[wyldckat]

(Note: I've had to write about this so many times already, that I'm writing this thread to make it easier to use it as a reference thread for future questions.)

Naming context

• OPENFOAM® is a registered trade mark of OpenCFD Limited, producer and distributor of the OpenFOAM software via www.openfoam.com.
  • OpenCFD is part of the ESI group, which is why it is usually referred to ESI-OpenCFD.
  • Please read the OPENFOAM Trade Mark Guidelines, if you plan on using the OpenFOAM name.
• The OpenFOAM Foundation - www.openfoam.org - was originally founded by OpenCFD and SGI in 2011, when SGI acquired OpenCFD back then.
  • This Foundation has permission to use the OPENFOAM® trade mark.
  • You can find a bit more about it in the history time line here.

Version numbering convention and features

• The OpenFOAM Foundation uses the numbering convention "I.J", for example: 4.0, 4.1 and 5.0.
  • There is no defined date of when a new version is released.
  • The development is done using many of the strategies advocated in the DevOps programming culture, which is why it provides weekly releases of the current development line (OpenFOAM-dev) here: https://openfoam.org/download/dev-ubuntu/
  • Code contributed to the OpenFOAM Foundation requires that all contributors sign the "Contributor Agreement" - see section "Getting Involved" in this page for more details as to why: How to Contribute to OpenFOAM
  • Access to the bug tracker and development line is available to the public without registration/login needed (unless you need to report a bug).
• ESI-OpenCFD uses the numbering convention "vYYMM", for example: "v1706", which stands for having been released in June (6th month) of 2017.
  • Major releases have a set date, as indicated in the previous bullet point.
  • Major releases sync up with the development line in OpenFOAM-dev that is provided by the OpenFOAM Foundation. You can find the following lines on each release note:
    • OpenFOAM v1606+: https://www.openfoam.com/releases/openfoam-v1606+/
      

      Includes all Foundation code changes to commit level 449a9ec 21-04-2016

      This means that it includes many of the features that were made available in OpenFOAM 4.0: https://openfoam.org/release/4-0/
    • OpenFOAM v1612+: https://www.openfoam.com/releases/openfoam-v1612+/
      

      Incorporates all relevant Foundation code changes to commit level 3140cfa9 09-12-2016

      This should mean that it includes all of features that were made available in OpenFOAM 4.1 (13th October 2016 -
      https://openfoam.org/release/4-1/) and some of the features that were made available in OpenFOAM 5.0 (26th July 2017 - https://openfoam.org/release/5-0/).
    • OpenFOAM v1706: https://www.openfoam.com/releases/openfoam-v1706/
      

      Incorporates all relevant Foundation code changes to commit level d2a62df7 26-04-2017 - with some additional ad-hoc fixes

      This means that it includes many of the features provided in OpenFOAM 5.0 (26th July 2017 - https://openfoam.org/release/5-0/).
    • OpenFOAM v1712: https://www.openfoam.com/releases/openfoam-v1712/
      • Features from the OpenFOAM Foundation (OpenFOAM.org) are now retrieved individually. Look for the lines that refer to "Community contribution" and look into each one to check where the contribution came from.
      • Should include most of the features present in OpenFOAM 5.x.
  • Source code contributions do not require transfer of ownership of the code to ESI-OpenCFD, while the OpenFOAM Foundation does require it.
    • There is a community page that lists projects being integrated at ESI-OpenCFD: https://www.openfoam.com/community/projects.php
  • Access to the bug tracker and development line requires free registration and logging in at https://develop.openfoam.com

For easier reference about each specific version, on the top right corner of the main page at https://openfoamwiki.net - you can find links to all of the release notes for each version.


Which version to choose?

Here are a few suggested steps for helping you decide:

1. Make a list of what features you need to use. Then cross-reference what you need with the ones listed for each version.
2. Still not sure? Pick any or both of the latest versions from each website, but start working with just one of them.
   • If you are not able to install one specific version, try the other one.
3. When you notice that you need a feature that is not in the current version, then go look into the other version(s) and release notes.
   • The other version should be compatible for the most part.
4. If you can't find a specific feature in the release notes, try searching online for it and/or check for other forks/variants or any other contributions (see here and here).

FAQ:

But why?

