[Bf-taskforce25] RNA properties sorting

vekoon vekoon at gmail.com
Wed Jan 28 02:37:17 CET 2009


Hi,
I just noticed that RNA properties get sorted alphabetically during 
makesrna preprocessing. Is there any technical reason for doing so?
The reason I'm asking is that a good automatic layout for RNA groups 
need a smart ordering of properties when these are fed in to the 
autolayout routine, as I explained in my original proposal's 
documentation. As of now I thought I was using definition order instead 
it just happened to match to alphabetic order.
In the process of implementing groups and autolayout I noticed that 
definition order would really be a decent choice in that this way you do 
the ordering only once (otherwise you'd have to decide on a specific 
order three times: when you define, when you sort and finally when you  
lay the properties out).
In the end I think this would also make sense for a possible API (and 
would make this API more flexible), where the plugin developer is left 
to the task of choosing the right order for the properties he defines. 
Otherwise what you're doing is forcing the developer to first define the 
properties he wants and then re-sort them with some kind of custom 
function because the system sorted them alphabetically by default.
So if there's no particular reason for alpha sorting maybe we could 
leave it up to definition order?


More information about the Bf-taskforce25 mailing list