All posts by Yannick Buron

Clouder version 0.9.0 and Odoo Experience

Hello everyone,

Last week we released the version 0.9.0 of Clouder, which bring a lot of improvements especially in the deployment process. Two softwares are added in the Clouder infrastructure : Salt and Gitlab.

Salt is an orchestrator, like Puppet/Chef/Ansible. The idea behind integrating an orchestrator is to delegate most of the configuration work to a tool which is really design for it, has tons of plugins, a great community, so it can replace the ssh calls we are doing now. We don’t have real use case yet which Salt can manage whereas SSH can’t, but we just feel this is the right direction to go.

Also, why choosing Salt over Ansible for example, the reason is mainly because it’s written in Python like Odoo (and so Clouder), which mean we can easily integrate it in the Clouder code without relying on SSH calls anymore. Moreover, thanks to the fact that commands are executed by Salt minion and not the master, a Salt infrastructure is incredibly scalable.

Next come Gitlab. No need to present it, it’s obvious that having a SCM in Clouder to manage the collaboration around projects is a must have, and Gitlab is probably the best open-source project for this today.

But what’s really the gamechanger here is Gitlab CI. Gitlab is now able to manage deployment pipeline, and manage for example the following process :

  • After commit, build a new Docker image with the last source code from your/Odoo/OCA repos
  • Push your image in the private Docker registry
  • Execute OCA tests in a docker container
  • If tests are green, deploy it in a staging environment
  • After approval, deploy in production environment and update Odoo bases.

And the pipeline itself is fully customizable through the .gitlab-ci.yml file, while the Dockerfile allow you to specify the Odoo version (8/9/10/OCB) and community modules you need, directly in the repository of your project.

This means Clouder is now capable to manage a complete Continuous Integration / Continuous Delivery process, this is a huge step forward. This is also why we needed to integrate Salt, we didn’t want Gitlab to call Odoo for the deployment of the new image, Odoo is not designed for this kind of usage so we needed Salt to be the real heart of the infrastructure. Right now this process only work for Odoo template, but now that we have one we can easily extend it to other applications (WordPress, Drupal, etc…)

Another feature, now some containers are able to auto update. This is especially important for exec container which have their ports exposed so it’s important to update them regularly to avoid security issues. Each day, theses containers check if a new image are available and download/redeploy if necessary.

Same for the files containers, if you are not using the CD process for an instance you now have a button to manually download the new version, deploy it and update the bases linked to it.

CI/CD, integrate an orchestrator, auto-update,  there is still one thing that is missing in Clouder if we want to have a fully scalable infrastructure : A tool to manage a Docker cluster. This is especially important to provision new nodes, create a private network between the containers to ensure security, make the instance scalable by duplicating exec containers etc…
Theses needs are usually done through tools like Kubernetes/Openshift/Rancher etc… But with the announcements with Docker 1.12, I feel like the future in clustering will be more and more ensured by tools edited by Docker Inc. so I think the core of Clouder will essentially support Docker Swarm which became a real beast with Docker 1.12 and manage all the features I mentioned. This will probably be our next move.

Also, as you probably see we changed the numbers of the Clouder versions, which are now 0.8.0, 0.8.1, 0.9.0 etc… We change this because we don’t want to stick to the Odoo version anymore, Clouder work with Odoo 8, Odoo 9 and will probably also work out-of-the-box with Odoo 10, the only requirement is the new api. Before the release of 0.9.0 some people thoughts that Clouder was only compatible with Odoo 8, this was too much confusing and we needed to clear that.

Finally, last but not least, I’ll give a presentation of Clouder during Odoo Experience, Thursday at 11h45 ( I’ll concentrate on the demonstration so it’s the occasion for you to see how it works, I’m publishing the slides below now because I don’t think I’ll not really be able to show them during the presentation :

I’ll be there during the whole week, including the OCA code sprint, so if you want to meet just send me an email.

I am really excited by the event this year and I can’t wait to see what’ll happen. See you there !

Preview of the Clouder hosting offer

Hello world,

It’s time to unveil the details of the main Clouder offer. Thank you so much for being so many to answer the market study, we worked hard since then to prepare a concept which is I think long awaited by everyone.

As you know, our motto is to give you peace of mind, we shall take care of everything and just let you focus on the relation you have with your customers. This include taking care of the infrastructure itself and this is why our offer is built around it.

When you will create your account on, you will have the possibility to buy nodes (servers) where you will be able to deploy your containers and your bases. Prices depends of the resources (CPU, RAM, Disk space) assigned to your nodes.
You will then be able to select your nodes on the Clouder interface and manage your infrastructure from there. This service includes :

  • No limit on the number of containers and bases you can deploy on a node. This is especially interesting for multi-bases deployment since each base consume very few resources.
  • Each node is hosted on a high availability platform, managed on the hardware side (So Clouder is not involved here). This means that if a server crash or there is a problem on the network, the service will still be ensured by another duplicated server with no downtime or loss of data.
  • The support around the use of the Clouder interface.
  • The possibility of creating your own Odoo verticalization through the Clouder applications. This means that you can pre-configure your Odoo containers so they include the modules you defined (either official, OCA or custom modules) and select which modules will be installed when the base is created.
  • And last but not least, you will be able to add a subscription form on your own website. Build entirely on javascript, you only have to add a tag on your website to display it, and this will allow your visitors to get their instance of your verticalization in minutes just by submitting the form!