Why not? It's open source. It's one of its major strengths! You can choose from at least two major types of development lines, both of which are evolving as best as possible, as well as several other forks and online contributions.

See for example RedHat Enterprise Linux (RHEL and CentOS and Scientific Linux) versus Fedora exist.

And there is even foam-extend, which was the second major fork of OpenFOAM till the end of 2014 and is hopefully going to gradually be integrated into the development line provided by ESI-OpenCFD.


Will features from ESI-OpenCFD be integrated into the OpenFOAM Foundation source code?

In the near future, it doesn't look like it, but hopefully this will change later on.
Right now, there are contributions being made by developer(s) who work(s) for ESI-OpenCFD, but they are done directly to the OpenFOAM Foundation (OpenFOAM.org) and then get integrated into the development line at ESI-OpenCFD (OpenFOAM.com).


Any quick guides on changes between versions?

At OpenFOAM.com you can find two pages with at least some information, which you can then try to cross-reference with the respective version from OpenFOAM.org:

• OpenFOAM® User Upgrade Guide
• OpenFOAM® Developer Upgrade Guide

[Ahmed Khattab]

Hi Bruno,

thanks for your detailed explanation.

But why both site claim that they own the OpenFOAM software.

- openfoam.org

OpenFOAM is the leading free, open source software for computational fluid dynamics (CFD), owned by the OpenFOAM Foundation and distributed exclusively under the General Public Licence (GPL)

- openfoam.com
claims that ESI group now own OpenCFD who created OpenFOAM and that OpenFOAM foundation was a part of SGI before the acquisition by ESI

Releases were originally issued by OpenCFD, up to version 2.0.1. A new entity, OpenFOAM Foundation, was created in 2011 following the acquisition of OpenCFD by Silicon Graphics International (SGI) with a board comprising both OpenCFD and SGI members. OpenCFD continued to release OpenFOAM via OpenFOAM Foundation to version 2.3.1. In 2012 OpenCFD became a part of the ESI Group, and ESI members replaced SGI members on the board of OpenFOAM Foundation. From 2016

or did i understand something wrongly?

and why openfoam.org states in their contributes list that

Note that there are no contributors from ESI-OpenCFD (openfoam.com)

thanks

[wyldckat]

Greetings Ahmed,

I had the following paragraph written at the end and I've brought it to the top instead:

Instead of bickering about who owns what now, I would rather that people contribute to the project they like/love the most; and if they can't decide which one they like/love the most, then contribute to both and possibly even to other forks as well: http://openfoamwiki.net/index.php/Forks_and_Variants

So, back to the quick'ish list of answers:

1. I'm not a lawyer and only one or two specialist lawyers can give you the pure unadulterated definition as to why both "own" OpenFOAM software.
2. I actually didn't know that this page existed: https://www.openfoam.com/history/ - or at least I don't remember seeing it before.
3. ESI-OpenCFD owns the trademark and owns part of the copyright over the versions they have been releasing since v3.0+, i.e, they own a good chunk of the new code they've put into they releases.
4. The OpenFOAM Foundation has a permission to use the trademark and owns all of the copyright over the code they release.
5. The OpenFOAM Foundation even owns some of the recent code from 2015 that was contributed by ESI-OpenCFD, as stated at openfoam.com:
   

   09/2015 — OpenCFD transfer copyright of its development repository to OpenFOAM Foundation Ltd
   22/09/2015 - OpenCFD publish its development line as the OpenFOAM-history repository

   but was not fully integrated, due to logistical issues...
   • And a bit of a nightmare to me, because I did want to integrate them all, but was overwhelmed by how much time it would have taken me to do it and who was it that was going to maintain it afterwards?
6. Given that both can use the OpenFOAM trademark and they own copyright of their released code, that means that both produce and deliver OpenFOAM as software...
7. Technically, ESI-OpenCFD does deliver with their releases code that has copyright of other authors, as disclaimed in the source code... but those authors only own their own piece of the code, they cannot use the trademark as if they owned the software, i.e. they don't own "OpenFOAM".
8. Last but not least:
   

