Re: Annouce: DBIx::Class, a.k.a "taking the easy way out"

[prev] [thread] [next] [Date index for 2005/07/27]

From: John Siracusa
Subject: Re: Annouce: DBIx::Class, a.k.a "taking the easy way out"
Date: 02:08 on 27 Jul 2005
On 7/26/05 6:35 PM, Sebastian Riedel wrote:
> Am 27.07.2005 um 00:03 schrieb Michael G Schwern:
>> On Tue, Jul 26, 2005 at 10:50:13PM +0200, Sebastian Riedel wrote:
>>>> Gah!  An @ISA copying and patching module?
>>>> 
>>>> You're already into scary hacks and things have barely started.
>>> 
>>> I think the only reason people are so scared about multiple
>>> inheritance is the fact that so few languages support it, if there
>>> were more common design patterns it would be all fine... :)
>> 
>> Given that I'm a native Perl programmer, have been using MI for as  long as
>> I've been using OO, have done scary @ISA hacks in the past and  have the
>> scorch marks to prove it, I have to call bullshit on the above.
> 
> I read that as "i am god, just trust me, i don't need no fsckin
> examples..."

The commonly accepted wisdom within the CS/language community is that MI is
rarely the best solution for a given problem.  Common wisdom is sometimes
wrong, of course, but I don't think anyone disagrees that the anti-MI
sentiment is the consensus today.

Given this environment, when MI is put forth as the best solution for a
particular problem, it's understandable that contemporaries will try to
dissuade the MI proponent.  In reality, one of a few things might be
happening.

It could be that, even according to conventional wisdom, this is one of the
rare cases where MI really is the best approach.

It could be that the MI proponent has discovered something that is not taken
into account by the conventional wisdom.

It could be that the environment has changed since the conventional wisdom
was arrived at, and that the new environment alters the merits of the
approach.

It could be that the MI proponent is going through a personal version of
what the community went through in order to arrive at its current concensus:
enthusiasm for MI, successful implementations, stability and maturity,
growing pains, and eventual roadblocks that cause a reassessment of the
approach.

Obviously each participant in this thread has a different opinion of which
possibility listed above is closest to the reality.  But regardless of the
truth of the matter, I think it'd benefit everyone to at least entertain the
alternate possibilities (however briefly :)

-John


(message missing)

Delegation vs Hooks (was: Annouce: DBIx::Class, a.k.a "taking the easy way out")
=?ISO-8859-1?Q?Ask_Bj=F8rn_Hansen?= 23:27 on 26 Jul 2005

Re: Annouce: DBIx::Class, a.k.a "taking the easy way out"
John Siracusa 02:08 on 27 Jul 2005

Generated at 16:36 on 28 Jul 2005 by mariachi v0.52