[20:13] <saiarcot895> Would adding multiarch support for a library qualify for a SRU?
[20:15] <cjwatson> That would normally be pretty dubious, because adding multiarch support to libraries has often been known to break things further up the stack (normally things that are misbehaving in some way, but even so).  Unless it was quite an obscure library, I think we'd require a fairly high standard of proof.
[20:16] <saiarcot895> cjwatson: True. Would the same thing apply if the maintainer thought he was adding multiarch support, but didn't really add support?
[20:16] <cjwatson> Err, SRUs should normally be cherry-picks of selected fixes, not wholesale backports of updated packaging.
[20:17] <cjwatson> If I saw an upload by that description, I'd probably reject it on the grounds that it clearly wasn't a minimal fix for whatever it was trying to fix.
[20:18] <saiarcot895> And I wouldn't blame you.