Community mailing list archives


Re: Odoo v9 Community and Enterprise editions

Akretion, Raphael Valyi
- 05/30/2015 19:18:18
=Unfortunately license are either copyleft, either they are not. So there is no middle way.

Well eventually Odoo could have had several components with different codebases, or services and eventually these services would have had different licensing schemes. That was a solution proposed by me in this list long ago BTW. Now, if a monolithic architecture is chosen and reaffirmed with that flawed web technology introduced 2 years ago in a monolithic architecture and again under the same AGPL license purposely.. Then yes it can have only one of these two licensing patterns. And we should adapt our business to its limitations instead of loosing the North and loosing the existing benefits by trying to move to the other side without being successful at doing it.

Honestly it's fine to have a copyleft license. If some startup really want to protect their IP, then they PAY YOU to put that IP logic OUTSIDE from the ERP runtime and this is just fine, this is even more scalable if you think about a web front. It's very little different than a company using MySQL or MongoDB and putting their IP logic client side instead of coding them as SGBD engine extensions where they would be subject to the copyleft licence of these SGBD's. It has always been fine, it's just an architectural detail.

And NO, not all microservices architecture imply the bureaucracy of something like the the Magento connector. Not at all. At Akretion we do it and it's in prod since 1 year already. If you go to the community days, I suggest you get in touch with Sebastien Beau to learn more about this open source technology (I proposed to present it at the last community days but the talk was rejected). A micro-service architecture can duplicate the data or not (like ooor, erpeek etc...). But when they duplicate the data, they can alleviate 90% of the SQL integrity burden by using NoSQL patterns like we do with our Apache SolR high availability cache data store.

Now the good thing about using a micro-service architecture to work around the copyleft limitations is that closed source is the exception, not the rule: a startup can still protect its IP while we mostly avoid the cannibalization of the existing eco-system with a market of proprietary modules and while we avoid to put the whole ecosystem under the threat of some Odoo SA buyout.

So you see, I don't have an angelic vision about this. I just say, we should adapt smartly to constraints and keep moving forward instead of destroying the open source capital of the project with licensing indecision going back and forth, licensing incompatibilities and unfair and doomed pressures to try to change the license of third party codebase like the one of the OCA.


On Sat, May 30, 2015 at 4:40 PM, Nhomar Hernández <> wrote:

2015-05-30 7:59 GMT-05:00 Raphaël Valyi <>:
Some partners don't mind too much when they work in some of the richest countries were that price is eventually possible. But for those like us who bet on developing markets and did tremendous efforts to create localizations from scratch, during all these years, we have seen these Enterprise price been totally out of the market by a factor of 3x to 5x without even considering the margin for our local support. So basically that wouldn't be possible to work with Odoo in these markets...

That being said, it's extremely unlikely that the license change of the core affects the whole ecosystem. The vast majority of serious modules authors seems to understand their interest goes with keeping an overall AGPL licensing. Also if you look at stories like Compiere/Adempiere, it's exactly because of developing countries economics that the Adempiere forks started (it started in Indonesia and Latin America and survived the Compiere proprietary fork).

I am agreed in terms of Price dude.

But I think we need to put our feet on ground and understand we have 3 kind of companies working with odoo.

1.- The ones which can pay (on big countries).
2.- The ones that can not pay because is not possible.
3.- The ones that "decide" not pay because they are not convinced of the price/model or whatever.

I think it is fear enough think in terms of what is profitable and what is necesary and what is good.

I mean.

The fact that (1) pay and finance the R&D and use last version and paid apps and the (2) use community and use the R&D with some kind of delay or make them own low cost R&D in community and/or privately and (3) make the same but with an stupid reason is good, it is evolution.

I dedicated some time to understand all  the necesary effort for an "Open" apps store and honestly I could not find a correct way where invest N thousand dollars on it and open a "Free as Free beer" market for (3) without feel it is not fair.

2 thinks can happen in the next future:

1.- Odoo fill the economic gap to open everything being profitable at same time or.
2.- Odoo fail and try other thing in the future.

Let's wait and support the fact that without a change of licence it will be almost illegal make a privative option forsome modules.

I always use an expression: "If you are not agreed with some law, don't brake the law, change it if you have the power".

Then until today you know several partners make A LOT of private investment which do not put publically available, then even if they respect the AGPL (using a parallel technology) or simply fail the licence (puting private code where they didn't paid the Licence) what we need to work on is continue demonstrating with facts and not with words that "open is Better" but without attack the fact that the actual law do not protect anything in a practical way, it only become "illegal" something that is necesary.

LEt's wait and work for the option that is convincing more to yourself.
Saludos Cordiales

CEO at Vauxoo Odoo's Gold Partner.
Nhomar Hernandez

Post to:

Raphaël Valyi
Founder and consultant
+55 21 3942-2434