Community mailing list archives
Re: Odoo / Python 3by
And you know that last 2 years Odoo take off is partly due to VCs cash. May be R&D of Odoo is lean. but not the business strategy.
Hi Zahid,Fabien said in another thread that they are following Lean developpement methodology... So that they don't see beyound three months. Of course that they will take action before 2020... But without the move of Odoo SA... I don't see what the community can do... community module won't move to py 3 without the core... and migrating the community version to py 3 without odoo means forking Odoo... And nobody want that...2015-11-16 11:12 GMT+01:00 zahid jalal <email@example.com>:Hi folks.This thread is open of a week. And I just can't understand why we don't hear from Odoo S.A about this.Their silence does not bode good things. It is not possible that they did not discuss py2.7 EOL internally. Every python developer is aware of it, they are not an excpetion.
I wont be surprised , if v10 or v11 core goes py3 and closed as well for entrprise edition.IMHO, Odoo community should take action right now, at least let's start writing a road map to migration to py3. Sustainability of Odoo as openesource software depends on it.regardsOn Mon, Nov 16, 2015 at 9:39 AM, Dave Ellison <firstname.lastname@example.org> wrote:
"Python 2.7 is not going anywhere anytime soon."
That's what worries me about the Python community. Coming from a recent convert from other languages. My experience for learning Python was mixed. Some recent material online is still teaching 2.7 and some have now moved to 3.x or their software is now using 3.x. Which means learning can be a pain. Yes I am aware moving from 2.7 to 3.x isn't too bad, but for people learning it can be a royal pain.
Although for different reasons, lack of movement killed Perl. I am not saying Python will go that way, but the slow movement to 3.x will put people off.
On 13/11/15 18:03, Luke Branch wrote:<blockquote cite="mid:3836284C-F4A2-4DEF-BF32-B45F6DB5B6AF@gmail.com" type="cite">
On an unrelated note, you wouldn't happen to have a link to that tool by any chance would you?
I think the reverse could also be true and some useful modules could be explored from the Tryton project for Odoo, particularly some of the integrations/connectors.
Regarding Python 3, I'm sure Odoo SA have a game plan and timeline in mind. Python 2.7 is not going anywhere anytime soon.
I believe the number of Python libraries written in 2.7 still far exceeds those that have been written in 3+. Perhaps it makes more sense to stick with 2.7 for this and many other reasons so that they keep their options open in terms of libraries they and the community can easily integrate into their projects.
Here's a few links that might interest you:
Sent from my iPhone
On 14 Nov 2015, at 1:03 AM, zahid jalal <email@example.com> wrote:
2020 is tomorrow, let's hurry. :)@houssine 4 years is tomorrow in IT industry.About Tryton, just think about it. if Odoo goes py3, what will stop Trytoners from borrowing some code from Odoo. I saw sometime ago a tool (ETL like) to convert openerp DB to tryton. so why not.
On Fri, Nov 13, 2015 at 2:47 PM, Houssine BAKKALI <firstname.lastname@example.org> wrote:
note sure this matter has something related to Tryton...Zahid,2020 is in four years... it let Odoo SA and community the time to think about without hurry...
2015-11-13 14:17 GMT+01:00 zahid jalal <email@example.com>:
rgrdsI think that one reason that there is no initiative to go python 3 is Tryton. When openerp was forked, Tryton guys choose to go py3 for the project. The base code is well structured and cleaner than Odoo. But tryton is ugly (no fancy web), and lacks of important fonctionnalities. but it is also fully GPL and includes a migration mechanism (no upgrade fees).@Hossine :The hole picture here is 2020 EOL of python 2.7.
If one day Odoo goes py3. Tryton can rapidly gain some missing fonctionnalties and gains market share ==> it's FP's nigtmare ( i am sorry for that , but it's just an opinion ).
On Fri, Nov 13, 2015 at 11:34 AM, Houssine BAKKALI <firstname.lastname@example.org> wrote:
Hi Zahid,You have to take the whole picture then. And ask the question : what would be the benefits of porting to 3.x to justify that you'll put 20 to 30 people for 6 months at least on this job and think that you'll need to stabalize it so and some months on top of that which is a huge task in final.
