Talked to roomservice - I felt a copy/paste would be better than me trying to re-explain it
<roomservice> the problem is that people don't understand how tx fee payout on dgm work
<roomservice> quote meni:
<roomservice> The correct way is much simpler, it's to use the block reward of the share itself. That is, the share is a hash of a block header you hand out for the getwork, which has a merkle root of a transaction list with a given total transaction fee, which together with generation is the total block reward. That's the value of B which should be used for this share.
<roomservice> link: https://bitcointalk.org/index.php?topic=39497.msg573349#msg573349
<roomservice> did you get it?
<roomservice> let me explain in easy words
<Graet> ahh yes
<roomservice> if you submit a share right now
<Graet> we caculate the actual share reward at the time
<roomservice> you will be credited for block reward + tx fee if a block is found now
<Graet> but total block reward may be different
<roomservice> what pool finds is another story yes
<roomservice> it could be possible that the pool didn't find this 150btc block
<roomservice> and would have credited tx fees nevertheless
<roomservice> we even credit tx fees right now
<roomservice> even if we don't find the next block
<roomservice> so in this case you were really lucky
<Graet> can I quote this conversation?
<Graet> heh yeah
<roomservice> people who submitted shares a few minutes before we found this block should see a really nice increase of score
<roomservice> but it won't be a REALLY big difference - i guess most people don't even realize it
So we calculate the value of the share +txn fee at the time it is submitted not after the block is found. Which looks to be fine if there is not much variation in txn fees. On this occasion with an extremely high (unusual ) reward I feel this has not worked out the way it was intended.
It is late, will PM meni and sleep on it I think