Ryan has posted 17 posts at DZone. You can read more from them at their website. View Full User Profile

Sun Microsystems Charity Poll

03.27.2009
| 5688 views |
  • submit to reddit

Tonight I ran across an online poll asking if businesses would be willing to donate $1000 per developer every year to Sun Microsystems, and if independent developers would be willing to donate $100 - $1000 per year as well. Some companies survive on donations, like Wikipedia's annual campaign to raise $6 million. Sun is a different kind of company and I think they would get less business if people thought they needed to rely on charity to survive.

Instead of donating to Sun, why not buy their products and services? For example, one service that not many developers know about is Sun Developer Expert Assistance. For only $249 per developer you get one year of developer support (not production support) for one product stack such as Java ME, Java EE, Java SE, GlassFish, etc. For only $549 per developer you get one year of developer support for all Sun products and technologies. Sun engineers can look at your source code to give advice on best practices, provide sanity checks, show you how to use an API, etc. I've used this service before and was thrilled with the level of support I received. At one point several engineers VNC'd into my computer for two hours to help me understand their product and to diagnose a possible bug in the installer. This kind of support is affordable and invaluable to independent and corporate developers. The public mailing lists are often helpful, but when you need an answer right away then paid support is the way to go.

The other day I read some insight into the real reason Sun might be sold to IBM:

If the reports are accurate and Sun has decided to sell, the decision was probably driven by outside investors, notably Southeastern Asset Management, which increased its stake in Sun to more than 20 percent last year and has been pushing hard for a bigger return on its investment.

"I think those guys are driving the bus at Sun," Olds said. "This isn't a strategic thing or a Jonathan Schwartz thing; it's purely business."

From http://www.ryandelaplante.com/

Published at DZone with permission of its author, Ryan Developer.

(Note: Opinions expressed in this article and its replies are the opinions of their respective authors and not those of DZone, Inc.)

Comments

Dominique De Vito replied on Fri, 2009/03/27 - 3:01am

IMHO, more open source projects could be run with donation. Small payments could make a river and may push forward these projects. And if donations might give rights to vote for a roadmap, it may introduce more community involvement for these projects. This being said, thanks for mentionning "Sun Developer Expert Assistance". Indeed, it is quite interesting given the small price in face of the support I may get.

Jeroen Wenting replied on Fri, 2009/03/27 - 3:32am

So once again it's the hedgefunds and other get-rich-quick-over-someone-elses-back schemes that destroy basically sound companies and lots of peoples' lives for a quick win. And those guys still claim they only act in the best interest of the companies they're "investing" in and the people that work there?
I'm not in favour of massive government regulation of private enterprise, but IMO the influence of investors in corporate decision making should be massively curtailed, especially that of investors with short term portfolios.
Maybe introduce a long waiting time before a shareholder gets a voice (let alone a vote) in shareholder meetings, say 10 years. There are already limits on the minimum percentage of shares you should hold to have a voice, so that wouldn't be a big change.

Álvaro Martínez replied on Fri, 2009/03/27 - 3:50am

Sure! Millions of people dying in Congo, for example, and I'm supposed to donate money to a corporation. How do you like that?

 Jesus, what a rotten world.

David Gilbert replied on Fri, 2009/03/27 - 4:45am in response to: Dominique De Vito

IMHO, more open source projects could be run with donation. Small payments could make a river and may push forward these projects. And if donations might give rights to vote for a roadmap, it may introduce more community involvement for these projects.

You're being naive. But don't take my word for it - try it yourself. Go create an open source project, put a big "DONATE NOW!" button on the front page of your web-site, then after a few months come back and let us know how you got on.

Funding is one of the big problems for open source software. Begging for donations is not the solution.

Dominique De Vito replied on Fri, 2009/03/27 - 6:08am in response to: David Gilbert

You're being naive. But don't take my word for it - try it yourself. Go create an open source project, put a big "DONATE NOW!" button on the front page of your web-site, then after a few months come back and let us know how you got on.

Funding is one of the big problems for open source software. Begging for donations is not the solution.

 

Well, you don't see my point.

I agree people are reluctant to donate for an open source project. But I think it's partly due to the fact that they don't want to donate for nothing, that is, without knowing the target.

First, I have seen different open source projects receiving company funding/support for developping such of such feature the company wants. This is close to the way I envision donation (see below).

Secondly, I think there are situations where an open source project may call for donation. Imagine a project listing all possible items in the roadmap and saying that some items will be done and others will depend on funding. And let's imagine also that the team is able to give the development cost of each item.

Then, in that case, such a open source project may call for donation if people are interested for such or such feature. I have listed here some Firefox micro-features I am interested in. If Mozilla would propose a call for funding for some small enhancements some users want, then I would give the amount of money I want to pay for the requested micro-features. But, remember, I think that if an open souce project wants to follow this path, it has to do it well :

(1) evaluating development cost,

(2) publish them on some web sites,

(3) call for donation candidates,and make advertisement about that call

(4) make different rounds in order to select/vote for possible features (to develop+fund during the time duration we have) while trying to match the donation for such or such feature.

(5) and then, develop

(6) give community feed-back about that experience and so, on.

So, you see, the way I envision donation is NOT like just introducing a button "donate" on a web site! My way is more complete approach, a community-oriented approach while interacting better with the community.

Note that, following this way, developers have to write first the documentation, that is, the specifications-like of the features the developers want to call funding for. So, indeed, there are different donation approaches and choosing one or another could change the development process too.

 

I have learnt recently "Lightning" won't be part of Thunderbirdv3 while I have read Lightning-in-Thunderbird status update. I was not happy with such a choice. So, I have proposed "my donation way" as a comment of this post, without success AFAIK. I have thought Mozilla was strong enough, confident enough, to support such a new process for donation, at least, due to the Thunderbird audience and the number of people wanting a calendar inside TB; it would have been a perfect case for testing a donation try. Unfortunately, it is not in the air.

 

This being said, donation does not apply for all cases, for sure. For example, one can't call for donation while just starting an open source project. Quite an amount of code has to be produced in order to prove the project is going somewhere and to let users confident about that direction and developer capabilities.

 

David Gilbert replied on Fri, 2009/03/27 - 6:30am in response to: Dominique De Vito

First, I have seen different open source projects receiving company funding/support for developping such of such feature the company wants. This is close to the way I envision donation (see below).

That's not a "donation", it is a company paying for services. And if you can convince multiple companies to share the cost of that service, they're still paying for a service - not donating. Maybe we just don't share the same definition of donation.

That said, your idea of specifying a feature enhancement and inviting companies to part-fund the development has some merit for an open source project. I may even try it sometime.

Guido Amabili replied on Fri, 2009/03/27 - 8:35am in response to: Álvaro Martínez

Do not think Money could make our world a better place....

Bob Smith replied on Fri, 2009/03/27 - 9:17am

There are corprorations who have a vested interest in keeping Sun independent.

For instance, any corporate user of Netbeans probably doesn't want IBM to take over Sun.   The Glassfish community (which I'm sure includes some corporate contributors) probably doesn't want IBM to take over Sun.   Any company that relies on OpenSolaris probably doesn't want IBM to take over Sun.   And I can almost guarantee than any corporate customer running SPARC most definitely does not want IBM to take over Sun.

These are the people who should be donating to Sun, not developers like us, many of whom are worried about losing our jobs in this recession.

Maurizio Turatti replied on Fri, 2009/03/27 - 9:45am

Opensource is a way to sell services, not products. If one thinks people is going to pay or donate for an opensource product, then he's going to be pissed off. Services can be in terms of training, support, consulting. Alternatively you can package closed-source add-ons on top of opensource (MySQL, for example, sells an Enterprise Monitor, which is not opensource). The idea is that you have a lot of people looking at your code and posting patches, so decreasing development costs. You can sell added-value services on top of it and you receive a lot of feedback from real users, like you are into a constant feasibility analysis phase which always drive the product where the market really needs.

Sun's model is interesting but difficult because they rely on support only to get a revenue, they are not a consulting company, so much of the real money of their open-source is going into somebody else pockets now. Usually support is a cash-cow, but an active opensource community can provide far better support that any single company, so the only viable solution is to sell added-value services on top of the open-source software. Sun's present model would require an incredibly huge users base to provide good revenues. IBM, for example, could provide a much wider set of services on top of Sun software, that's why I guess they could make a better deal of it: diversification.

On the other hand, at least in Europe, Sun has not really tried hard to sell their opensource products, mainly because they do not have the right people for the job. They instead gave the idea that opensource software was a give-away present to sell more hardware, which could be a viable solution if only in the menawhile the server market was not crashing. But the demand for Sun software exists, just many customers are a bit confused by what they hear from some Sun sales. As I said, Sun software is not an opensource solution with optional commercial support, instead it is a commercial solution with the added benefit of being totally open. But then you need to go out and sell it, as enterprise software is not exactly like an iron box that you plug into a socket and leave...