By creating this offer, we want to address a need which has been around for a very long time in the Odoo community: Being able to easily create an offer based on modules you set up in order to create a customized experience for your target market, without having to have high skills or time to invest on system engineering.
And by choosing a business model based on resources, this will give you the flexibility to decide of your own pricing for your customers, based on the usual price per containers/bases/users.

This is it! Now in order to start this offer, we need to reach a critical mass of customers so I invite you to manifest your interest in order to give you more precise informations and get things started.


On other topic, I have to say that the documentation is now vastly outdated. Despite the importance of this task, this is not in our priority in the next weeks so since it’s something everyone can contribute to, help will be strongly appreciated (I’m only saying that because that’s the first place people which are contacting me are proposing to contribute). Until then, note that a complete presentation I did was recorded by ITpedia (thanks guys!) and published on Youtube:

About the next steps, still for the Odoo integrators, I really want to make Clouder manage the full DevOps process of the Odoo development. So expect to see some new templates coming in the near future, including gitlab, travis, and integrate the scripts the OCA use for their test and CVS platform so you can have a similar experience for your own private business needs.

Finally, I understand that our team is difficult to contact right now, so a public chatroom was set in order for you to easily contact us: This place replace an IRC channel, you can come here for any functional or technical question as well as getting in touch with other members of the Clouder community.
As usual, all our work is under open-source licence and is available on Github :

Thank you, see you soon!

Preparing the Clouder offer

Hello world,

It’s finally time to go to the next step. We are preparing the details of the Clouder offer and I wanted to give you a little preview.

Our main target at first will clearly be the Odoo integrators. This is my feeling, but I think that most of them have a complicated relation with hosting right now. Their main values are their skills at development and the proximity they have with their customers.
They would like to provide hosting for their customers, but hosting require economy of scale to fund the development of the infrastructure and the human resources needed. Most of the integrators are small teams and not all of them want to resolve an emergency while they are at their brother wedding during the weekend.

On the other side, some customers want to have their ERP on their own infrastructure or the same hoster that take care of their website and is not an expert of the ERP hosting. In this case the infrastructure will rarely be at the state of the art, simply because it will only be tested by one company.

So of course, our goal with Clouder will be to provide to all of them the infrastructure and the support they need so they have the peace of mind they deserve, with an infrastructure tested by several dozens of customers and installable either on the cloud or their own servers.
Now the big question is what do we want to include in the Clouder offer. On the technical side this is what we currently have :

  • Deployment of Odoo. We manage both multi-containers and multi-base cases.
  • Each Odoo instance has several containers (data, files, and exec), for easier version management and scalability of the instance.
  • We monitor the instance downtime, and if at least one backup was done today
  • Backup each hour, with no large consumption of disk space thanks to the use of a git-based backup technology. You can restore a backup directly from the Clouder interface, either on the same instance or another one if you just want to consult data.
  • Proxy based on nginx, with configuration specific to Odoo redirection and https encryption either based on your own certificate or otherwise on LetsEncrypt.
  • Mail based on postfix, with support of SMTP relay, anti-spam, SPF and DKIM. The received mail are correctly sent to Odoo, so we support both in and out mails.
  • Asynchronous jobs on Clouder (thanks to the use of OCA connector module).
  • Link to the accounting module so you can easily invoice the instances of your customers.

Finally it’s designed to manage verticalizations. You can package an Odoo with your own specific modules (either yours, core, or OCA), and when you deploy a new instance it will git pull them and automatically install them on the bases created afterward. This is the real gamebreaker, thanks to this you can provide to your customers a package with a complete new user experience, an ERP adapted to the specificities of their industry right out of the box.

And this is only the beginning! We will not make other development before the launch of the offer but as soon as we can fund it we will add the next features (with no added price). To give you some insights here is what will follow :

  • Form on the CMS part of Odoo, so visitors can create their own instance without your interaction (already under development)
  • Management of the development circle, with a git repo container, CI with Travis, integration with project management, and deployment on production.
  • Add other software useful for companies, like webmail, dropbox-like document management system (seafile or owncloud), etc…
  • Deploy automatically complex infrastructure, like for example Magento + Odoo with OCA connector configured. Most of the work is already done.
  • Add other subcomponents, like LDAP, SSO to easily move between each deployed software, 2-factor authentication, piwik analytics etc…

Of course our plan is to provide support and live from it. We would like to propose the following services :

  • Your own Clouder instance (obviously it is itself managed by a Clouder).
  • Either you already know how to use Clouder, or we make for you the deployment of the infrastructure (with all components, registry, backup, monitoring, dns etc…).
  • Assistance for using the interface, if you have any questions etc…
  • We receive the monitoring alerts, if something happen on one of your instance or one of your customers we immediately take action to resolve the problem.

