Absolutely Bruno, what one might consider a base-of-the-pyramid approach will probably garner more trust and credibility amongst the community than a top-down approach – idealist that I am I’d hope we could get both working at the same time to speed things up a bit

As an aside, it amused me to see you’re French and based in London and I am English and based in Paris Noting you like fractals, I love them as well, you might enjoy this link: http://www.martinlaprise.info/2010/02/11/visualize-your-own-github-graph/ but the reason I share it is twofold – #1 is because I think you’d like that kind of thing having just read your blog posts and #2 you might realise that this type of functionality does not yet exist for the business community in open source – there’s not yet sufficient transparency as they’re either locked up in a black book, Siebel, SalesForce, LinkedIn, Xing, Viadeo or a.n.other walled garden but if the open source community can drawn them out of those places and into a collaborative ecosystem we might find we have the opportunity to improve our chance of fixing bug #1. The old adage of “content is king” has passed, for us “adoption is king” and it is in all our interests to ensure that our sales & marketing people have the best possible tools to make this happen – and as you say we should not wait for Canonical or RedHat to begin this process – we should share our tools and knowledge amongst each other and begin the empirical process of improvement through crowd-sourced wisdom that the internet allows us to achieve. I vow to do my bit towards this endeavour.

I didn’t realise I had reduced your idea to a wiki page in my comment, it wasn’t intentional. Anyway, I think we’re all in agreement but have different approaches. In the Ubuntu context, I would expect Matt to look at the strategy of such a mesh and how to expand it beyond Ubuntu, as you say. That doesn’t prevent us, as interested parties, to work together on local tactical solutions to start off the mesh.

My idea, if you excuse me for harping on, is far from ‘having an area of the wiki’… it’s about a united effort across the entire opensource ecosystem to provide a superior commercial engagement model that rivals proprietary competitors in both capability and results.

Sure, it can begin on a wiki and foundations and roadmaps can be laid out for discussion and cultivation though I’m thinking more along the lines of event driven architectures that allow the partner ecosystem to be both agile and responsive to client requirements – having heard and read plenty about how corporate IT is not sufficiently aligned to business goals it does not surprise me that we have a significant challenge and opportunity in front of us to help corporate IT deliver on the promise of being business-centric and providing reliable services and innovative solutions to support these businesses.

Furthermore, the concept should be generic enough and prove to be best practice for open source solution providers that the co-opetitive principles and opportunities begin to emerge very quickly and clearly (co-branding, co-marketing, co-sponsoring, co-etc.); thus an open source business (whether that be open source project, open source VAR/ISV or even freelancer) benefits more from participating in the co-operative (this would eventually be a community of co-operatives just as the open source ecosystem is a community of software projects) than participating outside of this model because the value of unity is far greater and more effective than that of operating independently. How are we going to enable an effective and mature approach to co-bidding on massive projects? How are we going to identify the right partners who have the niche expertise at a rate our client can afford? How are we going to engage with potential clients and “look the part”. How do we evolve an equitable partner reputation model (some have an answer to this question already though I do not yet endorse its adoption in an open systems content, see http://www.alliantist.com for further thoughts in that direction). Many of us have plenty of experience and wisdom to share. Certainly among the biz-dev community that are opensaucing there are some real pros yet we need them to mentor the rest of the community and help polish up the business practices so open source business is excellent. I’m looking to people such as Matt to step up to that challenge as part of the process of aligning Canonical with it’s strategic goals. He should observe the success and challenges Jono Bacon faces as a community manager and bring on board someone to support existing sales/biz dev at Canonical and ideally with the aim of having a similar impact on the business partner “community” as Jono is having on the user / technical communities. It would be unfair and unrealistic to let Jono do it all though between Matt and Jono there’s a good chance they’ll find someone who fits that bill – and then, this is the key bit, this person/initiative needs to be replciated across the entire industry enabling us to loosely couple the business of open source (collaborative CRM)… and at this point we arrive at Alan’s mesh

@Alan
That’s a fine example to share though it is isolated to one vendor and its community, if we can think OO for a second and look at your example as an “abstract class” that we would inherit and extend with event handlers that enable behaviour amongst instances of the new class in a multi-threaded real-time architecture (multi-vendor, multi-partner, multi-client) then I think we’d have a very powerful tool to “raise all the boats”. Networking / doing business is not new at all but the rules changed when we got the internet (for some of us it started when we got our modems in ’82 ). We need to take heed of that and look to create new games, not play out old ones, though that last comment in no way is directed at your suggestion re: mesh, it’s more a general statement about the state of software business today and I’m looking at how traditional go-to-market methods are being disrupted and where we can take advantage of those trends to improve open source success and create a sustainable and successful software industry that we can all benefit from.

