Open Source economics
Posted by Ian Holsman
TED recently released the talk Yochai Benkler did in 2005.where he talks about crowdsourcing
both the ASF and DMOZ gets a mention
Open Source -- who is getting the money?
Posted by Ian Holsman
I just noticed 2 interesting papers about open source and web-business coming out of harvard.
and
The Value of a ‘Free’ Customer
The first paper investigates where (and why) large IT organizations invest in open source. the 2nd paper is about “free services” and how they are used to generate network effects.
I find the first paper the more interesting for me, as It isn’t really a ‘discussed’ topic, alot of research has gone into why individuals contribute, but not as much about why corporations do. Most see these donations as benevolent, but they are a business decision.
I’ve included the 2nd, as it presents another way to value companies as it calculates Customer Lifetime Value in a way that takes into account the indirect network effect the customer brings.
too much reading for me.. I’ve racked up about 800 pages in 2 days.
Is your project a Hydra?
Posted by Ian Holsman
In a recent post to one of the private member lists inside of the Apache Software Foundation, we were told that one of a project’s major contributors (of the project, not the ASF itself.. don’t get alarmed) was pulling out, due to the company he is currently working for switching standards.
The poster was a bit glum, about it, but as one of the other members mentioned:Fortunately we have strong support from many sides.This is the key to open source. It isn’t about a individual or a particular company.
When reviewing your open source project for possible use in your organisation, you need to ask:
Is this project a Hydra?
If one of the heads gets chopped off, will 3 others grow to replace it?
If not, avoid it.
This is one of the key reasons for the incubator in Apache, and what makes Apache projects what they are.
Don’t get me wrong, there are other factors you need to consider, (the Hydra was vanquished after all), but I believe this is what seperates True Open source projects, and open source ‘marketing’ wannabe’s.
Openlogic Piece rates for IT? no thanks!
Posted by Ian Holsman
OpenLogic has recently announced that they will ‘pay’ open source people to work on issues for them.
I find it quite disturbing, and not sure if they can attract professionals to it.
why? by reading through the FAQ I noticed the following.
- They pay per incident ($US100). for US/European/Australian workers that means you need to complete it in 1-2 hours (max) for it to be profitable to you, as you could be working on another regular contract for a better hourly rate. Remember this is BEFORE tax.
- They do the easy stuff. According to their FAQ they have already sifted through the tier1/2 issues leaving only the curly ones for you.
- Tax: I’m not sure if they have thought this through or not, but it becomes your problem. For international people this can be a big can of worms.
- Local (non US) banks charge for conversion of $US cheques. my local bank charges a bad conversion rate, a $10 processing fee, and has a 6-8 week wait before you can touch the money. This can eat into your $US100 fee, so unless you expect to get several support calls in the month this cuts into your money quite a bit.
- Ownership: (I actually like this one). they assign joint copyright/IP to themselves, and allow you to put your fixes back into the open source pool.
- Confidentiality: you can’t disclose who you are doing the work for, but they get to find out who is working for them.
- People will compete for the ‘easy’ fixable projects which can be completed in a small amount of time. The large/complex issues will be left unanswered as the oppurtunity cost to solve them is too small. (that and if you try and fix it, but can’t you don’t get paid. There is no second place)
- People will approach openlogic’s customers directly and try to get a more favourable rate for both parties by removing openlogic from the equation.
- The ‘experts’ will avoid it as the money isn’t good enough for the effort which will need to be expended, leaving the in-experienced people to try and answer the issues.
If you are a potential customer of openlogic’s and are reading this. Remember. you get what you pay for.
You should go to a support provider who have invested in their staff, and who actively participate in the projects. That is how you ensure sustaintability in the project, and access to fix your problems when they arise. Support providers provide a ‘throat to choke’ which you are not getting from piece-rate providers.
In Singapore!
Posted by Ian Holsman
Well.. I finally left @1:30AM (9 hours after it was scheduled) and had a uneventful flight (yay).
I arrived with 2 hours to spare. I finally got to meet Greg Stein, which was cool.. It’s good to put a name to a face after 6 years.
The conference is going well.
One thing which kind of surprised me was where the people attending the conference were at with open source. Hopefully the conference will help organisations understand open source, and how it can benefit their corporations.