We want to provide with Clouder the best Odoo infrastructure on the market, before moving to the next software. And of course, as usual, all of this, every single part of your infrastructure, even what we use in the Clouder servers, is and will stay free software and available on our github repo.

This is what we have in mind, but now we need your help. In order to create the best possible offer we definitely need your feedback, and I sincerely invite you to give us more information about your needs through our contact page (https:/, or just to show us your interest so you can have an early access to the service.

As usual, thanks a lot for your support ! I received a lot of messages these last months and it helped a lot to keep moving forward.

See you soon.

Introducing Clouder, how you can easily create and maintain your infrastructure

Hello world,

Today I’d like to give you the last news regarding the evolution of my hosting orchestrator ( .
I am very excited to make this announcement because the system is finally ready for a wider use and I am really proud of the result.

As you probably already guess, the project has been renamed into Clouder. It is still a suite of Odoo modules which connect to your server to install your applications and maintain them in a professional way.

My goal is to provide a standard way for sysadmin, hosters or simple geeks to easily build a full open-source infrastructure, and thus popularize the great open-source applications we can find.
Today, even if you are a sysadmin, you have the choice between closed-source services immediately usable after filling a form or spending days trying to install open-source application, often without respecting state of the art rules of hosting (supervision, backup, security etc…)

I want to change that. We need to build together a technology which allow anybody with some technical knowledge to deploy his own open-source and state of the art infrastructure. This is the goal of Clouder.

Yes, there is other initiatives for this, but if you look closely do you know any open-source orchestrator which :

  • Allow to deploy application just by filling a form
  • Is extensible enough to integrate any application
  • Manage several servers, not the one where it is installed
  • Support links between applications, to build a consistent infrastructure, not just deploy isolated application
  • Is flexible enough so professional hosters can use it for their purpose so don’t have to create their own scripts anymore.
  • Cherry on the cake, is directly linked to the most powerful open-source ERP on the market, to manage tickets, invoicing, e-commerce website etc…

This is what I propose with Clouder. And the story start today, for more information I invite you to subscribe to the following websites :

Finally, for those who follow this project since my previous post about it, here is what was done since last summer :

  • Port the code from Odoo 7 to Odoo 8
  • Rewrite all the code with the new api
  • The code is now PEP8 and OCA-quality
  • Move some important services, like backups, archives and container registry inside containers. Now except docker itself, nothing is still installed outside of containers.
  • Prepare documentation and websites to manage the community I want to build around it.

So now everything is ready for the real start. If you want to join, subscribe to the mailing-list and follow the repo on github.
For the next step, I’ll need your help to build template modules for others applications or improve existing ones. There is already a good numbers for Odoo, Drupal, WordPress, Seafile etc… but we need more.
Also even if it’s really easy to use one configured, the first installation is still quite difficult so I need people to try to use it, help others, and maybe later try to improve it this first step.

I don’t want people to lose time installing Odoo on their server, so if you don’t have an Odoo already you can ask me to create a free Clouder instance for you, just contact me.
If you already have one, you just have to install the Clouder modules on it.

Unfortunately, I’ll not give a conference about Clouder during the Odoo Experience event in June, but I’ll be here if you want to meet me.

Thank you for your attention. Let’s make this real!

Presentation of my hosting infrastructure

Hello world,

Today I’d like to present you what I did for hosting the projects I am working on. I started to work on it one year ago and after endless refactoring I finally ended up making something I am quite proud and even maybe something other people will be able to reuse.

As you know, my very goal is the promotion of open-source softwares to people without technical knowledge. Either for Wikicompare or the community verticalization of Odoo, I want everyone to be able to easily create their own instances simply by filling a form on a website and for this I needed powerful, flexible infrastructure and something I’ll not spend too much time on.

Flexible, that’s the word. Wikicompare is based on Drupal and the community verticalization on Odoo. Moreover, I want to easily host any open-source application I’ll want to promote later.
This means completely different infrastructures and problematics and I iterated numerous time to make sure everything was flexible enough. Today I can already host Odoo, Drupal, WordPress and Seafile as availables applications and postgres/mysql/shinken/postfix/bind/piwik for the internal components. Now that the core is done, it take me only one day to add a new application.

The heart of my infrastructure is what I call an orchestror (can’t find any good English word for “chef d’orchestre” and I don’t like “conductor” please help me), I mean here the system which will contain the list of all servers/domains/databases managed by the infrastructure, call the executables on the servers when we launch an action and provide a backoffice for easily maintaining this list/launch actions.

Currently, I feel that all hosting companies developed their own scripts to make the orchestror, I guess they see it as a competitive advantage but this not my point of view. In my case hosting is only a cost and that’s why I open-sourced everything I did on this topic.

