[standards-jig] JEP 0038 ((Emot)Icon Styles) Updated

Adam Theo theo at theoretic.com
Thu Nov 7 18:35:53 UTC 2002

Well, the point is that the "ransom picking" only happens if everything 
else fails. That is, if there is no single match because either there 
are multiple text strings with the same language that matches what the 
user or sender is using. At that point, it doesn't matter what text 
string is randomly chosen because one choice is really just as good as 
any other. And also, keep in mind that those rules are just guidelines. 
Your implimentation can apply other rules to determine which one to use, 
such as asking the user in a pop-up (and then keeping track of it from 
then on), or pick the one that uses the least number of characters.

Michael Brown wrote:
>>I think you are right that the order in the icondef.xml file may not end 
>>up being the same order that is parsed by the client, but the point is 
>>that at that point, when all else fails, the client just picks the first 
>>one it comes across, no matter if it was the first in the icondef.xml 
>>file or not. Hope that makes sense.
> Umm....not really.  I don't want my client picking ASCII strings at random.
> Maybe an order tag that lists them in order of preference?  Something like:
> <icon>
>    <text order="1">:-)</text>
>    <text order="1">:)</text>
>    ...
> </icon>
> Or something.  (Anything!)

More information about the Standards mailing list