ransom software release model - pay up or keep Windows!
 

This is a really interesting model for OSS releases. What if Redhat was to release like this? Would it work? Is there a better model? Does the current model (patronage) really work? Give us your views

What do you mean by redhat releasing like this?90% of their distro consists of Gpl'ed software and for the rest they charge anyway if you buy it.
Didn't hear about any Gpl programmers commiting suicide because of economic problems either.I don't think many of the guys that are into this right now would go for this - why share with somebody if you can always make some addons that are not under Gpl and need to be payed for?

The low cost of entry into the distribution market enables individuals to have GNU/Linux running in no time. But there are a lot of companies which don't have the skill to roll their own distro, yet have the money to pay for an OS. Because of the free rider effect though, they see little point in parting with their money.

This is why it will most likely never come true, but redhat targets businesses which are flushed with money but are not keen on investing in expertise to put these distros together. At the moment, how do business-oriented distros find their way into businesses? The IT Manager says "Lets pay money to redhat to get service on their distro". No. The sysadmin's lackey is told to set up the ftp overnight and burns it to disk in the morning.

So this is a compromise not to get individuals to pay but groups of individuals who have a common interest in getting (relatively) low cost reliable systems - businesses.

You've misunderstood the concept I think. This has nothing to do with the GPL and the GPL has nothing to do with making money (or not). Have another look.

Bert

Modified versions of this are being done right now.

Lindows is doing this in that the source release is being staggered behind the memebr's paid release. Mandrake is doing something related to this with the Club and software only for club members (and some of the roll-your-own RPMS requested are being done by members even). This kind of trend is going to continue because while the base software can be provided at relatively low cost via mirroring, those without heavy skills can find customizing it theri business needs to suit is a very challenging job. So, as usual precanned software that meets the needs of submarket segments gets paid for in the business world.

This is in part because of support costs-- business demands prompt and very detailed support, and canning anything (including software and O/S bundles) results in more detailed support needs.

John.

They are actually refusing to supply access to the source code on already GPL'd software? That'll get them in trouble. And it's one more reason to forget Lindows.

Bert you are right.I really fail to follow this idea a 100%.The way I look at it people already do work for companies that want customized solutions - and of course get paid for this.So what would be the benefit in having this made like in the model on the website?The way I understand it is that basically a interested individual or company or pool or a combination of these would pay somebody to customize programs for them.As this is already being done - where's the point?

Customising another application is done already, yes. But it's expensive - someone from within that community (however small - could be a department) has to have written the software first. They can't get that sotware from their rival firm because they probably didn't release the source code.

The point is that hybrid solutions have a good future - most of the problems I've seen in development is that the solution's infrastructure (for example a databases' relationship model) is flawed from the beginning. So to meet the original specification, the solution has to bolt on a compromise and roll it out to the customer.

With OSS, the groundwork is there, and already coded. The relationship model works for the generic requirements. An example is a CRM system (or a bug tracking database app). The functionality is the same whoever wants to use it. But it will need tweaking to fit individual company specs.

Companies will pay for the generic CRM app model because it is the best of all worlds - no need to worry about the underlying model as it's been beta tested*. And they get the source code to hand over to their developers - it cuts out those expensive architecture consultants (who never usually do a good job anyway).
That's my take on why this is a good idea.

* Actually the ransom model does not sit well for the "thousand eyes" testing and debugging principle. The ransom released software would have to be beta IMO.

I think this topics recently been picked up at Slashdot.org so they will have better ideas.

Cheers
Bert

Most of the companies I worked at are extremely paranoid when it comes to any technical 'stuff'.There might be a niche for this but usually (in my experience - which is not IT related) they rather pay for their add-ons than let anybody else even look at it.
Hell - one place that works with NASA and defense projects made me sign an agreement that the FBI or CIA or whoever it takes has the right to track me down and finish me off like the rat I am if I ever squeal.

Yes, I have little doubt they might hunt whistle-blowers like dogs - the old favourite - "That security app we built for <insert big-brother organisation> is full of bugs, ha! ha!" would really get 'em going!

My last organisation asked us to write a large content management system (using .NET and VS.NET - eugh! ) and then told us they didn't need us any more when we'd released it! We're waiting for the call in about a month from them saying "We don't know how this or that works ... "

Hehe.

I am sure that in this particular case they actually would.They make aviation parts for anything high-tech in the US defense industry the can jump up higher than 1 meter - thats flying.
Anyway - I got a lot of customers that do just the usual plain everyday crap that won't even allow a rep of a supplier in their place.And you are right - I was working on it but rat was the best I came up with - DOG - damn my english is getting real rusty.