Tell me if I’m wrong, I tried to search if any open-source orchestror project already exist and I weren’t able to find any good enough and especially flexible enough.
Chef/puppet/salt aren’t designed to easily create and manage custom field/objects, and same for OpenStack I think. I tried OpenStack one whole day and it looked too complex for me and I wasn’t sure there was a way to create a backoffice with it, I think it will a be a good sub-component but not for the orchestror itself.
On the next part on this post I’ll present my orchestror, tell me if you can do everything I do with any existing open-source project, my goal is not to duplicate again the open-source effort, it’s the contrary.

So, I said I need the flexibility of specific development for the orchestror, with me this means odoo module. So yes my orchestror is some odoo modules you’ll find here : You can install them on an openerp V7 if you want to see the whole interface.

I think today it’s critical to separate each component on a server (separate each odoo install for each customers for example). Some may think about virtual machine but it cost too much CPU/RAM, you can’t have hundred of VM on a server, and so my preference goes to containers.
I’m not the only one, Docker is all the rage everywhere currently, and many odoo hosting company I spoke to use OpenVZ extensively since a lot time. I decided to go with docker, and my orchestror is designed around it.

I also currently use shinken for supervision, piwik for analytics, bind for DNS, postfix for sending mail/fetchmail and nginx for proxy.

Ok let’s show you some screenshot of the interface now, here is the main menu :

Main page

On the left menu you’ll find all concepts managed by the orchestror. Base, services and containers are the main ones.

Containers of courses represents the list of all docker containers.

Container list

You can for each one specify the application to host inside the container and the image version to use, the ports open and the volumes containing critical files. Container form

Services represent each install of the application inside a container (for example several odoo install on different ports inside the same container). Here you can specify the service name, the container it belong and the version of the application to use. A database user will automatically be created in postgres or in mysql. Service list Service form

Finally, base represent both a database and a domain. It is linked to one url which shall then be interpreted by the service. Here you can specify the domain name, the service used, the users credentials and the SSL certificate. It will automatically set the DNS, create the database, set supervision and configure the proxy. Main page

Base form

I think the three concepts container/server/base are keys to have the flexibility we need. With this you can do multi-base hosting like Odoo SA on few services, or on contrary create one container for each customer to ensure security. Multi-services is also really interesting to easily duplicate the production environment into a test one before a migration.

You can also of course manage the list of your physical server and domains you own
Server Domains

Saves are listed inside the orchestror, so you can easily restore them.Save list Save form

The list of docker images are also stored here. For each image you’ll be able to specify the ports, the volumes and the dockerfile used to build it.

When you configure the image, you just have to click on the build button, a new version will be created and you’ll be able to use it in the container list. Image

It’s the same for applications, you’ll be able for each application to manage the default image to use, the time between autosaves, the other applications they are linked to and the code used to build the application version.

The build method differ for each application type, for example for Odoo I used the anybox recipes while for Drupal I use drush files. Application

I think this conclude the presentation of my hosting infrastructure. There is still a lot to do to be state of the art but I truly believe that the foundation are solid and flexible enough.

I think there is a lot of dispersion of the open-source community on the hosting question, and even in the odoo community itself. If what I did seems good enough for you don’t hesitate to contact me so we can build a community around it.

A simple system with preconfigured profiles for every open-source application that every association/company in the world can use, in order to provide professional hosting to every people without technical knowledge. I truly believe this is what we need to promote open-source, maybe it can be this, or anything else, but we really need this.


With the many questions I received after this post, I moved the github repo into
and I created a mailing-list :
Feels free to subscribe!

Summary of all the companies/projects I am involved

Hello world,

I know my communication was quite confusing theses times so I think a little summary of what I am currently doing may be useful.

Yolo Consulting is the company I use to sell my services on Odoo. I am the only shareholder and it’s gonna stay like this. Other than my five-year experience on Odoo and the reliable teams I work with, I also mainly use this structure for providing hosting for the others projects (the hosting code is available here and I shall precise we are very good for hosting verticalization. I you have any verticalization development (Odoo for hotel or school for example) you shall contact me.

EmpowerUs is the main company behind the Odoo Community verticalization ( which will go under the control of the OCA. I am not the only shareholder this time, I am launching this company with Julien Dussart ( which take care of the commercial part. We will host Odoo instances for communities with a price per moderators (and not users) and a commission on transactions in €/$.
We also plan to build a network of integrators, with peoples specialized in assisting communities like “Les Valeureux” (, a french association and one of our close partner (many ideas for the community verticalization come from them). If you are providing such services, anywere around the world, feel free to contact us.

Finally, Wikicompare is a platform for collaborative comparisons, based on Drupal. The purpose of this platform is to create and organize communities about comparison, so they can find and debate the important criterions and then judge the products we use everyday with a benchmark approach. I strongly believe it’s the perfect counterpower to the companies’s marketing department and probably my project which will have the most impact.
While it’s already working quite well (, there is no business model at short term and so I needed to concentrate on EmpowerUs. If you find the concept interesting, we are looking for people who want to try creating some first communities.


Odoo verticalization for communities and OCA

Hello world,

Here is (finally) some news regarding the evolutions of the Odoo for Communities verticalisation (, I’m sorry for not giving any news since the OpenDays but the amount of work during summer was just insane.

