On 26/08/2010, at 9:26 PM, Erwan de FERRIERES wrote:

> Le 26/08/2010 10:19, Scott Gray a écrit :
>> Hi Erwan,
>> 
>> Why the unnecessary changes?  All that was needed to change was fix the 
>> ConditionObject constructor.  It should have been checking for "field" and 
>> then falling back to "field-name" (for backwards compatibility reasons), 
>> there was no need to change the schemas and I think the previous attribute 
>> name was more consistent.
>> 
>> And I definitely don't agree with backporting unnecessary schema changes to 
>> 10.04.
>> 
>> Regards
>> Scott
>> 
>> HotWax Media
>> http://www.hotwaxmedia.com
>> 
> Hi Scott,
> 
> no problem for reverting on the 10.04, it's on its way. But for trunk, I 
> really think the new names are more consistent than the previous ones. Once 
> again, if this is not ok with you, then I will look at another way to make 
> this work.
> 
> Cheers,
> -- 
> Erwan de FERRIERES
> www.nereide.biz

Thanks Erwan,

About the attribute name, there was effort put in not so long ago (maybe 2 
years?) to get the attribute names more consistent.  One of those efforts was 
to rename all attributes that expect a context field name to "field".
This attribute was changed at that time but a bug was introduced because the 
underlying code wasn't updated to reflect the change.  What is it about 
"object-name" that you feel is more consistent than the "field" attribute which 
is used heavily?

In the simple-methods schema there are currently no attributes called 
"object-name", although there are two named "event-request-object-name" and 
"event-response-object-name" (I'm not sure what they do).
For the "field", we have 35 attribute definitions.

Regards
Scott

Attachment: smime.p7s
Description: S/MIME cryptographic signature

Reply via email to