Thing is, GroestlCoin is actually deviating from the common standard, in that it (and few other new coins) is not using sha256d for the transaction hashes, but sha256 (single pass). To accomodate GroestlCoin, NOMP and sph-sgminer actually threat it as special case. In summary, it was never the new DMD wallet's fault that stratum mining did not work -- it was how that software assumed that when you said you want "groestl" you also said you want sha256 (single).
NOMP already has a setting to fix this, thanks to some discussion with the author. For coins that use any of (currently) keccak, blake, fugue and groestl -- you need to specify
in the coin definition.
By the way, the morale of this story is that the DMD team has been blamed without guilt on this subject.
This is what I'm harping on. This is another "now you tell us" example that comes to us in the form of a BTW, and
that should have been released in headline format all by itself. Of course, if you're not releasing all pertinent data in a timely and transparent fashion, it's a little difficult to do that.
This is all open source after all, correct? And anybody who really wants to know can always dive into the code, correct?
Again, I'm scratching my head and wondering out loud about the lack of specific data regarding all the issues DMD has been facing (both here and on github, and dmdcoin.net too).
I'm befuddled.
Add:
With all the previous talk about "hacking" things to adapt Diamond that created the impression that Diamond was deviating from the common standard, it's important to know that Diamond is not deviating from the original algo, at least in this case anyway. IMHO, these things need to be prominently reported in periodic status reports.