[Dbix-class] Decision time: which fork inherits the existing DBIx::Class namespace

[Jonathan Scott Duff]

On Fri, Dec 2, 2016 at 11:50 PM, Peter Rabbitson <rabbit+dbic at rabbit.us>
wrote:

> Greetings,
>

Greetings and salutations to you Peter  :-)


>
> As such, and despite protests[2], I am essentially obligated to
> kickstart a DBIx::Class fork free of "community bias" ( note:
> "community", which was clearly set apart from "user base" in the past
> months ).
>

Hmm.  This is unfortunate.  Even if the fork is for the good of the user
base,
there will still be confusion about which DBIx::Class is the "right" or
"best"
one from existing users and potential future users.  And what would be
available to help developers explain this fork to their managers
(especially if
the list of dependencies does change as mentioned below)?


>
> The distribution will be governed by the proven BDFL model, although a
> lot of the details will get formalized in a document supplied with the
> distribution. The effect of uncertainty in this area turned out to be
> too great not to address going forward.
>
>
> The name of the fork is not interesting - the important thing is that it
> will continue providing the DBIx::Class namespace ( via a novel method,
> with safe-by-default conflict-resolution at install time, very much
> *un*like the Alt convention ). This means that a user willing to
> "switch" will have to adjust nothing more than their list of dependencies.
>

I'd be interested in hearing more about how the fork would be providing the
DBIx::Class namespace safely.  If someone were to switch from one
DBIx::Class to the other, how difficult would that be?

With all that said the outstanding question to the user-base is: given
> that a fork is unavoidable, and given technical means allow both forks
> to coexist on CPAN you must come to an agreement and instruct the PAUSE
> admins which of the two (mildly incompatible dists) should `cpan
> DBIx::Class` be resolving to.
>

My vote would be for "cpan DBIx::Class" to install the team-led DBIx::Class.
I am swayed by pragmatism and the bus number arguments.

-Scott
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.scsys.co.uk/pipermail/dbix-class/attachments/20161203/b86f737c/attachment.htm>