I skipped the afternoon section, and am going shopping and then will try to get some sleep in.
Off to Singapore / Geek Dinner in Singapore?
Posted by Ian Holsman
I’ll be in Singapore for the next 2-3 nights doing a presentation at Open Source: The Business Case, Risks, and Oppurtunities on open source adoption in mainstream IT efforts (whatever that means ;-)
If you are in Singapore drop me a line and we can possibly do a geek dinner or something ;-)
Coal Mining
Posted by Ian Holsman
Marc Fleury’s last posting posits that the IBM’s strategy of rebranding a open source product is bad for the customer.
I’m not so sure.
and I’d like to discuss some of the points he raises.
A perfect example is IBM WebSphere Community Edition. It is based on the Apache Geronimo open source project and is “Bluewashed” whereby Geronimo is strip-mined for whatever is useful, then combined with other ingredients ….
First, I come from the position that there are different consumers out there with different needs. Some are quite happy with the levels of features and support they get from a open source project. Others want the open source project, and a throat to choke, while others need more ‘enterprise’ features and integration with their other systems.
The BSD model (and therefore the Apache Model) can serve the first 2 needs fantastically.
we enable a marketplace where multiple vendors can compete for service contracts, and by doing so you can get a competitive offering. With a single vendors (open source or commercial) product, you are stuck with whatever support they offer, and because all the leading developers work for one company you are S.O.L. going to a competitor.. they wouldn’t have the level of expertise required, or be able to contribute these changes back to the main source base.
Which leaves the 3rd need, the company whoose needs don’t match what the product offers, be it scalability, performance, or integration. IBM (and anyone else who wants to) has offered them a upgrade path to their enterprise product range. Marc calls this ‘Bluewashing’ I call it a upgrade path. It makes both the open source product, and the enterprise product more valueable to a customer. They can start with the free, or support only option of the open source product, and when their company expands, or hits the wall they can upgrade to a commerical offering with minimal disruption and effort.
What Marc also neglected to write is that it also provides funding and patches back to the community. IBM (and other support vendors) have a vested interest in having people choose that product in the first place, whats the point of having a useless open source offering if your entire strategy is to get people to upgrade from it to your commercial version. If it isn’t any good then people won’t bother.
Now does this mean crippleware? well.. yes for the enterprise customers with extreme needs it might, but there is nothing stopping those enterprise customers developing the feature themselves or banding together and contributing the feature back. (they won’t because a IBM offers much more in their value proposition than a feature list).
Because most Apache groups are not comprised of a single vendor they could get the feature in if it is useful enough.
It enables this “stealing” from the community, or in marketing-speak “leveraging”.
By stealing I belive that Marc is talking about firms or people not giving back their resources or money back to the community. but IBM does give back, as does a lot of other vendors who on-sell the Apache products. I belive the correct way to talk about this ‘market speak’ wise is called ROIC. Return on Invested capital. IBM and others have figured out that by having staff contributing back to the product is better for them than doing it by themselves. Personally I think ‘stealing’ in this sense are the people who use the product and don’t contribute their changes back, the people or corporates who just download it and use it, and keep their fixes/patches to themselves.
as for ‘waste dumping’:Another commercial software “strategy” is proprietary software “Waste Dumping”. Think BEA with Beehive. This is the opposite of strip-mining in that it entails the commercial vendors “donating” some technology to open source. Eager to be part of the open source wave, the vendor identifies some technology that is inferior or of limited value to them, and they dump (oops…sorry…they “donate”) it into open source.
one person’s garbage is another persons banquet.
How not to ask for help on a mailing list
Posted by Ian Holsman
I just saw this post on a list I’m on, and couldn’t resist commenting on it..
> I see the last post to this list was November of last year. Is this project still active?
yes.. very active.. as you can see by the traffic.
> I have a client who has XXXX installed and would like …
translated: someone is paying me money, so I need some poor smuckh^hh^hopen source developer to spend an hour or two of their time so I can look good/charge my client a bit
> If so would it be possible for someone to cut a new release?
translated: I don’t even have the faintest clue on the basic tools of the open source world (cvs/make/configure) and/or think getting a nightly build is unstable, while having someone package it up as a tarball and slap a release number on it makes it better.
and the best for last
> I recall that the [last] release was barely usable so if this project is still in the same state please let me know.
yep.. insult the guys who work on the project.. that’ll work!
ok.. I know I’m ranting a bit here, and this person has just not got a clue, but why do people still equate open source products with IT-geeks just salivating at the thought of doing unpaid work for some stranger?
so.. to step off a pedastal for a bit, how should have he moved forward with this?
- Frame the question in a direct manner, feature X is failing when I do Y. i’ve tried Foo, and bar but they don’t work.
- Ask of the developers that works for his company to actually make the project do what he wanted to (and hopefully submit that patch back)
- Looked at the companies around who claim to support the product, and hired one of them to do the work for him.
- Actually read some of the other mailing lists around the product and noticed that XXX is dead because the core product now supports something similar to what it was doing.
update: ben, the project maintainer replied…
>CVS is more recent than the newest release. But no, I would say the project isn’t really active.
> It’s got some known issues but is working well enough for me that I’ve gone on to more pressing problems in my life. :)
> If you have a particular problem with what’s in CVS, what isn’t working? Maybe I can point you in the right direction. Or….. feel free to fix it. ;)
> Plus, I told XXX I’d take a look at seeing what it would take for me to enhance his own XXXX to offer the same features as this one. I’ll get around to that one of these months….
nicely put.
Do People upgrade?
Posted by Ian Holsman
A lot of opensource companies are based on a business model where they release a ‘open’ version which has the basic functionality, and a ‘enterprise’ version that people pay for.
I’m just wondering how many people actually make the switch. Do people actually start using the open one, only to find it doesn’t meet their needs in 3-6 months time? or do they just use the open one for a pilot project to make sure it isn’t a dud before they actually shell out the dollars for the enterprise one.
If they are only using it for a pilot, wouldn’t it make more sense to just offer a time-limited version of the enterprise one?
and if this is the case, why do large companies with ‘enterprise’ software even think of buying/supporting the smaller open source product in the hopes their users will pay $$$ at a later time.
My basic thought is that users of open source ‘basic’ products have already shown their price sensitivity and have zero interest upgrading. The only way to capture these users is to lock them into your ‘basic’ product, and remove it from the market at a later stage forcing them to upgrade then.
This is what I dislike about this business model the most, in order to sell the enterprise version of the software, you need to control the basic one.
And one method of controlling it is to not let the design, direction, or the IP of the product be ‘opensourced’, making it so only your companies people are responsible.
For me, a true ‘opensource’ product is where multiple vendors co-operate together to build a full-featured product.
Is there money to be made? sure.. in the complements. Tools, Support, Training, Books, and Consulting.
Communities and open source
Posted by Ian Holsman
In Matt Asay’s blog he mentions how he shared a flight with Clint Oram of SugarCRM, where clint quotes..
Of the top 50 contributors to SugarCRM’s development community, 95% of them are SugarCRM partners
And goes on to say that this is a good thing.
Well.. count me in the ignorant pile, because I’m not so sure. For me the signs of a vibrant community is more that the USERS of the product are contributing patches back to it, not the people getting paid to install it.
This might be because we have a different definition of what a community is, or maybe the client base of sugarcrm aren’t technically sophisiticated enough to contribute back to it..
why is this important?
Well there are two major benefits I’ve heard that a ‘open source’ product has.
The improved sales:revenue ratio, with JBoss’s Marc Fluery saying recently :
An optimally functioning FOSS business model needs 20 cents of sales and marketing to acquire 1 dollar of maintenance, where a traditional software company will have to spend around 2 1/2 dollars.
and a decreased R&D:sales ratio (can’t find a quote..sorry). SugarCRM needs to find a way to get it’s users to contribute back into the development. only then will it get the full benefit of being a open source company.
There are 2 other important reasons why you want the users to contribute back as well. Product Innovation, and ownership. the Users are the ones who know what their business needs, not the ‘partners’ who install it, they don’t use the product day in and day out.
The second reason is ownership, if you have put your own sweat and tears into making something work, it will be much harder to give it up for some other product later on. If you can get your users to feel that they play a bigger part than the person signing the check they won’t leave as easily.
of course.. this is all a moot point if they call their ‘users’ ‘partners’ isn’t it ;-)