I’m with you 100%. But do we have to wait for Canonical to do anything about it? What about starting something like this within the UK LoCo team? It’s small enough that we should be able to make it work and that should give a good template to extend it later. On a related subject, I’ll be at the British Library on the 10th at a workshop on planning and developing your business so I’ll try to ask about the value of networks and how to go about making one work.

Alan, sorry it looks like we’ve hijacked your blog post. The points you raised seem to have struck a chord

I see where you’re coming from, but I don’t actually think that we’re at crossed purposes at all. ISVs in FOSS tend to offer integration services on existing software, often using custom web or native apps to handle configuration. See eBox for en example. Creating an easier road for new developers strengthens this area. I’m not sure you’re going to get “partners” the way MS has them since Canonical is pretty much just a service company already, but Canonical can make it easier for others to sell integration services.

Bruno,

You caught my point exactly about “making opinionated choices.” (The quote is from Quickly) I don’t think there’s anything wrong with Ubuntu creating the pages with a specific language, toolkit, IDE, and VCS in mind. I’d love to see it all set up OOTB via a special CD or meta-package, so that devs can get to work and publish. There are a million choices in Linux: those choices remain on Ubuntu. Looking at history, though, Ubuntu came out as a single-CD distro with one of each app at a time when multi-CD installs with five word processors and ten text editors was the norm, and Ubuntu was remarkably successful as a result. I don’t see why the same attitude can’t be taken with development in order to make the decisions easier for new devs.

1. Daeng’s idea

Having an area of the wiki dedicated to development, with tutorials and examples for people who are new to Ubuntu development would be great. However, right now, tools like Quickly, Ground Control, Lernid or Acire are still in the incubation stage so are not quite ready for prime time.

On the other hand, to do this properly, we need to take into account developers who come from other platforms. If you look at the development world today, you can roughly identify the following groups: Windows devs, Mac devs, Linux devs, Java devs. Windows and Mac devs work on the same model: you follow the standards issued by Redmond or Cupertino and you only use tools that have been certified by Redmond or Cupertino. Java devs are an interesting hybrid (I know I’m one of them): they are used to open source, because some of the best Java libraries and tools are open source, in particular the ones from the Apache foundation; they have a choice of tools to do their job, a lot of them are open source too; they code against open standards; etc. But there is one major aspect that they share with Windows and Mac devs and that is alien (or at least strange) to the Linux world: they use an IDE (Integrated Development Environment): in the morning, they come in, they fire up their IDE (MS Visual Studio, XCode, Eclipse, NetBeans, etc) and they do everything in it: update source code from the repository, create a new project, create and use code templates, compile code, run and debug it, package it, run test cases, fire up the coffee machine, etc. By contrast, on Linux, you end up having to use half a dozen different tools to do the same thing, unless you use Emacs. For someone coming from Windows, Mac or Java, the Linux dev environment looks very fragmented and difficult to understand.

2. Ed’s idea

Having an area of the wiki dedicated to ISVs and VARs would be great and would complement the developers’ area. This should be targeted at the sales departments of ISV and VAR companies.

One issue in particular with open source is how do you build a successful business around it? In the Microsoft world it’s a no brainer: an ISV will sell his own software solutions on top of MS solutions for some good money => profit; a VAR will sell MS solutions, with services on top for money => profit. Customers will not hesitate to pay for the ISV’s software or the VAR’s services because they are already making an investment in MS software to start with and the extra services on top are meant to reduce the risk associated with that investment. How does that work in an open source world where the original software is open source and therefore free to use?

Up to now, I have seen two successful models for companies who work with open source. #1 are the ones that have a product, for which there is an open source “community” edition but who sell an enterprise edition for a lot of cash, as well as services behind and have a support model that is such that, for most companies, there is less risk in buying the enterprise edition which includes bundled support than using the open source edition and risk spending gazillions on support costs. #2 are the ones who sell support and consultancy services on a well known open source product: they don’t have to convince the customer to get the product in the first place, most of the time they get a call because the customer is using an open source product and realises they need help with it.

Sorry for this very long comment. It started simple enough and then it got out of hand