So here’s what happened :

-A considerable rework of the code was done in July, to improve the usability, the overall consistency of the project and better separate each module (Mainly, the module managing the alternate currencies and the marketplace are now separated in account_wallet and marketplace).
You have a demo platform available here :

-Tests scripts was written for each modules, I added some demo data (I used the OCA as example) and the code is now PEP8. In general, I think the code is now clean enough for contributors.

-A web module for the marketplace, funded by a customer, is currently under development and shall be delivered soon.

-During August I reworked all my hosting scripts at my company Yolo Consulting to use Odoo as controller with good use of docker, shinken, bup etc… Since I also need it for Wikicompare I designed it to be able to host any applications and I hope I’ll later be able to promote some open-source softwares thanks to it. You know my motto, I want to bring open-source software to people without technical knowledge, so they only have to fill a form to be able to use them.
It’s a key point for both EmpowerUs (my brand for Odoo for communities) and Wikicompare and I prefer sharing this cost. So everything is available in AGPL and if you need to create a hosting platform you shall really have a look.

-And last but not least, as maybe some of you saw in the last OCA minute, the Community verticalisation was accepted inside the OCA. I know it was a difficult decision because it’s the first time the OCA include a project with so much code lines and I want to thank everyone for this. The modules shall be included one by one to make sure reviewers have enough time to inspect what is already done.

So what will happen now? As you know I am the only developer in the Community verticalization for now and so every decision and conception was done by very few people until now.

This stop now, once the OCA repo will be created every bug and every suggestions will be reported on github tickets so the community will know what happen. Of course all OCA rules, especially the MP approved by 2reviewers, will also apply and contributions are more than welcome. Until the repo is created, all the code is available here :

Also for all important discussions I’ll throw a topic on the mailing list of the verticalisation (, same than before since mailinglist are still an issue for the OCA) so if you’re interested by the project I strongly recommend you to subscribe!

Thank you and see you soon.

Odoo CommunityTools mailinglist

Hello everybody,

A little post following the Odoo OpenDays. I was there to present the Odoo CommunityTools and it was a great success, many people were interested by what we are trying to achieve, including the main members of the OCA which now I know are watching upon us.

Currently we are four people working of the project, after the Odoo OpenDays we’d like to give the chance to people which are following us to have more details about what’s happening and to participate to discussions about the conception of the modules.

So we need to coordinate. I wanted to ask to the OCA to provide us some tools for this, but with the migration to Github and the general assembly they need time to first decide if CommunityTools should be an OCA project, and second to think to the tools they will themselves use. Which is all perfectly understandable.

Awaiting some news from their side, I propose to use a simple launchpad team to use as mailing list, since everybody know how to use it. This is temporary because we will leave launchpad soon or later, but I believe this is the good thing to do for now.

Please, if you are interested to see or participate to the discussions, join the mailinglist.

I plan to post soon two important mail :

-The first presenting the modules already developped.

-The second presenting the improvements we’d like to do to the Odoo groups management.

I hope I’ll have time to write them this week.

See you soon.

Presentation of Odoo (formerly OpenERP) CommunityTools

Hello everybody!

Today I’d like to present you another project I am working on. I already made a presentation here about the Wikicompare project [french], so to give some news about it we already have a fairly good MVP (Minimum viable product) and we are waiting for a good occasion to start using it.
You can see the MVP here :, it contains some demo data on the ERP topic (false data of course, please no troll :) )

During the development of Wikicompare, I had the chance to meet “Valeureux”, a french association who want to create tools to dynamize communities. They think, and do I, that communities will take more and more importance in our world, but many fails to develop their members size, to really implicate them, argue for personal reasons when decisions must be taken and simply fails to reach the objectives which first bring his members together.

We think that there is currently a lack of tools for them, and also that not many people know which process shall be used to manage communities. This is what we want to resolve, we want to create a framework for communities which will contains all the tools they need, kind of an “ERP for Communities”.
There is some other peoples who try to achieve the same goal but not with the same technology. They call this kind of system NRP, for Network Resource Planning, of course derivated from ERP. We like this acronym, as this demonstrate clearly that NRP is an evolution of ERPs adapted to manage communities, and we will call the Odoo CommunityTools as such.

We think that Odoo is the right technical base for this project. The OpenObject framework is one of the most efficient for such process-oriented project and Odoo already contains a lot of modules needed, such as messaging, groups, project and timesheet, forum etc… and many communities will also need more business oriented features like sales management, CRM or accounting.

Which features ?


The first problem we need to resolve is to better implicate the members. The interactions between members are critical for this and one way to improve it is to allow them to exchange goods and services, this is why we developed a marketplace module, so everyone know the needs of the other members and can contact them.

