5:11 p.m.
Hi folks,
I'm packaging tpcclib (https://bugzilla.redhat.com/show_bug.cgi?id=1832562)
and as per the review, I wanted to confirm the licence for tpcclib (
https://gitlab.utu.fi/vesoik/tpcclib/-/blob/master/license.md)
Should it be "GPLv2+ and GPLv3+ "or something else?
Thanks,
11:33 a.m.
On Fri, May 08, 2020 at 03:41:36AM +0530, Purusharth Saxena wrote:
Hi folks,
I'm packaging tpcclib (https://bugzilla.redhat.com/show_bug.cgi?id=1832562)
and as per the review, I wanted to confirm the licence for tpcclib (
https://gitlab.utu.fi/vesoik/tpcclib/-/blob/master/license.md)
Should it be "GPLv2+ and GPLv3+ "or something else?
The copying.md file includes this:
"This program library is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by the Free
Software Foundation; either version 3 of the License, or (at your option) any
later version."
The '+' on the GPLv3+ means "GPL version 3 or any later version.
It's also a good idea to check for license text in individual files in the
project. For GPL projects, I like to do this:
find . -type f | xargs grep "General Public"
Which does a more or less ok job of finding files with what is probably a GPL
boilerplate. That gives me 39 files. Now, that's all files including
non-source. But in this case I am looking for any file that would indicate
something other than GPLv3+ Further refining:
find . -type f | xargs grep "General Public" | \
cut -d ':' -f 1 | sort | uniq
Gives me 10 files. I can do this:
find . -type f | xargs grep "General Public" | \
cut -d ':' -f 1 | sort | uniq \
xargs grep -i "any later version"
And see it matches 6 files. So 4 of those original files found lack the same
kind of boilerplate. Running the previous command and comparing it to what
was found, I see the sounds files in v1/ and v2/ were left out. Specifically:
v1/tools/htmlfind.c.bak
v2/libtpcabss/temp/algtimes.c
v2/libtpcabss/temp/blotimes.c
v2/libtpcabss/temp/scanditronics.c
Forget for a moment that there's a ".c.bak" file. That's odd to find
in
version control, but we'll look at it anyway. I am manually inspecting each
of those files with less(1) and I find:
v1/tools/htmlfind.c.bak: GPLv2+
v2/libtpcabss/temp/algtimes.c: GPLv2+
v2/libtpcabss/temp/blotimes.c: GPLv2+
v2/libtpcabss/temp/scanditronics.c: GPLv2+
So these files carry a GPL version 2.0 boilerplate. I am not sure if
libtpcabss is built as a separate shared library or not. And I also don't
know if all of the code it links in carries GPLv2+ boilerplates. My guess is
no, but you should verify.
Since the project is GPLv3+ and it incorporates GPLv2+ code in some capacity,
that makes the entire work GPLv3+ from our distribution standpoint. Unless
the GPLv2+ files are compiled and linked in to a standalone library with no
other GPLv3+ files, then you can't say there is some GPLv2+ component in the
project.
I would say the License for this package is GPLv3+
NOTE: The package maintainer is responsible for understanding the licenses
that apply to the source in the package. You should do an inspection of each
source file to see if there are other non-GPL license headers and how those
are compiled in to binary objects. The entire work looks to be a derived
GPLv3+ work, but it may have some shared libraries that are GPLv2+ or BSD or
MIT.
Thanks,
--
David Cantrell <dcantrell(a)redhat.com>
Red Hat, Inc. | Boston, MA | EST5EDT
12:11 p.m.
On 08/05/2020 17:33, David Cantrell wrote:
On Fri, May 08, 2020 at 03:41:36AM +0530, Purusharth Saxena wrote:
> Hi folks,
>
> I'm packaging tpcclib
> (https://bugzilla.redhat.com/show_bug.cgi?id=1832562)
> and as per the review, I wanted to confirm the licence for tpcclib (
> https://gitlab.utu.fi/vesoik/tpcclib/-/blob/master/license.md)
> Should it be "GPLv2+ and GPLv3+ "or something else?
The copying.md file includes this:
"This program library is free software; you can redistribute it and/or
modify
it under the terms of the GNU General Public License as published by the
Free
Software Foundation; either version 3 of the License, or (at your
option) any
later version."
The '+' on the GPLv3+ means "GPL version 3 or any later version.