What about backward compatibility? And as said Graeme new version of 2.7 is going to be released next month with improvement on performance...
And i don't say that it hasn't to be done but to evaluate the main benefits and then so see if it's worse to dig in it or not...
2015-11-13 11:37 GMT+01:00 zahid jalal <email@example.com>:
RegardsI hope i am wrong about this. But unless we have a response from Odoo S.A. Our concern and doubts are legitimate.I did sone gross stats. core code is xml+python+js. python portion of code is less than half in total. Migrating to python 3 can be done by 20 to 30 developpers within 6 to 7 mounths.IMHO : Something is going on inside R&D departement at Odoo SA. Next move will be when releasing V10 (entreprise+closed core+ another licence change+ python 3) and V10 community in python2.7 as there is some backward compatibility.On important thing about this topic is official Odoo SA position about python 3. Why we don't hear from them ?Hi allPlease guys. This is meant to be serious. ( so please stop feeding the troll).
So try to fix your "Website Edit Button not working for Website Editors" problem and then maybe you'll maybe get some credit... until now this is just trolling and hijacking mailing list...yeah for sure... any company running a business of hundred thousand of euros care about python 3.x more that have their employees performing their day to day operations which enable the company to make money in...You talk about big NGOs, and you talk about going right now on Odoo9 which still unstable for now... This is madness..
@HoussineWhen you start working on projects the size we do since over ten years and for customers which are world wide recognised NGO Organisations, Companies and Universities, than your point of view probably changes a lot. If you are talking about the small shop around the corner you are probably right. They simply have no time and no resources, as they are mostly run by their owners themselves and they really don't care as long as it is running and not breaking down - you are right at that point.
Universities and NGO Organisations have one in common. They don't like so much the use of proprietary things and like to stick more and more to Free Open Source Software Products released under GPLv.2,, GPLv.3 or AGPLv.3 - They are not so willing to get involved in LGPL stuff due to the reasons already mentioned many times.
But much more important is that they would need to run a system which gets maintained on state of the art software and not outdated versions far beyond 2020. Have a look what the needs are of many of those organisations. On the one hand side they need to maintain the website portal with lots of new articles and have spaces where they can represent their projects. Easily you will have sites with much more than 5000 Pages, over 15.000 content elements and perhaps more than 25.000 Digital assets to be managed and not to forget the authors which could be i.e. about 500 not counted the dependencies they have all around the world.
For their internal Project management, for their Human resource management etc all the stuff Odoo has actually already available, those people are using not seldom multiple products which can't even communicate properly with each other and interfaces need to get programmed and maintained etc. It is very costly. To reduce those costs it would be great to have only one centralised system which can be run for all those instances world wide. ODO actually has now since it has the website builder, everything they would need.
BUT and this I write in capital letters as it is a big but. Unfortunately the license has changed to be LGPLv.3 and this makes many of them very sceptical on what the future of Odoo would be!
Right now many of them try to centralise their efforts to one system and to one space. Many of them are using for their websites TYPO3, and others Drupal and some even try to grow up Drupal to be getting something like ODOO. All those systems they can maintain from one single spot in the world, they are multi domain, multisite capable and multi tenant capable and of course they are multilingual (Drupal 8 is the best example for that point as this was very important to keep up with those customers). ODOO has already all those features. it is multi-domain, multi site, multi database, multi core and multi language capable and it is much easier actually to get backups than from any of the other systems. It is very easy to maintain and easily to extend.
Even they are huge organisations they often work mainly with their internal staff which needs to maintain the infrastructure and the sites. They need training again and again and they also need services - this is why they sometimes externals like us to do exactly that.
Right now all of those - also those universities and school organisations are NOT using Odoo but Odoo would be an ideal platform for them in many cases. But have a look to Kuali ERP, Kuali ERP is not a mobile app by the way. You probably have never heard of it before but simply have a look also to their business model and why they are using AGPL! And most important why the use of AGPL brings them those bigger customers which could be very easily also customers for Odoo if the License would have not been changed again.