Alternative currencies are getting popular these days, for example with global digital currencies like Bitcoin or local currencies like “Abeille” in France which aim to improve local exchange by keeping the currency inside a territory. We may also quote the very-used WIR in Swiss which exist since quite a long time.
We’d like to help theses alternative currencies and so the CommunityTools support many form of them :
-External currencies, which means all currencies not generated by the framework. It can be official currencies like euros or dollars, or bitcoins, WIR etc… You can use them to pay an exchange in the marketplace, it can be declarative (receiver confirm on NRP that payer gave him the money) or trusted third party (payer first charge his account by giving the money to the owner of the NRP, which then update the receiver account after the exchange).
-Auto-generated currency, this currency only exist in the NRP and is very useful for exchange inside the community. It is also controlled by the community leaders, and can be easily used to reward the people who are most implicate in the community goal. The way this money is generated shall be decided with caution, for example a fix amount is distributed at the beginning or a fix amount for each new members, to avoid inflation.
-Abundant, there is no real limit to the generation of this currency and it can’t be used to pay an exchange in the marketplace. You can for example define an abundant currency “Thank” so a user can generate a “Thank” when he want to simply reward someone for an action. You can compare this kind of currency with a “Like” in Facebook.
With the exception of the abundant currency, each currency is manage with the accounting module of Odoo. Each user has on his personal page the balance of all the currencies defined in the NRP.

Of course, you can also create on the marketplace announcements about the main needs of the community, to more easily find people to resolve this need while motivate them with the currencies managed inside the system. Thanks of this, the marketplace module is one of the best way to implicate members while finding resources to reach the community main goal.


Another important module is the governance module, which allow the community to make decisions.
Most of us know the vote by majority, decision is taken if 50%, 2 / 3, whatever, of the people vote yes. This is not really the best vote method, as the decision is taken while many people were against it, and the will of the minority is simply ignored.
There is many others decision method, like the decision by consent ( where the goal is to debate until nobody has justified objections (not meaning everyone have to agree).
Also for election, there is another interesting process called election without candidate, where no candidate are proclaimed before the election but instead each people spontaneously say a name. The one with the majority of vote is then elected, if nobody has justified objections.
Most of theses ideas come from alternative ways of governance like Sociocracy ( which we would like to promote. Within the governance module, you’ll find classical governance method like majority but also alternate ones so you can decide which one you want to use. We planned to design the module so facetoface event is unnecessary, so decisions can be taken with people geography distant.


For a community to be efficient, you often need to divide it into group to better implicate their members. For example one group is in charge of the development, another the design etc… We use the group concept already existing in Odoo, we improved them so they can be in arborescence, in marketplace there can be announcement in a group context or even created at the name of a group, decisions can be made within a group etc… In short, this is a concept we find in every CommunityTools modules and which define the context.

Another important point is the communication. Here there is nothing to reinvent since Odoo already had it : internal messaging linked to record, notifications, instant messaging, forum. Not only communication between members but also to the exterior, the blog, the events, the CMS itself will allow the community to better communicate about theirs actions and what they want to do.
Personally, I am still unsure about the quality of the CMS, I worked in a webagency and I have some doubts that it can really be used for a professional design for now, but the new CMS for sure is simple to use and I’m sure this simplicity can be a really good point for communities which will concentrate on the content they want to publish and need a tool they can immediately use.

I only gave you some of the features we plan to develop, Social CRM, Crowdfunding, Reputation management and many more are also on the pipe.


The power of theses modules may be difficult to understand, so let me take an example. If they are interested, what would theses modules bring to the Odoo Community Association if they have an NRP?
An integrator member from Spain may have a bug he isn’t able to resolve, he will post an announcement on the marketplace proposing 100COM (auto-generated currency) for anyone able to resolve the bug. Another member from Brasil help him and then obtains the 100COM that he will then be able to use if he need some help from another member. Of course, you can also use the marketplace to find a subcontractor for a customer project, with currency in euros or dollars.

Similarly, if a module need to be developed, the decision to fund his development will be taken thanks to the governance module, and then first an announcement will be posted on the marketplace to make the conception. People interested use the communication tools included in Odoo to make the specifications, and then another announcement is posted to make the development.
Once the module is ready, the people invested in conception and development are rewarded either in COM, so they can use it with the other members, or real currency if the module was funded by crowdfunding. They also win some reputation points, and so gain some visibility inside the community.

This is only a fraction of the potential, I hope there will be many people to share this vision with us.

Who might be interested by this platform ?

First of course the associations. Due to their often lack of money, they don’t have real good tools currently and an open-source platform which have all the features they need would really be a relieve for them. We also want to give them some way to improve their financial situation, for example by controlling the access to the platform with a periodic payment, taking a tax on the transactions of the marketplace or simply crowdfunding.

Local authority (at least in France) also shown some signs of interest in such a platform, especially the marketplace. Their main preoccupation is to develop the local economy, and there is already a lot of experimentation to implement marketplace + alternative currency. We would like to provide the framework they need for this kind of projects.

Co-ownership, and in general any neighbor community. I personally see a great potential here with the marketplace, because it’s a way to know who my neighbors are, they skills and interest and ask their help if I need them, and conversely. Also, the governance module may be a revolution, by allowing to make decision for the coownership without waiting for the yearly meeting.