I could say something very un-American (not that I'm American) about suppliers to the military being the least of their worries at the moment.
You can say rat too! Your English is fine ...
:D

Now might be a good time for one especially good distro to consider this financing model.

I really hope it will be a good Christmas for them ...

Bert

If RedHat were to release like this then they won't get any _more_ of my money. Neither wil SuSE.

Granted after a quick look at the site I don't think they're talking about rolling distributions (which are the collective works of hundreds? thousands? of developers). That unto itself opens up a giant can of worms. Sand worms. Like the ones in 'Dune'.

It's an interesting idea. Too bad it won't work. Open source code is more than the work of one person. I doubt there would be anything remotely resembling Linux if M. Torvalds decided to halt his kernel source publication via anonymous FTP because he was underpaid. That project would have been stillborn and wouldn't even be indexed on any search engine in that bleak future. Hell - there wouldn't be an Internet - Just a few SCO boxes hidden away in the basement (for nostalgia) and a gazillion MSN subscribers.

Ransom is what Redmond and its sycophantic patsys are having us live under in the present day. The upgrade mill has been relentless. It is a simple model as well - "ALL YUOR DATA ARE BELONG TO US". You bought Office 4.2 for your entire organization? Great! Then when every company you deal with buys Office95 you're going to have to upgrade. (lather, rinse, repeat for Office97, 2000, XP) The method is simple - make it as difficult as possible for your customers to change to the competitors. Lock in their most precious asset - their data.

Even the mere _threat_ of a ransom would have me running away from any package licensed as such.

BEEN THERE. DONE THAT.

Well, that's a point of view certainly.

I think you're confusing the 'ransom disclosure model' and 'holding your customers to ransom with version churn'. And it's an idea only - it's hosted on a domain http://www.theoretic.com/?Ransom - does that look like it's going to jump up kill your firstborn tomorrow?

So, Mandrake goes to the wall - everyone liked it but, when it came down to it, none of it's users wanted to pay because ... well ... there's always redhat, right?

That might not be what you meant but it's certainly what you're saying. Becuase they have finance problems (again).
There is always a getout clause for aompanies that wish to allow individuals to download - "free for non-commercial use" - this can be put on an FTP server too! Companies threaten themselves when they don't adhere to licensing laws. Individuals are rarely worth holding liable. And the ransom model targets businesses which are using the platform and distros as a base for hybrid solutions.

Then again, your rejection of such ideas would seem to indicate you'll be putting your .. erm ... money where your mouse is and donating to Mandrake to keep them going ... can that be done through Affero ...? ;)

Happy Christmas
Bert

Actaully, Mandrake had some interesting figures just posted. They earned about -1.500 KEuros on Income of 5,000+KEuros for 2002 so far. This is a DECREASE in annual losses of a very large amount versus 2001. Essentially, what they have is not a pure long-term problem, they have a cash crunch due to the economy and investor jitters caused by that economic slump. At this point they have investors waiting for their funding to come through so they can fulfill about 5,000KEuros in investment promises made to Mandrake. They need roughly 4,000KEuros to see the next version out as a breakeven thing. I will be contributing some funds of my own free will toward this.
If MandrakeClub is considered a Ransom by some, perhaps THOSE folks do not recognize that ALL software devs need to eat.

The idea of Ransom or prerelease to those who pay with free release following for most work done by a publisher's internal employees is one that has been followed for decades-- shareware was one vector for this, those who paid got free support and also in many cases free upgrades.

The software that got supported let the dev folks afford to buy hardware and test software on it and also invest in operating systems to test on and port to. The software that did not get supported got its code sold to commercial software publishers or simply vanished.
PC File started this way, folks supported it some but not enough, and it got sold. Peter Norton's Optune started as shareware but is now known as Norton Speedisk and is commercial. BRU started as something that was pure open source and now is a multi-version software that is one of the best backup softwares available for Linux. This is economic reality-- just as authors and artistws need recompense for their efforts to continue, so also do systems analysts, systems builders, software developers and publishers, and those who integrate software and driver and sets with the Linux base to get a cohesive and better and easier to get to function whole package. Linus gets speaker fees and funds for writing and is an officer for a software publisher as well as being a developer.

Mandrake does some of this with the club and actively seeking investors and stock holders and selling tuned OEM software that works with its distro. They use partner offers as draws, but in fact if more folsk do not contribute in some monetary form then the expensive bandwidth expenses and the cost of developing web sites will need to be absorbed by not releasing proprietory code, ALA RedHat's way of charging large amounts for support and for commercial use fo teh software at least. They do not want to do this, but they need to break even in 2003.

John Danielson.

They are being very responsive to user's needs and I for one will be actively supporting them here in the US.