On 13/01/26 at 12:17 +0000, Ian Jackson wrote:
> Simon Josefsson writes ("Re: Please don't stop using uscan (Re: Include git
> commit id and git tree id in *.changes files when uploading?)"):
> > Indeed. I think we all seem to share the same concerns, and everyone
> > try to establish some order to the mess we are in, and we all seem to
> > create slightly different approaches when establishing order, and then
> > try to argue that their approach is the most reasonable one.
>
> Sean and I are the only people who have a plan[0] which meets all of
> our needs. For exmaple:
>
> Our plsn is the only one which lets people build a Debian derivative
> based on Debian's git, readily traceable to the upstream git history.
>
> Our plan is the only one which will enable a user to get the source
> code for a Debian package with "git clone" and then build it! [1]
>
> Sean and I have the only plan which will enable everyone to engage
> with Debian without needing to deal with dscs. [2]
>
> Indeed, many of our critics don't even have these as *objectives*!
I was thinking about something along the same lines. Maybe it would be
useful to try to write down a table comparing end-to-end workflows, with
a defined list of features/objectives of those workflows.
For example, you mentioned "traceability to the upstream git history",
but "traceability to the upstream released tarballs" is also useful.
Lucas