And Enterprises, by many ways. It can be a way to improve exchanges between employees and so better implicate them or trying a less hierarchical governance.
Even in prospection and sales, I believe the concept of community will take more and more importance. Even now social network like Facebook/LinkedIn etc… take more importance in prospection and we start to see many tools to manage the communities on social network, this is called Social CRM and we can quote tools like Nimble, NationBuilder etc… I really think that it’s critical for Odoo to have such module, I don’t know if it’ll be in Odoo SA official module or in CommunityTools but we really need it.
Community management for business is often simply summarized by some communication from the entreprise to the community. No co-building, no co-decision, no real exchanges between members, this is by no way real communities according to me. Everything still need to be invented and by being the best open-source ERP and with the CommunityTools, Odoo might very well become the perfect tool to achieve this.

Thank you for reading this long presentation.If you want to look at what is already done, you can go look the code here : .

We don’t have any website yet presenting the project but we will do a presentation at the Odoo OpenDays 17h20 on Thursday “Use Odoo as NRP, Network Ressource Planning”. If you have any questions feel free to contact me at


Présentation du projet “Wikicompare”

Bonjour à tous,

Après une très longue période sans poster de nouveau billet, je reviens vers vous aujourd’hui pour vous présenter le projet que je prépare depuis maintenant près de deux ans.

Il est probable que j’arrête de poster au sujet d’OpenERP à l’avenir car quoi qu’il arrive je vais passer à autre chose dans les prochains mois. Mais vous pouvez être sûr que je vais continuer à suivre son évolution de très près.

Le projet Wikicompare

Le projet consiste à mettre à la disposition des internautes une plate-forme de comparaison open-source, complète et collaborative.

En d’autres termes, un internaute pourra aller sur le site du projet, rentrer ses coordonnées et le type de produit qu’il souhaite comparer, et nous mettrons alors gratuitement notre système à sa disposition.

Il va commencer à saisir ses critères de comparaison, les premiers produits et si les produits supportent ou non les critères en question. Les premiers visiteurs pourront alors apporter leur aide en suggérant de nouveaux critères qu’ils estiment importants, et qui seront ou non validés par le créateur.

De cette manière, les données de la comparaison proviendront non pas de partenaires mais des internautes eux-même.

Historique et genèse du projet

Comme vous le savez je viens de la communauté OpenERP, et de part mon implication dans celle-ci j’ai mené de nombreuses réflexions sur comment promouvoir OpenERP.

Les ERPs sont des logiciels complexes. OpenERP est un jeune logiciel qui a une dizaine d’année, à comparer aux leaders du secteur comme SAP qui a plus de 40 ans d’existence.

Par ailleurs, les ERPs sont assez opaques. Les éditeurs ne communiquent pas les détails du fonctionnement de leur produit au grand public, des détails comme ce qui se passe lorsque l’on valide une commande, comment se gère la date d’échéance des factures etc. Ce genre de détails que je cherchais lorsque je me formais sur les ERPs, c’est à dire finalement à connaître l’état de l’art dans le secteur.

Parmis mes contributions, j’ai lancé le site openerp-universe qui visait précisément à déterminer les fonctionnalités d’OpenERP et celles qui pouvaient manquer. C’était avant que je réalise qu’en fait c’était l’ensemble des ERPs que je devais comparer, j’ai donc commencé à travailler sur un autre plus ambitieux et qui comparerait l’ensemble des ERPs.

Par exemple, Dire que l’ERP N°1 implémente la gestion des dates d’échéances de telle façon, l’ERP N°2 l’implémente d’une autre façon, mais l’état de l’art pour l’implémentation de cette fonction est la façon N°3. Et de dire que tel secteur d’activité a surtout besoin de la fonctionnalité A ou B, donc le plus adapté est l’ERP N°2.

Cela aurait permis de crédibiliser OpenERP, en démontrant que certes il reste encore beaucoup de travail, mais sur les fonctionnalités vraiment importantes il est au point. Il aurait permis aux personnes qui se forment sur les ERPs d’apprendre beaucoup plus rapidement l’état de l’art, ou permettre aux développeurs de savoir comment implémenter les fonctionnalités. Enfin, il aurait permis aux contributeurs de mesurer le travail qui reste à accomplir, ce qui est une profonde source de motivation pour un logiciel libre.

Comparer les ERPs à un tel niveau de détail nécessitait des millions de critères, j’en étais conscient. C’était quelque chose de jamais-vu, et je savais qu’il n’y avais qu’un seul moyen : Le crowd-sourcing.

Cette idée est en quelque sorte la conclusion de mon mémoire de fin d’étude, et j’ai travaillé un long moment sur sa conception. Et une fois la conception terminée, j’ai réalisé que la problématique de comparaison des ERPs était tellement complexe, que le système que j’avais conçu pouvait en fait comparer presque n’importe quelle thématique.

C’est alors que le projet a totalement changé. Il n’était plus question de mettre en place un site collaboratif de comparaison des ERPs, mais de mettre à disposition ce système aux internautes.