Dominique De Vito replied on Fri, 2009/03/27 - 9:51am in response to: David Gilbert

Maybe we just don't share the same definition of donation.

Thanks for this sentence. It clarifies things.

Indeed, my definition of donation is quite between feature-oriented donation and funding ('funding" is somewhat the right term but "feature-oriented donation" is interesting too while both defining a goal and giving some freedom to developers).

While thinking about this "concept", I have been inspired by the ones I have seen here:  http://www.datanucleus.org/project/donors.html

 

This being said, I don't know why there are many "donate" buttons out there on open source web sites, and not so many button "fund". It could be worthwhile to create more of these last buttons because funding is also another popular win-win option if both sides agree.

For example, I don't want to donate for OpenOffice, because I don't want OpenOffice to spend it on advanced features while some very useful basic features are not there (OpenOffice lacks a little to take into account the user experience). But if OpenOffice enables funding-oriented developments, changes its mind in order to ease such kind of developments, I am OK for funding, at my level, the development of micro-features I would like (read about them here), or the development of any feature set that goes in a similar direction.

 

That said, your idea of specifying a feature enhancement and inviting companies to part-fund the development has some merit for an open source project. I may even try it sometime.

While thinking about the development process I envision for feature-oriented donation, in order to emphasize this process, I have thought about the term "Donation/Funding Driven Development". While the TDD approach moves the Test as a first step, the FDD moves Funding in a similar way. So, I like the FDD term.

As you may know, there are some soccer/football teams out there, that try to follow nowadays the FDD approach. You pay some fee, and then, you get some rights in order to vote for training/strategy/game choices. The subscribers get also some privilegied access to trainings/videos/interviews. But there is still a team manager, who has some vote rights too and a limited amount of jokers for choices (in order to have the final word for some special occasions).

That's a similar approach I would like to see for some open source projects. It may work, for some kind of organizations.

First, such initiative needs collaboration infrastructure and computer scientists are the best placed in order to create such infrastructure.

Secondly, while getting more and more connected, we still have to imagine, to learn, and to try, new forms of organizations and IMHO, such organizations as above are worth trying in open source.

Dominique De Vito replied on Fri, 2009/03/27 - 10:14am in response to: Maurizio Turatti

Opensource is a way to sell services, not products. If one thinks people is going to pay or donate for an opensource product, then he's going to be pissed off.

For big companies, yes. But for small to medium open source projects/companies, IMHO (feature-oriented) donation, or funding, could make sense.

 

Sun's model is interesting but difficult because they rely on support only to get a revenue, they are not a consulting company, so much of the real money of their open-source is going into somebody else pockets now.

I think you can get the precise information from Jonathan's blog, but they are not doing that bad, and their open source business looks like growing if I remember correctly.

 

Dominique De Vito replied on Sat, 2009/03/28 - 6:34am

While this post and various commenters have been able me to express my opinion, and to make it better (thanks !), I have gathered all my ideas into a new post :

Revisiting donation/funding for open source projects - let's talk about FDD

I hope it will be helpful.

 

Alex(JAlexoid) ... replied on Sun, 2009/03/29 - 5:12pm

I bet a lot of consultants and developers could heard the money together and just buy out Sun.
I mean it would take only 1 million people with 6.5 k USD to buy Sun. Taking into account that a lot of developers actually depend on Sun and a lot of their income comes from developing with Java, that is no such a crazy idea.

Sun could even be a non profit organization. The returns to the shareholers would come only in form of assured future in their field of expertiese and having a say in that area. And sure enough, income from consulting opportunities would be in any case better than dividends, rise in share prices and possible time lost on spending time developing the Java platform.

Just imagine if IBM buys Sun:
 - IBM becomes Java
 - IBM can get anyone an Indian guy to replace you, and still have some left on the side
 - IBM will be able to offer support to anyone
 - IBM is not known to like competitors, in fact Microsoft took over from IBM with their "love" for the competition
Would that be a positive thing for a Java consultant/ freelance developer or a Java shop?
No, unless you are IBM's partner. And being partner to IBM is like being a corporate whore's dealer.
(Sure, it's a doomsday scenario, but still...)

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.