   Originally Posted by Ahmed Khattab

   and why openfoam.org states in their contributes list that
   

   • Did you know that the expression "Knowledge is power" is a lie by itself?
     • Knowledge alone does not bring you power, it only brings potential... power is only achieved when you work on all of the intermediate requirements needed to use said knowledge, for it to indeed become power.
     • However, knowledge is the first step to potentially reaching power, so it's sort of true and it makes it a lot easier for people to understand its benefits and convince them to study.
     • That said, how many times have I heard children and adults complain about mathematics being too hard and "why do I need it? I don't understand it, so I don't care for it"... even though "knowledge" in this world is as valuable or more than money or gold.
   • This to say that technically that sentence at openfoam.org is true, because the code that was contributed by ESI-OpenCFD to the OpenFOAM Foundation on that specific period that I quoted above from openfoam.com, was not delivered in a way to easily integrate and maintain as it was delivered. Further, that page: https://openfoam.org/dev/contributors/ - refers to "Current Contributors to OpenFOAM"... which within the context, refers to the code released/maintained at the OpenFOAM Foundation.
   • On the other hand, technically there are people working at/for ESI-OpenCFD that have/are contributing code to the OpenFOAM Foundation...

But maintaining open source software does cost money/time to do it and someone has to foot the bill. And we have not just one, but at least two producers of OpenFOAM that are doing just that, both by spending the time needed to raise the funds, as well as doing the work and contributing with their own free time as well! And we even have forks/variants of OpenFOAM, that go beyond what the base code provides...

That said, anyone can have access to the source code of both producers and can do anything with it, as long as it is within the bounds stipulated by the GPLv3, i.e.: don't go closing the source if you provide only the binaries to other people, or else we will deal with you... either with niceness or in a court of law if you don't comply.

So in a very loose interpretation of the whole legal mumbo jumbo: Anyone (almost) can own a copy of the source source and modify it to one's own content. And anyone (almost) can provide services without costing them a single penny in licenses (i.e. make money for free) by either:

1. Simply using the software and delivering the results to clients.
2. Providing training in how to use it.
3. Providing services in modifying it and/or even contributing directly to it.
4. And any other service that I can't think of right now...

However, "almost anyone" refers to the edge situations, such as:

• According to the GPLv3, those who have broken the license agreement will no longer be able to use it? I can't remember exactly, let me quote the license document: https://www.gnu.org/licenses/gpl.html - uh, simply read chapter 8 "Termination", because a pardon can be made if things start to be done properly...
• Any situation where workers are refrained from using open source software, due to contractual issues...
• And any other situation that I can't think of right now...

So, all of this to say that: from my own experience with software names, people don't really care about the strict name that a software piece has got. They will rather call it by the shortest name that most agree with and roll with it... if the name has more than N number of characters, the excess is lopped off without concern for trademarks and such.
For example: "Linux"

• Ubuntu, Debian, CentOS and hundreds of other Linux Distributions are usually simply referred to as Linux when talking with others, because it's the minimal moniker that people can understand each other with.
• The actual distribution name is only referred to when asked to name it.
• "Linux" is just the kernel itself, i.e. the minimal software needed to talk with the computer and all of its components.
• "GNU/Linux" was heavily defended by RMS (Richard M. Stallman) as the proper name of the software, mostly due to... uh, what was it again? I vaguely remember that it was due to license and due to the need to have additional software to actual use the computer, aside from the Kernel. Namely the GNU software stack: https://www.gnu.org/software/software.html - uh, then again:
  

  GNU is an operating system which is 100% free software.

I honestly preferred when ESI-OpenCFD was delivering their more recent versions of OpenFOAM under the name "OpenFOAM+", because it was OpenFOAM.org+com, but I guess that the majority of the human population is unable to care about details, so "OpenFOAM" is the single name shared by both producers, hence the importance of a trademark.


And mind you that what I originally wrote on the last sentence above, I had written "both projects" instead of "both producers", because to me I see their efforts for their source code repositories, the associated issue trackers and the massive effort they put into them and the benefits that both bring to the community.

I'm tired of the bickerness of the details about who owns what and how, because there is so much work to be done and that needs to be done, that I've grown too tired to care about who owns what exactly... get work done first, complain when you are able to do something better about it...

Best regards,
Bruno

PS: I've spent well over an hour writing this post... so I ask you: Was this time well spent? Or would it have been preferable if I had answered to 2-10 other people who asked questions here on the forum about issues with their cases?

[ErenC]

I just realised openfoam.com version has additional forces inside src/fvOptions/sources/derived, I tried to copy it to openfoam.org version and wmake, but It gaves absurd errors(yes I added _USER_ to make file). Is it possible to compile it(like, joule heating) on openfoam.org version?