A travers ce projet, les internautes vont soudainement pouvoir débattre et définir les critères qui sont importants pour eux, ils vont pouvoir communiquer autour de ces critères, ils vont pouvoir décider quels produits les supportent ou non, et enfin ces informations vont remonter sous la forme d’une note qui sera instantanément compréhensible par les autres internautes non-experts.

C’est un très grand pouvoir que ce projet va confier aux internautes.

Nous sommes en train de parler d’un projet qui pour l’instant n’existe absolument pas. Un projet qui est potentiellement disruptif dans un très grand nombre de secteurs, pour ne pas dire tous. Des start-ups qui proposent des produits innovants pourraient grâce à cela avoir un départ canon, tandis que les entreprises établies pourraient voir leur produits se faire très fortement critiquer par les internautes.

Là où il a du changement, il y a des opportunités commerciales. En ce sens, en lançant ce projet et en tant que personne qui entrevoit ce qui va arriver et peut les anticiper, mes motivations sont en grande partie financières. Je ne le cacherais pas.

Néanmoins, c’est loin d’être ma seule motivation. Je considère que notre société actuelle est déséquilibrée, les entreprises ont trop de pouvoir, et c’est mon choix de vouloir donner aux clients, à travers les internautes, les moyen de se défendre.

En plus de cela, la riposte des entreprises ne pourra se faire que par la transparence. Elles devront communiquer sur plus de points de détails que ce qu’elles ne font actuellement, et surtout sur les critères pointés par les internautes.

Je viens du monde du logiciel libre, je crois en ses valeurs, et j’espère que ce projet pourra apporter plus d’équilibre, plus de transparence à notre société.

Où j’en suis actuellement?

Tout le cœur du développement a été fait en préparation du projet, même si ce développement doit être revu par des développeurs plus expérimentés. Le système est d’ores et déjà opérationnel mais a besoin d’être testé: (login : wikiadmin / admin) .

Je me suis basé sur un système Drupal. La raison principale c’est que Drupal est très bon pour gérer tout ce qui est collaboratif, et m’a donc permis de gagner beaucoup de temps sur toute la partie authentification, commentaire, historique, traduction etc… Par contre j’espère qu’on arrivera à l’optimiser suffisamment pour gérer les millions de critères que certains wikicompares risquent d’avoir.

Pour lancer le projet il reste donc, sur le plan technique, à :

  • Réaliser les maquettes graphiques, et les intégrer.
  • Définir les contenus.
  • Développer le site central, qui permettra aux internautes de créer leurs wikicompares.
  • Développer les scripts d’hébergement qui permettront de gérer le parc de wikicompares.

Le système étant open-source, et en vu d’assurer la qualité du système, il est important de former une communauté autour du projet. C’est la prochaine priorité et le but de ce billet.

Enfin bien sûr, avant de lancer le projet, il me faudra également finir de mettre au point toute la partie commerciale, chercher des partenaires etc… Pour cette raison, il est peu probable que dans les prochains mois je puisse passer beaucoup de temps sur la partie technique et je vais avoir besoin d’aide pour stabiliser le système.

La licence ?

Sauf si quelqu’un me trouve un contre-argument valable, ce sera de l’AGPLv3.

Comment aider ?

Je vous invite dans un premier temps à vous inscrire sur le google group du projet, qui au début fera office de forum et de mailinglist :!forum/wikicompare.

Je communiquerai principalement par ce biais, notamment pour vous tenir au courant de l’avancement des étapes du projet. Si vous avez besoin d’aide, que ce soit pour installer ou pour utiliser le wikicompare, poster un message sur le groupe est la meilleure manière d’obtenir une réponse rapide de ma part.

Le code, qui est un module Drupal, est disponible sur mon github :

Toute ma TODO list est dans la gestion des « issues » de Github. Néanmoins comme je vais passer plus de temps dans les prochains mois à faire de la gestion qu’à coder, un petit coup de main pour m’aider à clôturer les tickets ne sera pas de refus.

Si vous ne pouvez m’aider au niveau du code, juste utiliser le wikicompare sera déjà beaucoup. A ce stade, avoir quelques wikicompares qui fonctionnent apporterait déjà beaucoup à la crédibilité du projet, sans compter les retours d’expériences qui permettront d’identifier les points à corriger. Attention néanmoins, établir une liste de critères est un travail long et fastidieux, je vous recommande de vous y mettre à plusieurs et de ne pas cibler une thématique trop complexe.

Enfin, vous pouvez m’aider sur la documentation via la partie wiki sur le github. Je vais essayer de documenter autant que je pourrai mais je ne garanti rien.

Enfin dernière précision, je vais surtout publier en français au début sur le google group histoire de garder une certaine maîtrise sur le démarrage du projet, mais je commencerais à publier en anglais dès que le projet sera lancé. Le github fait exception, tout est déjà en anglais de ce coté là.

Je vous remercie d’avance pour votre soutien et j’espère que ce projet saura vous intéresser.