My goodness, the trolls are thick tonight.

I've posted in other threads about the comparison between XMR and BBR's proof of work schemes.  The bottom line is that they're optimizing for fundamentally different things.  I have a hunch that in the long term, both will turn into DRAM bandwidth-limited currencies, but for very different reasons, but that's an attribute that places ASIC-accelerated approaches on a reasonably similar bar.  The GPUs will be faster than the CPUs.  The ASICs will be faster than the GPUs.  But the ratio will be less large than it is with Bitcoin.  In the short term, BBR will be relatively more GPU-friendly than XMR is, but I don't think that there's much difference for them with ASIC-based implementations.

The statement about the "seems to lack entropy" is as vague as AnonyMint's other attacks against the XMR proof of work for having weaknesses related to its use of AES.  Remember - he doesn't like either of them, if you're going to start dredging all this crap up.  I'm in the opposite boat - I suspect they're both fine for now, but I make that statement cautiously for both, with the understanding that they may need to be modified (in relatively small ways) at some point in the next few years.

There is one known "weakness" in the BBR proof-of-work that should be fixed at some point in the future, but it's similar in importance to the way AES is used in XMR:  Both might make a serious cryptographer nervous if the function was being used **as a hash function** that needed all of the strong properties of a hash function, but neither is overly-flawed as a *proof-of-work* function, barring future analysis.  The truth about them both is that they've both used well-studied cryptographic primitives (good) and changed the *inside* of them to achieve a goal of memory hardness.  It's always dangerous to poke around inside the algorithms, but proof-of-work leaves a lot more breathing room as far as we know -- this is still a fairly under-studied field.

All of this misquoting of the mythical man month is silly, and ignores a more important point:  1-2 developers can do great things.  So can 7.  So can a dozen -- that's Amazon's preferred team size, for example.  With all of these, the proof will be in the pudding, as measured by robustness and security, usability, efficiency, and important features.

Stop grasping at bull$#!#.  Go read Zoidberg's presentation about the pruning, as I'm about to do, since it seems to be the only bit of actually useful information added to this debate in the last day.  People have expressed skepticism about whether it can work, and others have said it's one of the two most important features that BBR has over the rest of CryptoNote -- now is the time to start figuring that out.