[wyldckat]

Originally Posted by ErenC

I just realised openfoam.com version has additional forces inside src/fvOptions/sources/derived, I tried to copy it to openfoam.org version and wmake, but It gaves absurd errors(yes I added _USER_ to make file). Is it possible to compile it(like, joule heating) on openfoam.org version?

Quick answer: It depends on what you want to compile. If you only want a single class, then you might be able to adapt the code, although you might have to start working from an older version of OpenFOAM.com, in order to make your life a bit easier.

But the short answer is that if you really want to use certain new features from OpenFOAM.com, it will be a lot easier to simply switch to a version of theirs instead. They integrate the majority of the features from the OpenFOAM Foundation's developments.

[roenby]

Thanks for the comprehensive guide, Bruno.

A minor update: This line is no longer valid:

Access to the bug tracker and development line requires free registration and logging in at https://develop.openfoam.com

You can now access https://develop.openfoam.com directly without login.

Cheers,
Johan

[mm66]

Thank you very much Bruno for this very needed topic.

Just wanted to add another layer to the topic. I came across the cloud option by OpenFoam foundation (openfoam.org) but have not seen anything similar for that of ESI (openfoam.com). Moreover, in educational institutions, I have frequently seen the foundation version being used but not from ESI. Please correct me if I am wrong. If not, do you think if there is any specific reason for this?

Thanks

[HPE]

ESI-OpenCFD Ltd. does not provide a cloud service, but CFDDirect Ltd. does. It is not related to OF releases or their content.

>> Moreover, in educational institutions, I have frequently seen the foundation version being used but not from ESI.

Do you have an extensive statistics on this remark? How many institution have you observed?

[mm66]

Originally Posted by HPE

ESI-OpenCFD Ltd. does not provide a cloud service, but CFDDirect Ltd. does. It is not related to OF releases or their content.

>> Moreover, in educational institutions, I have frequently seen the foundation version being used but not from ESI.

Do you have an extensive statistics on this remark? How many institution have you observed?

Hi HPE,

My observations are not very extensive unfortunately (that's why I asked others to correct me if I was wrong ). Nevertheless, I have observed at least three institutions which were all using the foundation version...

Regards,
MJ

[HPE]

I see. I think the reason is that potential users or even novice users do not know there are three main variants of OpenFOAM.

[Tobermory]

This is an interesting case study of the problems that arise when a corporation enters the market for an OpenSource product. The corporation feels it has to raise barriers (grab the trademark, police its usage, formalise operations procedures etc - all common sense business practices) that can alienate the developers who made the product in the first place, and can distort the GNU-environment for the product.

In my mind, I am happier with the code development being managed and done by OpenFOAM.org, and that's the flavour that I use. But that's my opinion ... your mileage may differ, and that's perfectly fine.

Note that there are a lot of corporate customers for OpenCFD's version, and so
don't expect any merging of the versions any time soon.

Now, as per Wyldcat's advice - let's just get on with our lives!

[HPE]

I am truly sorry, I am just an ahole sometimes without any excuse... and become extremely stupid... realising what I have done in few tens of minutes later.

I behaved in a bad manner, and I do apologise you again. Please let me know if I can compensate my behaviour in some way.

And I highly appreciated your respectful response. If I were you, I would just swear at me.

Many thanks for your kindness.

[Tobermory]

<message no longer necessary and so deleted after a friendly discussion with HPE>

[Tobermory]

The more I use OpenFOAM, the more my conclusion is: use both! There are some great features in both distributions - why limit yourself to just one?

[elochana]

Every time I happen to re-install OpenFOAM, I find myself pondering on this very same dilemma. I then somehow settle for .com version just because it has that 'grad' operator symbol which makes it more familiar in remembrance of the college days' Fluid Mechanics/Aerodynamics courses.. It makes it feel that they are the 'real nerds' .

Good enough reason?

[mikethe1wheelnut]

Originally Posted by Tobermory

This is an interesting case study of the problems that arise when a corporation enters the market for an OpenSource product. The corporation feels it has to raise barriers (grab the trademark, police its usage, formalise operations procedures etc - all common sense business practices) that can alienate the developers who made the product in the first place, and can distort the GNU-environment for the product.