It's also a good idea to check for license text in individual files in the
project. For GPL projects, I like to do this:
find . -type f | xargs grep "General Public"
Which does a more or less ok job of finding files with what is probably
a GPL
boilerplate. That gives me 39 files. Now, that's all files including
non-source. But in this case I am looking for any file that would indicate
something other than GPLv3+ Further refining:
find . -type f | xargs grep "General Public" | \
cut -d ':' -f 1 | sort | uniq
Gives me 10 files. I can do this:
find . -type f | xargs grep "General Public" | \
cut -d ':' -f 1 | sort | uniq \
xargs grep -i "any later version"
And see it matches 6 files. So 4 of those original files found lack the
same
kind of boilerplate. Running the previous command and comparing it to what
was found, I see the sounds files in v1/ and v2/ were left out.
Side tip: I'm guessing you don't know about `grep -rl`, & maybe a little
`sort -u`? These could be much simplified as:
grep -rl "General Public"
grep -rl "General Public" | xargs grep -i "any later version"
(which I guess eliminates the `sort | uniq` step anyway).
--
J. Randall Owens | http://www.GhiaPet.net/
12:58 p.m.
On Fri, May 08, 2020 18:11:32 +0100, J. Randall Owens wrote:
On 08/05/2020 17:33, David Cantrell wrote:
Thank you for both your replies.
We know the breakdown of the licenses in the different files
(`licensecheck` provides us with it, and is run as part of Fedora
review). The query primarily is whether the licensing of the various
individual files makes the "effective" license to be GPLv3+ or if it
should be noted to be GPLv2+ and GPLv3+.
https://fedoraproject.org/wiki/Licensing:FAQ?rd=Licensing/FAQ#What_is_.22...
The licensecheck output from the review is also attached.
--
Thanks,
Regards,
Ankur Sinha "FranciscoD" (He / Him / His) |
https://fedoraproject.org/wiki/User:Ankursinha
Time zone: Europe/London
1:41 p.m.
On Fri, May 08, 2020 at 06:58:51PM +0100, Ankur Sinha wrote:
On Fri, May 08, 2020 18:11:32 +0100, J. Randall Owens wrote:
> On 08/05/2020 17:33, David Cantrell wrote:
Thank you for both your replies.
We know the breakdown of the licenses in the different files
(`licensecheck` provides us with it, and is run as part of Fedora
review). The query primarily is whether the licensing of the various
individual files makes the "effective" license to be GPLv3+ or if it
should be noted to be GPLv2+ and GPLv3+.
https://fedoraproject.org/wiki/Licensing:FAQ?rd=Licensing/FAQ#What_is_.22...
The licensecheck output from the review is also attached.
In this case the effective license would be GPLv3+ since the GPLv2+ code
becomes GPLv3+ when combined with GPLv3+ code.
I think you should use 'GPLv3+' in the License tag.
Thanks,
--
David Cantrell <dcantrell(a)redhat.com>
Red Hat, Inc. | Boston, MA | EST5EDT
1:47 p.m.
On Fri, May 08, 2020 at 06:11:32PM +0100, J. Randall Owens wrote:
On 08/05/2020 17:33, David Cantrell wrote:
> On Fri, May 08, 2020 at 03:41:36AM +0530, Purusharth Saxena wrote:
>> Hi folks,
>>
>> I'm packaging tpcclib
>> (https://bugzilla.redhat.com/show_bug.cgi?id=1832562)
>> and as per the review, I wanted to confirm the licence for tpcclib (
>> https://gitlab.utu.fi/vesoik/tpcclib/-/blob/master/license.md)
>> Should it be "GPLv2+ and GPLv3+ "or something else?
>
> The copying.md file includes this:
>
> "This program library is free software; you can redistribute it and/or
> modify
> it under the terms of the GNU General Public License as published by the
> Free
> Software Foundation; either version 3 of the License, or (at your
> option) any
> later version."
>
> The '+' on the GPLv3+ means "GPL version 3 or any later version.
>
> It's also a good idea to check for license text in individual files in the
> project. For GPL projects, I like to do this:
>
> find . -type f | xargs grep "General Public"
>
> Which does a more or less ok job of finding files with what is probably
> a GPL
> boilerplate. That gives me 39 files. Now, that's all files including
> non-source. But in this case I am looking for any file that would indicate
> something other than GPLv3+ Further refining:
>
> find . -type f | xargs grep "General Public" | \
> cut -d ':' -f 1 | sort | uniq
>
> Gives me 10 files. I can do this:
>
> find . -type f | xargs grep "General Public" | \
> cut -d ':' -f 1 | sort | uniq \
> xargs grep -i "any later version"
>
> And see it matches 6 files. So 4 of those original files found lack the
> same
> kind of boilerplate. Running the previous command and comparing it to what
> was found, I see the sounds files in v1/ and v2/ were left out.
