppport, and less bugs, and this really will get a HUGE market share. I mean that truly and honestly.

[totallymaxed]

Yeh, but OSS is hurried along by donations, and seeing as we are neither requiring or even advertising donations at the moment (other than a very small link on the main page), I do not see how this goes against Libre software ideals. It's merely showing appreciation for (an amazing) piece of development, which will also speed up and improve said development. Mozilla is an example of this. Entirely open, but also very fast development time. Why? Donations.

I truly believe that LinuxMCE could be the Firefox of the Media Centres, as it is just so damn amazing, how can't it be? All we need is more hardware suppport, and less bugs, and this really will get a HUGE market share. I mean that truly and honestly.

Well I am 100% for supporting development by donations...we've always donated in various ways to this project. We also contribute back software we've developed ourselves or supported the development of by the community.

Andrew

[alx9r]

With respect to merits of paid development in FOSS projects:

Having a means of supporting a FOSS project in a way other than by contributing time (i.e. by donation) is important because it can increase the likelihood that a popular feature will actually get developed.  Surely this idea is actually some well understood economics axiom, but I will try to explain it in my own words.

If the only way a person can contribute to a project is by actually writing code, every person who does not have the skill to do such work is eliminated from the pool of people who are able to contribute.  I greatly benefit from LinuxMCE and would like to contribute somehow. I do not, however, have the skills (at least not yet) to actually write code to further its development. 

If a person can also donate money as a way to contribute to a project, look at all the people who can contribute to the project: All people who have money they would like to donate AND all people who have specialized skills and time they would like to donate. 

I suspect that currently the group of people who would like to contribute to the development of LinuxMCE but do not have the specialized skills required to actually work on the code-base hands-on is significant.  By creating a means of transferring money from this group to members of the group of specialists who actually write the code, the specialists would be able to spend more of their time on LinuxMCE than without the cash injection.  More time spent working on LinuxMCE by the specialists means a better LinuxMCE product with more features.

If those who are donating money can somehow steer what development it furthers, then features that are popular will garner support in the form of money. 

The previous sentence is important.  Without a donation system, the only features that will be implemented are those that are important or interesting enough to people with the specialized skills to create them.  However, with a donation system, there is a significant chance that features that are important to all users but perhaps not to developers will be implemented.

A good donation system should both increase the number of users whose contributions can be harnessed, and make the list of features that is actually implemented more democratic.


Alex

[USMC96]

First off I haven't even gotten Linux MCE functional yet, but I am working on it. I have made some rookie mistakes and I am currently trying to educate myself before spending more money!

I too agree that $ makes the world go around.  Personally I am the kind of guy who will buy a $50 book/manual to tell me how to repair or build something worth $40 because the knowledge is worth it.  I however have minimal (compared to those who will read this) hardware knowledge, And no knowledge of writing code, compiling, etc.  I am planning to try and learn, but in the meantime I am incredibly thankful for the support of the various open source communities.

The money donation to make a driver or other tech related experience already exists, but the point this post has repeatedly mentioned is a "general fund" sort of concept.  I myself would donate on a regular basis, I Love this program (even though I don't really have it functional yet) and I wish to give back to the community that has helped, and I am sure will help me in the future. 

My biggest thought is how would such a general fund be disbursed?  Who would get paid?  Would there be enough money to pay everyone? I doubt it, for reasons mentioned above re: pluto.  So who would be in charge of disbursing funds? And bear in mind their is a huge difference in doing something for the community which gives you the warm fuzzy feeling inside, and getting paid $1 for days of writing code, and checking it's function.  Or maybe it could be used to purchase new hardware to check for compatability? But who gets to use that equipment if it works? 

Most people are not happy with the way their government spends money, because governments are so big....so is this!

While I support the idea the general fund concept may cause more problems than it may help, people may start to feel they should have a voice (or a vote) in the way it is used.

Just my two cents.     A HUGE thanks to the devs, and everyone who works so hard on this!!