search

cernopendata / opendata.cern.ch

Source code for the CERN Open Data portal
http://opendata.cern.ch/
GNU General Public License v2.0
656 stars 147 forks source link

Licence for CMSSW still unclear #1232

Open RaoOfPhysics opened 7 years ago

RaoOfPhysics commented 7 years ago

cc: @tiborsimko, @katilp, @ekenn003

davidlange6 commented 7 years ago

On Mar 10, 2017, at 4:24 PM, RaoOfPhysics notifications@github.com wrote:

• The two CMSSW records on CODP mention the licence as being Apache License, Version 2.0: • http://opendata.cern.ch/record/221 • http://opendata.cern.ch/record/220 • However, the CMSSW repo on GitHub does not have a valid LICENSE file for the whole project: https://github.com/cms-sw/cmssw • Although, some sub-directories and files seem to be accompanied by licences: https://github.com/cms-sw/cmssw/search?utf8=%E2%9C%93&q=license (Some Apache, some GPL, etc.) • Question: Can the whole of CMSSW be licensed under Apache when some components are GPL? • When contacted in 2014-09, @davidlange6 said, “The CB has endorsed Apache License, Version 2.0 - but strictly following the requirements of that is not yet achieved [for CMSSW].”

and indeed, the context of that discussion was about analysis code examples and less about CMSSW as a whole.

As I mentioned at the time, in addition to the code license questions (which are still not solved), there are additional issues if you really mean to license a VM with libraries of CMSSW+external dependences.

• When I mentioned this in #71, it may have been interpreted as CMSSW having a confirmed licence, which isn't the case. • How should we proceed? At the moment, those directories/files without licences technically have all rights reserved -- you, you can view the source (“open source“) but they are not free to use/modify. • We need confirmation of the licence of CMSSW as a whole. cc: @tiborsimko, @katilp, @ekenn003

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.

katilp commented 7 years ago

@davidlange6 : @RaoOfPhysics refers here to a proposal that was discussed and approved at the CB meeting, independently of the Open Data, and to my understanding it is intended to CMSSW as a whole

https://indico.cern.ch/event/143964/ (CMS internal) https://indico.cern.ch/event/143964/contributions/1378788/attachments/133233/189074/LicensingCB77.pdf (CMS internal) (In the minutes: "8. The CB approved the proposal for Copyright and licensing of CMSSW software" )

davidlange6 commented 7 years ago

Right- thats for the code (and while proposed and discussed, the implementation of the specifics has unfortunately not been completed as I said)

On Mar 10, 2017, at 9:13 PM, Kati Lassila-Perini notifications@github.com wrote:

@davidlange6 : @RaoOfPhysics refers here to a proposal that was discussed and approved at the CB meeting, independently from the Open Data, and to my understanding it is intended to CMSSW as a whole

https://indico.cern.ch/event/143964/ (CMS internal) https://indico.cern.ch/event/143964/contributions/1378788/attachments/133233/189074/LicensingCB77.pdf (CMS internal) (In the minutes: "8. The CB approved the proposal for Copyright and licensing of CMSSW software" )

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.

davidlange6 commented 7 years ago

In any case, "CMSSW as a whole" means different things to different people. Its good to be more specific.

On Mar 10, 2017, at 9:13 PM, Kati Lassila-Perini notifications@github.com wrote:

@davidlange6 : @RaoOfPhysics refers here to a proposal that was discussed and approved at the CB meeting, independently from the Open Data, and to my understanding it is intended to CMSSW as a whole

https://indico.cern.ch/event/143964/ (CMS internal) https://indico.cern.ch/event/143964/contributions/1378788/attachments/133233/189074/LicensingCB77.pdf (CMS internal) (In the minutes: "8. The CB approved the proposal for Copyright and licensing of CMSSW software" )

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.

RaoOfPhysics commented 7 years ago

@davidlange6: Thanks for joining the discussion! :)

In any case, "CMSSW as a whole" means different things to different people. Its good to be more specific.

Indeed, this issue was triggered by the fact that as part of the open-data efforts, we have also been pointing to the relevant versions of CMSSW from the CERN Open Data Portal (records 220 and 221, as indicated in the OP). There, we erroneously state that the licence is Apache. Should we therefore simply remove reference of a licence or do we say "components licensed modularly"?

I suppose part of the concern is that while we can point to CMSSW we cannot in good faith ask users to use the software (let alone re-use/modify it) in the absence of some formal statement permitting them to do so (through some licence). Perhaps @tiborsimko or @TimSmithCH can chime in?

[…] there are additional issues if you really mean to license a VM with libraries of CMSSW+external dependences.

This will also need to be addressed if we are to adhere to the standards/practices of open research.

davidlange6 commented 7 years ago

On Mar 13, 2017, at 9:22 AM, RaoOfPhysics notifications@github.com wrote:

@davidlange6: Thanks for joining the discussion! :)

In any case, "CMSSW as a whole" means different things to different people. Its good to be more specific.

Indeed, this issue was triggered by the fact that as part of the open-data efforts, we have also been pointing to the relevant versions of CMSSW from the CERN Open Data Portal (records 220 and 221, as indicated in the OP). There, we erroneously state that the licence is Apache. Should we therefore simply remove reference of a licence or do we say "components licensed modularly"?

again, what is CMSSW and a "version of CMSSW" precisely?

I suppose part of the concern is that while we can point to CMSSW we cannot in good faith ask users to use the software (let alone re-use/modify it) in the absence of some formal statement permitting them to do so (through some licence). Perhaps @tiborsimko or @TimSmithCH can chime in?

[…] there are additional issues if you really mean to license a VM with libraries of CMSSW+external dependences.

This will also need to be addressed if we are to adhere to the standards/practices of open research.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.

katilp commented 7 years ago

@davidlange6 In my simple-minded view: for the versions we recommend with the open data, CMSSW is the code that can be found in the repositories: https://github.com/cms-sw/cmssw/tree/CMSSW_4_2_X https://github.com/cms-sw/cmssw/tree/CMSSW_5_3_X

davidlange6 commented 7 years ago

so you mean code in the CMSSW repository. There (afaik) we intend (and desire) to have an apache-like license but have not gotten there yet.

katilp commented 5 years ago

The CMS collaboration board approved at the CB meeting on February 8, 2019:

Adoption of Apache-2 license for software developed in CMS, unless other licenses apply