On Dec 1, 2:38 pm, Gary Doades <[email protected]> wrote:
> On 01/12/2010 10:11 PM, Jeremy Evans wrote:> On Dec 1, 1:56 pm, Gary 
> Doades<[email protected]>  wrote:
> > It is the intended behavior.  It's how the master/slave database
> > support works without any configuration other than setting up the
> > database connection.  If you think it should be documented somewhere,
> > I'll certainly consider a patch.
>
> I think it just needs stating that if you override the :servers_hash
> option then you MUST add a valid :read_only entry to the servers hash.
> That seems to do the trick.

Except that that isn't the case.  For example, consider the case where
you override the :servers_hash option with Hash.new(:bev1234).

Like I said, if you'd like some documentation added/changed, please
submit a *patch*.  If it's not important enough for you to submit even
a simple documentation patch, it's probably not worth modifying the
documentation in the first place.

> Clearly without the :servers_hash option you don't need to do that, or
> at least I never did before when only using the :servers option.
>
> Maybe if the call to Sequel.connect (with a :servers_hash override)
> doesn't include a :read_only entry you can add one that is the same as
> :default. That seems to adhere to the principle of least surprise?

I don't think this is a good idea.

Jeremy

-- 
You received this message because you are subscribed to the Google Groups 
"sequel-talk" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/sequel-talk?hl=en.

Reply via email to