Side tip: I'm guessing you don't know about `grep -rl`, & maybe a little
`sort -u`? These could be much simplified as:
grep -rl "General Public"
grep -rl "General Public" | xargs grep -i "any later version"
(which I guess eliminates the `sort | uniq` step anyway).
I do actually know about those options. They've just never made it to my
muscle memory. cut(1) is my hammer and stdin are nails. ¯\_(ツ)_/¯
In this case, my needlessly exhaustive and verbose command makes it clear to
the reader what I'm doing where condensing it with options could be somewhat
more confusing depending on the audience. (though cut is also cryptic to
anyone who hasn't used it...oh well)
Thanks,
--
David Cantrell <dcantrell(a)redhat.com>
Red Hat, Inc. | Boston, MA | EST5EDT
12:24 p.m.
Hey,
I confirmed it with upstream.
The files with GPL 2 license were actually not included in compilation and
are now removed. Most of the small files without license are test
data
which obviously cannot be edited to contain license, but I have omitted the
license from individual code files too, because I hope it would be
sufficient to have GPL3 in the main docs, covering the whole library.
The winpthreads.md needs to be included with Windows binaries, but that
library is not used in other platforms, and thus not relevant to this.
Should I go ahead with GPL3 then?
Regards,
Purusharth S.
On Sat, 9 May 2020 at 00:18, David Cantrell <dcantrell(a)redhat.com> wrote:
On Fri, May 08, 2020 at 06:11:32PM +0100, J. Randall Owens wrote:
>On 08/05/2020 17:33, David Cantrell wrote:
>> On Fri, May 08, 2020 at 03:41:36AM +0530, Purusharth Saxena wrote:
>>> Hi folks,
>>>
>>> I'm packaging tpcclib
>>> (https://bugzilla.redhat.com/show_bug.cgi?id=1832562)
>>> and as per the review, I wanted to confirm the licence for tpcclib (
>>> https://gitlab.utu.fi/vesoik/tpcclib/-/blob/master/license.md)
>>> Should it be "GPLv2+ and GPLv3+ "or something else?
>>
>> The copying.md file includes this:
>>
>> "This program library is free software; you can redistribute it and/or
>> modify
>> it under the terms of the GNU General Public License as published by the
>> Free
>> Software Foundation; either version 3 of the License, or (at your
>> option) any
>> later version."
>>
>> The '+' on the GPLv3+ means "GPL version 3 or any later version.
>>
>> It's also a good idea to check for license text in individual files in
the
>> project. For GPL projects, I like to do this:
>>
>> find . -type f | xargs grep "General Public"
>>
>> Which does a more or less ok job of finding files with what is probably
>> a GPL
>> boilerplate. That gives me 39 files. Now, that's all files including
>> non-source. But in this case I am looking for any file that would
indicate
>> something other than GPLv3+ Further refining:
>>
>> find . -type f | xargs grep "General Public" | \
>> cut -d ':' -f 1 | sort | uniq
>>
>> Gives me 10 files. I can do this:
>>
>> find . -type f | xargs grep "General Public" | \
>> cut -d ':' -f 1 | sort | uniq \
>> xargs grep -i "any later version"
>>
>> And see it matches 6 files. So 4 of those original files found lack the
>> same
>> kind of boilerplate. Running the previous command and comparing it to
what
>> was found, I see the sounds files in v1/ and v2/ were left out.
>
>Side tip: I'm guessing you don't know about `grep -rl`, & maybe a little
>`sort -u`? These could be much simplified as:
> grep -rl "General Public"
> grep -rl "General Public" | xargs grep -i "any later
version"
>(which I guess eliminates the `sort | uniq` step anyway).
I do actually know about those options. They've just never made it to my
muscle memory. cut(1) is my hammer and stdin are nails. ¯\_(ツ)_/¯
In this case, my needlessly exhaustive and verbose command makes it clear
to
the reader what I'm doing where condensing it with options could be
somewhat
more confusing depending on the audience. (though cut is also cryptic to
anyone who hasn't used it...oh well)
Thanks,
--
David Cantrell <dcantrell(a)redhat.com>
Red Hat, Inc. | Boston, MA | EST5EDT
_______________________________________________
devel mailing list -- devel(a)lists.fedoraproject.org
To unsubscribe send an email to devel-leave(a)lists.fedoraproject.org
Fedora Code of Conduct:
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives:
https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org
1434
days inactive
1443
days old
6 comments
4 participants
participants (4)
-
Ankur Sinha -
David Cantrell -
J. Randall Owens -
Purusharth Saxena