In my mind, I am happier with the code development being managed and done by OpenFOAM.org, and that's the flavour that I use. But that's my opinion ... your mileage may differ, and that's perfectly fine.

Note that there are a lot of corporate customers for OpenCFD's version, and so
don't expect any merging of the versions any time soon.

Now, as per Wyldcat's advice - let's just get on with our lives!

I completely appreciate the sentiment here, but it can also be a hard sell. Fresh new users coming in and wanting to get a job done quickly, if they can, will be drawn to the version that can both do what they need it to do, -and- is better documented. At this point in time, it looks like .com has tutorials that will get me going faster than those at .org.

[pbachant]

Originally Posted by Tobermory

The more I use OpenFOAM, the more my conclusion is: use both! There are some great features in both distributions - why limit yourself to just one?

Personally I find it annoying to install and manage multiple OpenFOAM environments. It would be better if there were a single core library that all entities contributed to, with the commercial or niche features available as extension libraries rather than a complete repackaging of all the basic low level stuff, plus a little more.

[lourencosm]

I've been using OF for 10 years, having used mostly the ".com" fork.
Recently, I started to question about 2 topics, regarding the "www.openfoam.org" and "www.openfoam.com". I'd like to hear from you all about this:

1) Credits:
it's very frequent to have citations/references in publications only to ".com" of projects that only use tools that were developed totally (or at least, mostly) by the foundation (".org").
Is it fair, just because you're using the ".com" fork?

Specially because an average user has no capacity/time to inspect the git-logs and understand who developed the solvers he's using. Even if you use some extras exclusive to ".com", such as boundary conditions, is it fair to only give credits to this fork? I think the interFoam solver is a pretty good example.
Look at the copyrights of interFoam.C header in ".com" :

Copyright (C) 2011-2017 OpenFOAM Foundation
Copyright (C) 2020 OpenCFD Ltd.

I assume thati, in my PhD thesis and associated publications, I used mostly interFoam from ".com", and only cited ".com", but I'm having serious regrets not giving ".org" the propers credits.

2 - New users go for "openfoam.org" - the foundation
Regarding new users that use basic tools present in the foundation release (and also in some forks). Why not recommending to start always by the foundation release (".org"), as they are the ones who invested, and keep developing, the OpenFoam structure?
And afterwards, users will go to forks naturally following their needs.

Maybe, I'm missing several important perspectives, maybe not. I think the users should question themselves when using forks if they are giving the proper credits to the foundation.

[Henning86]

Its complicated details see:

Original OpenFOAM author(s)?

[ancolli]

Hi everyone,

I would like the opinions of experienced users and developers regarding the directions that each version is following. I believe that this information can be valuable for inexperienced and new users.

I started using OpenFOAM (the .org version, I did not know about the 3 branches) some years ago. What was more attractive to me was that there were a lot of solvers available, and I could explore them, study them, select the ones most similar to my needs, and (with the huge help of CFD Online and many of you) modify them to suit my needs.

As far as I understand, the .org version is focusing on usability and maintainability, thus reducing the number of solvers by "merging" some of them into what is called modules (https://cfd.direct/openfoam/free-sof...dular-solvers/), while the .com version maintains and adds new solvers each year.

I continue to use the .org version up to version 9. However, as things are going, I think I will switch to the .com version because while studying a given solver in each new version of the .org, it is more difficult (at least for me) to understand what is going on. I have had to go back to older versions (many of them) and, by following the new commits, understand what was happening.

My concerns are the following:

a) Will the .com version maintain their policies, or will they follow what is being done by the .org version?

b) Regarding the .org version, I understand that the code is open, but with the reduction in the number of solvers each year, is it becoming increasingly difficult to follow and modify them in a simpler way (at least for me). In terms of practicality, does this mean that the .org version is gradually becoming more like a black box?

c) If a) and b) are somewhat true and not the result of my misinterpretation, does that mean that the .org version will be more suitable for a better user experience (like Comsol) without the user needing to know most of the details behind the code, but the code solving their needs? And does that mean that the .com version will be used for testing new models (knowing more in detail the hypotheses behind each model)? Like an industrial use for the former and scientific use for the latter?

I do not want anyone to be offended, as English is not my mother tongue, and I cannot follow the more than one million lines of code for each version, as many of you do.