corinthia-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From jan i <>
Subject Re: Copyright notices
Date Fri, 13 Feb 2015 09:42:20 GMT
On 13 February 2015 at 10:25, Peter Kelly <> wrote:

> > On 13 Feb 2015, at 3:32 pm, Dennis E. Hamilton <>
> wrote:
> >
> > Oops, let this sit in my draft folder for hours. I blame the wonders of
> oxycodone and pain management.
> >
> > The TL;DR: You're doing great on the notices.
> >
> > Some nuances and clarification of NOTICE below.
> >
> > - Dennis
> >
> > Peter,
> >
> > Thanks for asking about this.
> >
> > I think the new notices on files are just fine (and COR-41 is totally
> elective but valuable as a casual undertaking).  The probably
> needs an ASF notice too.
> >
> > Some ASF purists expect that NOTICE and LICENSE will appear without a
> .txt extension.  I don't expect any/much push-back about that on a release,
> and if there were, it is probably something that could be fixed on a
> following release.
> My gut instinct would be to go without the extension. I noticed there was
> already a LICENSE.txt there (which may have actually been my addition way
> back when) hence I used that. But the instructions did say NOTICE (not
> NOTICE.txt), so i propose we use the former.
> > After a few years watching the lists about the proper use of these
> files, I think you are fine except for what might be needed for external
> dependencies, etc.  That can be dealt with as platform and
> external-dependencies (including for incorporated source code) are handled.
> >
> > It would be useful to obtain an appraisal from the mentors on this.  I'm
> confident that we have enough information to avoid marching over to
> discuss-legal and general-incubator to hammer anything out.
> >
> > We will need to run RAT on the repository as part of diligence with
> regard to third-party license notices and claims, and that should be done
> before declaring every release candidate.  I think this can be done on an
> unzip of the source archive for a given release, since there is more time
> to clean up IP on unreleased code/dependencies.
> One other dependency which I think should also be mentioned that I don’t
> think has been discussed the list or JIRA yet is showdown.js inside the
> Editor (inside the Editor/src/3rdparty directory). This is a Markdown
> implementation written in Javascript; there’s a license.txt in there which
> is the same as the original markdown program by John Gruber, which I think
> is BSD-style.
clearly something we should mention.

> This library is used when the paste command is executed in the editor and
> the clipboard contains only plain text. It converts the text into HTML,
> assuming Markdown syntax, and then goes through the normal paste codepath
> which works with HTML input.
> I’m not sure if we necessarily need this dependency in Corinthia. UX Write
> uses it but it’s not a fundamental requirement; it could instead be moved
> out into the application level (that is, UX Write itself and any other
> applications build on top of the Editor library). When we have Markdown
> support in DocFormats itself, an application that uses Corinthia could use
> DocFormats to do the conversion to HTML on paste.
We need to think a bit more about how to have the editor working, I am
thinking that it might be split in a couple of atoms (editing functions,
load/save functions, UI).

> >
> > It has been made very clear that NOTICE is not an attribution or
> acknowledgment file.  It must be limited to *legally-required* notice
> information.  Moving your copyright notice there is perfect under the
> third-party rules.  Any associated licenses that are required to be
> included are appended to LICENSE, and multiple uses of common licenses only
> needs to appear in LICENSE once.
> >
> > (Aside: My inclination would be to include the git commit hash for the
> code as it was before you made the contribution cited in NOTICE, but I
> don't think that is a requirement.  Anyone willing to do some forensic work
> in the git can find that point on their own, a nice feature of Git having
> all history in each clone.)
> Yeah the question popped into my head “how does who owns copyright over
> which part of the code?” and I realised Git history is the answer. Even
> with the initial commit hash in place, this serves only for UX
> Productivity’s contribution, and then only the contribution that was made
> at the point of initial commit. Any new code that is added by myself or
> others still needs to be identifiable as being under the copyright
> ownership of the companies/individuals who wrote it (in case of any
> disputes) - something which in non-Apache projects can often be done by
> looking at the names at the top of the source files, though with the Apache
> approach relies on version control. But I don’t see that as a problem given
> how easy it is to track the history of code through VC systems

Please do not make this more complex than it already is. One line in the
NOTICE file is what ASF needs, do not start to split it into who did what
before the code entered incubator.

Any code added after the code entered incubator is pr definition from the
start of the repo, so ASF has everything it needs.

> .
> >
> > I believe all of my contributions of any substance came after the move
> to the incubator, although I was made a member of the UX project before
> that.  In any case, I have not applied any copyright notices to files from
> me (except ASF notices) and I am in complete accord with the code being
> licensed to the ASF.  There is no requirement to do anything in NOTICE on
> my behalf.  Having a CONTRIBUTORS file would be valuable though.
> If you write new code though, you still own the copyright on it, correct?
> An example would be the scripts you wrote for externals required by the
> windows build - you own these but have granted a license for their use to
> Apache, is that right?
That is nearly correct. The copyright is Apache (see the ICLA) but you have
the full right to use it. The ICLA transfers ownership to Apache, this is
so that ASF can protect you in case of problems, if you owned the code ASF
could not protect you, and ASF could not license it to others (the essence
of ALv2).

jan i.

> --
> Dr. Peter M. Kelly
> PGP key: <>
> (fingerprint 5435 6718 59F0 DD1F BFA0 5E46 2523 BAA1 44AE 2966)

  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message