|
As to prior posts for PoS, thanks.
I might talk over some weakness. First off there are not many PoS coins on the top of the list -- a simple feel; there were some top coins but seem having dropped down. PoS gets value diluted in particular with high interest; PoS make people tend to accumulating -- seems good, but not quite; PoS could be easier to manipulate, and tend to be a P & D which can quite be done if one is able to do that. These issues wouldn't matter much if the market value is so huge that no single person is able to do that, but not good for young magi at this stage. And importantly, PoS means one must buy in order to engage into it.
Speaking of the security, there exist weakness in PoS too. People who hold a great deal of coins having the ability to launch an attack, by having offline stake days and coming online at a moment; people might not intentionally fire up an attack but simply a desire to get more PoS coins and they know how-to. The consequence might be an attack like, minimum - hard fork, that would be a severe incident enough to a coin. Again this is easy to be accomplished for a young coin. Such issues were talked over in NOBL when switched to PoS (and given the try to PoS-II then, to be exact). Magi's PoS-II did some tricks along that, but I won't be able to guarantee 0% risk at this time.
In my view, PoW & PoS is most secure, but feel free to discuss, I might be wrong.
Thanks for the prior PoW comments; I'd think & figure & talk again, it may take time as I am still limited to free time.
|
|
|
|
price is going down every days.... make me want to sell all my magi coin
The situation is not bad actually, though price rising seems everyone's dream and can be one of the ways call for wide attention. The coin is yet happy to the specs, and concerns need a care; all are just happening before on rocketing if it would be. |
|
|
|
You don't need to clone the algo - as a matter of fact, take the opportunity to tune the parameters of the KDF to your liking. While GPU miners for it are obviously possible (as for M7M), the gap in between CPU and GPU is at least currently barely existing - factoring in power makes it look even better to mine on CPU. This is about as close as you'll ever get to your stated goal - i.e. GPU mining impossible.
this makes sense for me, but what about all those buldozers that will mine it only to fu*k the block rewards, this will really solve it? i mean even if is more profitable with CPU after the algo change, they will loose hash power? The big miner things as to PoW is about GPU or CPU farm. I guess we're try to kill those mining with cheap by algo change, and people then mine at the same level (mostly). The Magi's block design is supposed to take care of the CPU farm thing (to a certain degree, don't want to say 100% without apparent proof  ). We can tune Yescrypt's params such that it'll trip more resource alarms on a farm - using up FAR more RAM. Thanks, I'll have a look at it and discuss with you. |
|
|
|
wow i can see here lot of good options in the last day, thats cool for me there is only one big problem that i already talk about, the buldozers, i hope you guys find a way to at least, minimize that. on the other hands, i would like to see a % increased on the pos rewards, for me to have more than 1k of magi and get a reward of 0.01xmg from stake is so low, but i will understand if you dont want to change it, i mean increased it. I guess it's dependent; by the current market, 0.01xmg by 1k magi staking values trivial, but would be not that low when magi is valued as much as btc, though the latter situation is very much no more than an imagination for now; imo it'll depend on how we keep the way going then and the overall view. |
|
|
|
You don't need to clone the algo - as a matter of fact, take the opportunity to tune the parameters of the KDF to your liking. I disagree. It will cause only confusion. People still confuses groestl^2 with groestl-myr. If it is called the same, it must be the same. But I'll agree the KDF of yescrypt has... some very interesting properties. Hmm nice to see the ideas. Think Magi has to look into those options. So there is option 1) M7M with PoS-III and more adjustments option 2) KDF option 3) yescrypt. Or are there more good options? How about option 4) Go PoS only. Add a block reward cap and other inflation control measures patterned after a coin like Hyperstake. Or option 5) PoS only with something new like have a percent of the PoS block reward be sent to a common fund to be used to promote the coin. Only a board of directors voted in by the community based on their holdings could have access to the common fund. We could use the common funds to make tee shirts, drink holders, pay for development work, entice new community members, bribe/lobby congressmen,  ... I am kinda like to be traditional, in way of btc or the like, and PoW is a need for that, but feel free to discuss on pure PoS. |
|
|
|
You don't need to clone the algo - as a matter of fact, take the opportunity to tune the parameters of the KDF to your liking. I disagree. It will cause only confusion. People still confuses groestl^2 with groestl-myr. If it is called the same, it must be the same. But I'll agree the KDF of yescrypt has... some very interesting properties. Hmm nice to see the ideas. Think Magi has to look into those options. So there is option 1) M7M with PoS-III and more adjustments option 2) KDF option 3) yescrypt. Or are there more good options? Hi all, my opinion is: good for option 1, ok for option 2 very bad for option 3 (i'm playing with yescypt from a long time...) I see a lot of additional idea, like anonymity (i2p,etc), different db for wallet speed up, masternode,instant transaction,etc. but change the algorithm into yescript maybe not the best deal. Yescrypt made for a different usage, with different machine, different power and not for CPU coin. But this is only my opinion. Regards Thanks mate, I have no opinions yet but have to look over. |
|
|
|
You don't need to clone the algo - as a matter of fact, take the opportunity to tune the parameters of the KDF to your liking. While GPU miners for it are obviously possible (as for M7M), the gap in between CPU and GPU is at least currently barely existing - factoring in power makes it look even better to mine on CPU. This is about as close as you'll ever get to your stated goal - i.e. GPU mining impossible.
this makes sense for me, but what about all those buldozers that will mine it only to fu*k the block rewards, this will really solve it? i mean even if is more profitable with CPU after the algo change, they will loose hash power? The big miner things as to PoW is about GPU or CPU farm. I guess we're try to kill those mining with cheap by algo change, and people then mine at the same level (mostly). The Magi's block design is supposed to take care of the CPU farm thing (to a certain degree, don't want to say 100% without apparent proof ). |
|
|
|
I think that if the stake rate is increased in a way that is consistent with the goals of Magi it could actually help reduce the current inflation from over PoW mining. This could be done by improving PoS-II into PoS-III, which should reward smaller holders the most. The stake rate should go up the most on small blocks. I think 10% for any block under 100 XMG wouldn't be too generous, while 100-500 XMG could stake at 7.5%, 500-2500 at 5%, and any block over 2500 at 2.5%. I get a lot of stakes from small blocks, and the reward is often 0.02 XMG for anything small, and around 0.15 XMG for a 2000 XMG block. This is much less than any PoW block, and could be safely increased without causing inflation. Being able to set the split threshold would also make staking much easier.
If we want Magi to stay a CPU only coin, changes to the algorithm should be made, if only as a preventative measure. Even if the miner that turns on 48 MH/s isn't using GPUs, that doesn't mean that one couldn't be developed in the future. If it can be done, it will if the profit motivation is there. I would prefer that changes were made to the existing algorithm to make it more GPU resistant, instead of basing it off of something completely different. Maybe M7M2? This should only be done after there is an optimized CPU miner that is ready to go so people only have to download an update to the Sweetspot miner, which would be easy for the common user.
The PoS-II approach was designed with similar feature, though not direct inverse rewarding route. I am sure I'll take a closer look systematically. A simple procedure for a big holder to take the advantages is to split his coins into small pieces, and as to each piece he'll gain equal shares as small holders. An optional approach is given "rules" and "efforts" are made barriers to the big holders and mitigate advantages gained by them in a way unusually. This is not straightforward, though PoS-II is supposed. You're right, everything needs to be ready before going, mostly launching testnet for the test. |
|
|
|
Joe - the PoW you want to use is probably the KDF Yescrypt. I'd wanna talk on IRC before recommending parameters to it, though.
Thanks for suggestion, Wolf. I'd look into Yescrypt. I read the IRC message, and will send message too; kinda busy during weekday, will find time for talk at some point. My concerns as to the current situation which seems in existence all along are some hidden facts. Like you mentioned before, gnu miner for magi algo is possible; if you say so, I guess so. Had review of past pages and also told by some people; the fact is that most are wondering the mystery big miners in and out. This IS a concern, and to make magi step forward better to be taken care. What I wish and probably most of people wish is that magi stays at a level that everyone can mine, that equally means CPU mining, and mining should be fair that also tells any other means of mining with advantages over CPU mining should be avoided. To correct (to who like other means of mining), this is not because I hate GPU miners or things like that, but just because magi is to be what it should be and that leads to a fact that those mining wouldn't do any good to the coin. In brief, solidified CPU mining, no others. From this point of view, Yescrypt seems an option. I am kinda avoiding simply algo clone though; so the thing gets me complex; either way, what happens now are simply open discussion. You don't need to clone the algo - as a matter of fact, take the opportunity to tune the parameters of the KDF to your liking. While GPU miners for it are obviously possible (as for M7M), the gap in between CPU and GPU is at least currently barely existing - factoring in power makes it look even better to mine on CPU. This is about as close as you'll ever get to your stated goal - i.e. GPU mining impossible. Thanks again for inputs. I'd check out details and think of things to tackle. Thanks for the link. |
|
|
|
Hey Joe.
IMHO. We need to implement masternodes to fight with the inflation. PoS % needs review too. I2P maybe ?
PoI is not necessary anymore.
Cannot understand why changing the algo? ( I'm not against this ) but what is the advantage?
This are known features.
Thought about master nodes, but didn't yet look into it. What is I2P in details? Algo thing mentioned in another post. Any reasoning why masternodes are terrible?
From my perspective if masternodes were added to Magi it would simply be a way for the rich to get richer. I think the same could be said for increasing the POS %. I am interested in developments that make the playing field even for people that want to get involved, which is a good way to attract new interest. Masternodes are pretty much the opposite in my view. Thanks for the input about master nodes. Exactly one of the disadvantges of PoS, rich getting richer; PoS-II is supposed to do something along that; it would be needed to take cautions to increase interest in a straight way. I believe people should be aware of that when generic gains shares easily, there exist people who can get more share much quicker than rest of us, like the PoW mining things. |
|
|
|
I think the servicenodes are very fashionable these days. Such implementation should prevent the inflation and if there is some king of additional tech through the network, the value will be much more reasonable.
I still think we dont need changing the algo , or if we proceed to that , if should remain CPU-only!
You're right, mate, try to stay CPU mining; I'd like to some more details of service nodes you mentioned. |
|
|
|
Joe - the PoW you want to use is probably the KDF Yescrypt. I'd wanna talk on IRC before recommending parameters to it, though.
Thanks for suggestion, Wolf. I'd look into Yescrypt. I read the IRC message, and will send message too; kinda busy during weekday, will find time for talk at some point. My concerns as to the current situation which seems in existence all along are some hidden facts. Like you mentioned before, gnu miner for magi algo is possible; if you say so, I guess so. Had review of past pages and also told by some people; the fact is that most are wondering the mystery big miners in and out. This IS a concern, and to make magi step forward better to be taken care. What I wish and probably most of people wish is that magi stays at a level that everyone can mine, that equally means CPU mining, and mining should be fair that also tells any other means of mining with advantages over CPU mining should be avoided. To correct (to who like other means of mining), this is not because I hate GPU miners or things like that, but just because magi is to be what it should be and that leads to a fact that those mining wouldn't do any good to the coin. In brief, solidified CPU mining, no others. From this point of view, Yescrypt seems an option. I am kinda avoiding simply algo clone though; so the thing gets me complex; either way, what happens now are simply open discussion. |
|
|
|
Lionheart78 nice to see you back here!
heya 111magic!!! Thanks  , XMG community is blessed to have a great dev and team of people like you to manage XMG and vice versa, it is always refreshing to visit your lively thread! More Power to XMG Dev and Community!!! Very true, nice family. |
|
|
|
I've always thought that around 10 percent would be a good rate for PoS but it is just a gut feel. A hard fork or algorithm change might cause issues with the exchanges and I'm not sure that would be worth the risk. I'd vote for adding setstakesplitthreshold and some feedback on coin weight to the advanced coin control dialog though.
I would not do any major hard fork unless there is a clear indication from the community – perhaps a vote could be setup for this?
