The case Peter mentions was one where the (String) value of the constant 
was correct (and did not change). However the (public final static) 
variable name of the constant was misspelled. So we added a new variable 
with the correctly spelled name having the same value while deprecating 
the misspelled variable.
-- 

BJ Hargrave
Senior Technical Staff Member, IBM
OSGi Fellow and CTO of the OSGi Alliance
[EMAIL PROTECTED]

office: +1 386 848 1781
mobile: +1 386 848 3788




From:
Peter Kriens <[EMAIL PROTECTED]>
To:
Niclas Hedhman <[EMAIL PROTECTED]>
Cc:
OSGi Developer Mail List <[email protected]>
Date:
2007-11-01 11:36
Subject:
Re: [osgi-dev] Resolution policy for code vs spec mismatch



The spec overrides the code. However, we had one such case where we
deprected the old constant and just added the correct constant. Not
nice, but the best we can do.

Kind regards,

     Peter Kriens
 
NH> Hi,
NH> I was wondering what the resolution would be if the specification says 
one
NH> thing and the commonly distributed sources/jars used by framework 
NH> implementations isn't according to the spec.

NH> Note: This is a hypothetical case, if we (God forbid) end up in such 
NH> scenarion, I just want to know what is authorative, since such 
scenario would
NH> break compatibility.

NH> Cheers


-- 
Peter Kriens                              Tel +33467542167
9C, Avenue St. Drézéry                    AOL,Yahoo: pkriens
34160 Beaulieu, France                    ICQ 255570717
Skype pkriens                             Fax +1 8153772599

_______________________________________________
OSGi Developer Mail List
[email protected]
http://www2.osgi.org/mailman/listinfo/osgi-dev



_______________________________________________
OSGi Developer Mail List
[email protected]
http://www2.osgi.org/mailman/listinfo/osgi-dev

Reply via email to