[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Sponsorship requirements and copyright files

On Sun, 22 Mar 2009 02:47:04 +0000
Noah Slater <nslater@tumbolia.org> wrote:

> On Sat, Mar 21, 2009 at 03:58:34PM +0100, Joerg Jaspert wrote:
> > Honestly, if you cant deal with listing the Authors/(C) holders - dont
> > maintain a package. It is not much work to list them. (It might be a lot
> > of work using the "new" format, but noone *requires* this format, especially
> > not ftpmaster. It has *no* gain for us at all, we couldnt care less if
> > you use it or not).
> I resent the implication in this remark. 

Then reconsider the remark. The proposed format is more work for many
overworked maintainers, it presents no clear gain for those maintainers,
it overly complicates the file and file handling. There is no point
arguing about these facts, overworked maintainers have made their
feelings clear and no amount of bleating from those supporting the
proposal will now change their minds. Only a complete restart will be

I, for one, would rather see the entire proposal backported to revision
50 or thereabouts and left at that. I refuse to use later revisions as
a basis for my own packages and I will not sponsor packages that, in my
own view, overly complicate debian/copyright by using this proposal as
a template for their own packages. The proposal, as it stands, is
insane and anyone recommending it needs to review the reasons for
recommending such a grandiose waste of my time.

> The copyright proposal is not complex,
> not verbose, and does not require any extra information from developers. The
> only thing it does is to mandate a machine readable format, in a similar vein as
> debian/control, for whatever information you might have already been using.

No, the current proposal is overtly complex, unnecessarily verbose,
requires enormous amounts of extra time from maintainers, the proposal
itself still has no clear structure and is completely unusable. The
machine readable format is not human readable and bears no comparison
with the clarity of debian/control.

> This has clear advantages for being able to post-process, check, search, and
> navigate copyright information using whatever tools the community decides would
> be profitable.

It has many clear disadvantages in wasting maintainer's time,
unnecessarily complicating a free form file and making it harder for
humans (i.e. us, the ones who need to understand copyright data in the
package) to read the data. Machine readable copyright data is not the
aim in and of itself. The aim has to be to make it easier to maintainers
to maintain debian/copyright and easier to build tools that support such
efforts. The current proposal makes that HARDER, not easier and it is
not surprising that no such tools have been devised. IMNSHO, the only
logical step for the progression of the current proposal is the
conversion of debian/copyright to a binary file that needs a parser to
be made human-readable as that illustrates just how insane the proposal
has become.

The proposal has been drowned in pointless edits and is unworthy of
further consideration, IMHO. Either backport to a point where the idea
is once again deemed sane by those maintainers who would most benefit
from tools to support updates of debian/copyright or abandon the entire
proposal as a good idea gone bad.

Maybe someone else can look at it after Squeeze and raise version 2.0
from the ashes.


Neil Williams

Attachment: pgpe5Ezs6qRnU.pgp
Description: PGP signature

Reply to: