After studying
http://starflakenights.net/crypto/blockreward/xmg.html, I think I understand it, but I have some some questions/remarks left.
As far as I understand, MAGI implemented a blockreward that depends a lot on optimum hashrate used. Interesting to see is that, when optimum hashrate is used the blockreward is 50 XMG, when for instance the hashrate is double from optimum, the miners are punished very hard, there blockreward will only be around 2 XMG.
In case with all other coins blockrewards are fixed. When miners use double from optimum hashrate, they will only see their ROI ( rendement on investment) halved. Where in case of MAGI, the blockreward is divided by 10 or maybe 50. Please correct me if I am wrong.
Almost right there. The block reward can go all the way down to zero (or "transaction fees only")
I try to think about the consequences of this system. Putting myself in a the spot of a miner first. A miner of course likes to mine the coin that brings him the best result. To estimate his ROI on MAGI, he cannot just calculate, because blockrewards differ a lot. He probably has to mine some period of time and then make his calculations. And then even this result gives him absolutely no indication regarding results in the future.
Absolutely, yes that is correct. I have my own thoughts about ROI and profitability etc of mining in general, which I do not share in public. It is linked inextricably with the spot price available on the coin exchange(s)
Mining is just one way of acquiring coins. Buying is another (and the easiest way). Trading coins for profit works as well. Participating in XMG Magi promotions is yet another way of acquiring coins. Staking coins in the wallet generates a modest return too. Then there are XMG faucets to grab the occasional drip from as well. Some are tempted to obtain coins by dishonest means ... innit.
Let me ask a question. When you go to work and earn some money, is that profit? I don't think so. Is it more profitable to go find a job with more pay? No it isn't, but it would be a better return on your investment. Now how about you decide to get a job with more pay and hand in your notice to quit with your current employer, only to find that they were planning on promoting you with better pay than the job you are now going to? Gutted? I hope you are happy in your new job lol
I think that the true state of affairs is that many miners will stick with a coin in the hope of a massive increase in spot price in the future. If you have ever dabbled with "penny shares" then you will know exactly what I mean. If you were lucky enough to acquire some Bitcoins when they were worth a fraction of a dollar, then I am already preaching to the converted here.
Looking through the eyes of someone who is interested in a 51% attack on MAGI. As far as I Always understood, a 51% is possible when someone someone has more than 50% of the hashes used on a network. Please, correct me if this is wrong. On first site a 51% attack on MAGI looks very simple, because of the variable blockrewards. When someone wants a 51% attack, he just puts in a lot of hashes. After some time the regular miners see that the hashrate is very high and the blockrewards are only peanuts. So a lot of miners will go away from the network and mine some other coin. Resulting in a situation where the guy who wants to do the 51% attack, has an even bigger percentage of the network.
Yes, that could happen. People are at liberty to do as they want. I would like the big miners to go away. I don't always get what I want. I keep plugging away nevertheless.
But on the other side, I read MAGI is only cpu. Very to hard to believe, that when all coins, starting as cpu, never could stay only cpu, that MAGI will be able to stay purely cpu-mining. If this is the case, I can imagine a 51% attack will be extremely hard ( or better almost impossible) to do.
People are at liberty to write their own code to run an XMG (M7M algorithm) miner on a GPU, or indeed any other computing device. The mining code is open source, so grab yourself a GPU compiler, get coding and knock yourself out. When everybody is at it then expect XMG block rewards to remain steadfastly close to zero, like now. It rapidly becomes a pointless exercise. Personally, I think it would be more entertaining to employ an army consisting of an infinite number of monkeys with an abacus each to see if they could solve the next block purely by chance (either that or to come up with the entire works of Shakespeare).
Happy hashing
Wow, I really had to laugh a lot. Thans for your excellent explanation. You convinced me to buy some MAGI, I have no clue about mining.
Then after some days, i'll mine with full speed again 
