[Dbix-class] Fwd: DBIx::Class patch: better error diagnostics

Marc Espie espie at nerim.net
Wed Dec 6 16:17:51 GMT 2006


On Wed, Dec 06, 2006 at 04:01:04PM +0000, Matt S Trout wrote:
> And immediately *BEFORE* that is -
> 
> The community can be found via:
> 
>   Mailing list: http://lists.rawmode.org/mailman/listinfo/dbix-class/
> 
>   SVN: http://dev.catalyst.perl.org/repos/bast/trunk/DBIx-Class/
> 
>   Wiki: http://dbix-class.shadowcatsystems.co.uk/
> 
>   IRC: irc.perl.org#dbix-class
> 

Seriously, that's not good enough, for at least two reasons.

There is ways too much information. Several lines of addresses, all
related to some big framework. The lone email address is more attractive.

Also, let me explain where I'm coming from. I expect you'll understand it,
since we're both busy people. ;-) 
I'm one of the guys working on the packages in OpenBSD. In this line 
of work, I routinely write corrections and bug-fixes to hundreds of programs.

I've noticed, over the recent years, that `the entry barrier' has gotten
worse.   In many cases, if you want to contribute in a casual way, you
now have to jump through hurdles, like subscribing to a mailing-list,
waiting for the email to come back, or subscribe to some mozilla mechanism,
on the web, or some such, and get all the stuff that goes with it: a new
password to remember, and possibly spam for months to come, when your
report gets handled.

I'm interested in making things work. If I have twice as many administrative
hurdles in the way, I will deal with half the number of problems...
(you must also realize that I'm not paid for that work, so as long as it
gives me pleasure, I will continue doing that. Administrative hurdles do
not give me pleasure).

As a result, when I read through documentation, I will go to the path of
least resistance... If I see an author's address, and a mailing-list, I will
choose the author.

In the case at hand, I'm interested enough in DBIx::Class (I'm not just
fixing it, I actually want to use it) that I'm willing to spend some time
to explain what's going on.

In my opinion, if you want people to contribute to the DBIx::Class
ml directly, you have to write, seriously that people should write directly
to DBIx::Class, and that no subscription will be required... some people
will make the effort to contribute anyways, some others will simply go
away or keep their patches for themselves.

Maybe I'm wrong, and nobody else thinks like I do. Feel free to ignore this,
I do not want to start a huge flamewar over it... ;-)



More information about the Dbix-class mailing list