[development] Have you ever laughed fate in the face?

Dries Buytaert dries.buytaert at gmail.com
Tue Jul 25 08:25:46 UTC 2006


On 24 Jul 2006, at 17:53, Chad Phillips -- Apartment Lines wrote:

>
> On Jul 24, 2006, at 6:00 AM, Gabor Hojtsy wrote:
>
>> IMHO what Dries suggested with the status column is simple. status  
>> = 1 is already in queries. status = 0 means unpublished, another  
>> status can mean deleted. No need to modify queries (just review  
>> existing ones for the status queries).
>
> once again, let me reiterate that this is an insufficient solution-- 
> it only applies to nodes, and there are other forms of content that  
> need to be addressed--

The main candidates would be nodes, comments and users and they all  
have a status field today.  This should make 95% of the users happy.   
What else do you want to un-delete?

Is it worth to be able to un-delete, say, a path alias?  Loosing a  
path alias is not big of a data loss, and re-creating the path alias  
is nearly as fast as un-deleting it.  Of course, it would be nice if  
we could undelete path aliases so we can make that happen in  
incremental patches; just add a status field to the table, and add an  
undelete action to the operations column (or something).

> not to mention that afaik node dependent tables don't have a status  
> column, which could also cause problems.

I don't see why node-dependent tables need their own status field.   
They implicitly inherit the node's status value (through SQL joins).   
Can you think of a concrete example of where/when this would be a  
problem?  I can't think of any ...

--
Dries Buytaert  ::  http://www.buytaert.net/



More information about the development mailing list