Thanks for suggestions; got excited about many old faces around, and yeah, it won't be just-do-it, will look around carefully before taking the leap. setstakesplitthreshold thing talked before and will take it. Some thoughts about PoS made below. |
|
|
|
To all, It was of course a frustration that the issue took this long to be fixed; no room would be given for excuse. Instead of speaking these things (feel free to PM lionheart anyways), I want to issue the following fix to the block stuck which was at #257053 (edit). If your current block No. is less than #257053 (edit), download the following wallet to get over where stuck (win and mac os, respectively). Also you can download the block chain without syncing from scratch if this is the case; provided github source code is attached below as well. Github source code: https://github.com/ermis-project/ermisBlock chain: http://coinmagi.org/files/ermiscoin/ermiscoin-blockchain-257185.zipWallet for Windows: http://coinmagi.org/files/ermiscoin/ermiscoin-qt-v1.2.0.1-win.zipWallet of Mac os: http://coinmagi.org/files/ermiscoin/Ermiscoin-Qt-v1.2.0.1-macos.dmgBy the way, this is Joe from Magi - XMG, and would like to push ERM forward in a certain way, at minimum fixing the things. lionheart is my friend for more than 1 year, and I am quite sure he is a good person and persistent to the initial plans. As far as the wallet tech is concerned, his specialty is mostly in the business instead of dev. All of people have a real life and some life pain, and that did happen. And yet the team still has a strong wish to go ahead with this coin. lionheart does have the plan and has actually started his business in some form; what he is doing is exactly coincidence with the plan of ERM as planned, and I believed he will simply carry on his plan though real life sometime offsets the steps. In any case, I simply wish this coin proceeds, and also quite sure lionheart will take the role as done in the past. Regarding ERM itself, I very much like pull the PoS-II into ERM, with some features specifically designed for ERM. Specs would be announced later on. p.s., the block chain has moved few blocks after where stuck 257052; feel free to check the block chain, all clean and no hidden facts; let the team know per your suspicion. |
|
|
|
I wish to speedup wallet startup time. On slow atom cpu it is annoying. Maybe we can change DB format (like novacoin, dash).
And I like idea to some increase of POS reward.
Thanks for comments, will check out DB things and see if that improves startup time. Regarding PoS, what's interest you like; the APR is not that low actually by comparing with other coins, and the staking can be offset by PoS-II at some conditions. |
|
|
|
|
We talked about changing the algo, and/or going a smarter adjustment system regarding the block reward. Magi operates on the basis of dynamic adjustment; further optimization would be beneficial and that will surely mitigate in part the issue of big miners. Frankly the best effort in magi algo - M7M is no widely the best, and there always are smarters who tend to tackling. I heard quite a lot talking about the gnu miners.
Here I'd open up discussion about the algo switch strategy: anyone interested in a switch of algo? I'd rather go that way straightforwardly, as I have been already blamed when we changed to M7M-II; equally changing algo simply because of the suspicion is superficial and not a solid reason for the practical action. In any case, I hope people in this community can vote for changing or no changing, and if you want to change, what's algo you're up to?
There are a lot pain in switch; unless we can take care of algo properly and would no longer need to look back and forth, bringing a new algo could be less meaningful in the long run. Having all the pools switching is another pain. Provided that most of people say NO, we'll solidly head for taking measure and improving the block reward system.
|
|
|
|
thats pretty awesome, i was inactive at the irc when you have the meeting  but magic told me what you talk about, and im really think that some of these new features are simply needed and awesome. No problems, shoot a post/PM anytime. btw i didnt know if you guys already talk about it or not, but i was thinking on a new console command like other coins already have and is very usefull, i mean something like this: setstakesplitthreshold 4000, then that would mean your wallet won't split the blocks if it results in each block being less than 4k XMG. So in the example above, it would have produce one block of 6k and not split them. If you set the threshold to 2000, then it would have split them.
Yes, this is exactlly what we thought about. I'd work on it whenever I can. The major / biggest concern is the big miners yes this is a big problem but im sure you will find a solution This issue is essentially in the center of magi at this moment. I pull this question out below so that we can give it a focus. |
|
|
|
|
magic and I had a talk over IRC, a brief list:
* wallet issues as he received from few people
* concerns about PoS
* Many people lend help to magi -- thank you
* PoM-II / PoI -- will complete PoM payout and end this campaign soon
* Launching PoM/PoI as well in the future
* New code base, features and misc; all point to a thing, making magi to be something useful.
** The major / biggest concern is the big miners, while this is a dilemma, and certainly this is the top one in the list to be looked into and solved/improve in a timely fashion.
Almost mentioned all above, not a NEXT plan list though but concerns mostly; let this community know if you come across any interesting or concerns to you.
|
|
|
|
Think it would be good to add some items we would like to talk about on a list. That would also be good for the new roadmap Magi needs to add here.
If you have a specific questions you can post them here or send it pm to me.
Any thoughts on how to combat the whales that are mining and destroying the block reward or will this just stay like this? What about increasing the POS % to encourage more people to save in wallets? mmm i didnt think on it, i mean increased the POS%, but i start to think this will be a good idea, right now the pos % is very very low, maybe by increasing it more ppl will hold his coin and maybe will be more interesting. i think we should talk about it on saturday, dont gorget it guys #magi at freenode Personally not a fan of high PoS, but keen to be convinced, welcome to have a talk. |
|
|
|
|
Show Posts