All those organisations are NOT living in the past but when they intend to change to another system they are looking in a very long future and this future means 10 and more years. i.e. East West Center was TYPO3 for many many years and than they converted their site to Drupal in 2010 as TYPO3 was changing over and over again with each new update at that time - by the way today the new Version 7 has been relaunched which will be also a new beginning! Here it was the community!!! which was standing up and simply started as Community to take action against some others who tried to make things less looking for the needs of the projects which community members had. Quite similar to what is happening right now here at ODOO.
Anyway, or take as another example UNESCO which was running on even two CMS Systems and multiple Systems to maintain all their other needs. Until 2007 they had simply the need to have two CMS systems as one was proprietary but could not write arabic and other let's say "utf8 languages". This was the reason on developer of them decided to get involved with TYPO3. In 2009 they decided to put all instances centralised where possible to Paris and of course based on TYPO3 - and still using multiple other products for all their other needs. Since they are using TYPO3 and the other one is using Drupal they actually have absolute no intention to switch back again to an proprietary system I guess. They actually promote the use of Free Open Source Products on many occasions and even on CDs and DVDs and downloads etc.
To change those kind of customers to something more mature and much better suited to their over all needs, than you need a product which will be available for at least the next 10 years and this is not the case with ODOO when the Python 2.7 libraries get no more maintained and they are already like said - outdated and no more state of the art.
So if you are viewing the past you are absolutely right, keep them on 2.7 until they die and probably you with them as with one simple "bum" your revenue will be gone. But if you are looking to discover new markets and get new customers your point of view should be just the opposite of what you just wrote.
If you start right now and convert ODOO to Python v.3 you could have a very stable product already in 2017 which gives you and your customers enough time to get things done, which means trainings, modules, themes - also this takes time after the core had been converted to Python v.3.4 and later
We are focussing on the future and not living in the past!
You won't get any of those kind of customers with Odoo as long as it is using a library which is already outdated and beside that not even gets maintained anymore in 3-4 years!!! and many parts of it already don't get properly maintained anymore right now.
I really recommend setting up an ODOO AGPL Branch without all those modules and start converting the core to Python v.3 in a Community effort. I am pretty sure that as soon as that branch would be available people from the community would start contributing to it The AGPL Licensed OCA-ODOO base on Python3 could than and will be for sure the base for all future community developments which will focus on projects in the next decade and in parallel the outdated ODOO 9 Python 2.7 Branch or even the than in future outdated ODOO 10 Python 2.7 Branch could still be used for all projects from the past!
As an inspiration the TYPO3 7 Movie
The OCA could do exactly the same and get things done instead of waiting until big boss is changing licenses again and again!
Odoo SA planned to rewrite all modules to the new api for version 10... they previously aimed 9 and they did not did it... So python 3 is a big step and the release date for 10 is around october 2016. They clearly have to made a choice and for a limited added value job that could not just be in the Entreprise version, I doubt that they will put python 3.x as a priority because even Odoo SA has limited ressources...the pills part wasn't for you... but for someone a bit excited in several threads...Never got this problem on prod env for standard installation... maybe some lib aren't maintened anymore but most of them are... Just keep in mind that most of odoo modules still coded in the old api and that lot of still delcaring model through osv.osv
I may be wrong or not...
Customers are not early adopters, they just need a tool, and Odoo in one of many they use, to run their business. Customers doesn't car about python 2.7 or 3.4 so as far 2.7 is maintained it's ok to stick with it as 2020 is in four years. Their are other big important step before the absolute need of python 3.x
So keep calm and take your pills ;)
Ok, i'll remind you to keep calm by then. When you try to install a module or even the server itself in a prod environment, and some buggy py2.7 lib you depend on is not maintained any more.
It is stated clearly from python maintainers that "standard library improvements, for example, are only available by default in Python 3.x. "
It's time to move on.