[ANN] [XMG] MAGI | CPU mining | mPoW | mPoS | [MagiPay]

[okae]

Post in progress!
Try to find all different Magi threads on the different forums


https://bitcointalk.org/index.php?topic=735170.0
https://bitcointalk.org/index.php?topic=1164424.0 (Spanish)

http://bitcoingarden.tk/forum/index.php?topic=4544.0
http://bitcoingarden.tk/forum/index.php?topic=4554.0 (Spanish)

https://www.cryptopia.co.nz/Forum/Thread/329
https://www.cryptopia.co.nz/Forum/Thread/351
https://www.cryptopia.co.nz/Forum/Thread/332 (Spanish)

http://cryptos.us/forum/announcements/33-ann-xmg-magi-1e-posii-1e-pom-m7m-cpu-bittrex-poloniex-1e-poi

http://www.bitcoinaltcoin.com/index.php?topic=778.0

http://altcointalk.org/index.php/topic,531.0.html (Spanish)
http://altcointalk.org/index.php/topic,520.0.html

not bad, but if we can find more translators for more languages then it will be perfect

you know, all help are welcome, let's hope that people watching this thread and come from other languages encourage to do it too

[thdim]

To force the mindset change I have an idea that could be benefit (or not) to Magi : why not open a pool that only confirmed Magi member could join ? That could spread the hashes a bit more and if we have a total pool power bigger than big miners, their rewards would be decreased a lot and maybe they will give up.

That's the spirit for me.

[cryptonit]

i think spreadcoin invented a technology that only allow solo mining

now if u combine that with a check to accept new blocks only for example once every hour from same address

u would make GPU mining senceless he would just mine to many blocks and all beside the first each hour would be rejected by network

nice sideeffect u would have a lot nodes in network because mining would always required to have a node active too

[111magic]

i think spreadcoin invented a technology that only allow solo mining

now if u combine that with a check to accept new blocks only for example once every hour from same address

u would make GPU mining senceless he would just mine to many blocks and all beside the first each hour would be rejected by network

nice sideeffect u would have a lot nodes in network because mining would always required to have a node active too

Wow good tip.
Joe could you check this.

[ocminer]

i think spreadcoin invented a technology that only allow solo mining

now if u combine that with a check to accept new blocks only for example once every hour from same address

u would make GPU mining senceless he would just mine to many blocks and all beside the first each hour would be rejected by network

nice sideeffect u would have a lot nodes in network because mining would always required to have a node active too

Actually that "technology" was already circumvented several times, I have (had) a pool up for SPR but closed it due to no interest: https://spr.suprnova.cc  I might reopen it if interest comes back.

You can mine every block to a new address without problems.

[kondiomir]

[111magic]

Post in progress!
Try to find all different Magi threads on the different forums


https://bitcointalk.org/index.php?topic=735170.0
https://bitcointalk.org/index.php?topic=1164424.0 (Spanish)

http://bitcoingarden.tk/forum/index.php?topic=4544.0
http://bitcoingarden.tk/forum/index.php?topic=4554.0 (Spanish)

https://www.cryptopia.co.nz/Forum/Thread/329
https://www.cryptopia.co.nz/Forum/Thread/351
https://www.cryptopia.co.nz/Forum/Thread/332 (Spanish)

http://cryptos.us/forum/announcements/33-ann-xmg-magi-1e-posii-1e-pom-m7m-cpu-bittrex-poloniex-1e-poi

http://www.bitcoinaltcoin.com/index.php?topic=778.0

http://altcointalk.org/index.php/topic,531.0.html (Spanish)
http://altcointalk.org/index.php/topic,520.0.html

not bad, but if we can find more translators for more languages then it will be perfect

you know, all help are welcome, let's hope that people watching this thread and come from other languages encourage to do it too

There are a lot more.
I will add this list later. Have little problem with my laptop atm. Wifi is not working ok. Hope to fix it soon so I can join the irc also again.

[111magic]

I really like this picture.
Thats awesome Kondiomir.
I'am very proud you are in our awesome Magi community! Well done!

[starmman]

testcc90cc         KH/s-----23,657      XMG/Day-----    2,726.157

Wonder what they are testing  

Now 44,231 KH/s, that's not normal.

This is a concern...

smells like a gpus on it..

Maybe that explains why the hashrate has been so high recently =(

[joelao95]

https://bitcointalk.org/index.php?topic=735170.0

...

Thanks for the collection of info, mate, I have changed / updated the OP (nearly took me a day lol). I guess some info still missing, let me know that.

[joelao95]

[img ]https://www.filepicker.io/api/file/0VYh3ak6Ther5FZtFwWs[/img]

Nice pic. 

[joelao95]

i think spreadcoin invented a technology that only allow solo mining

now if u combine that with a check to accept new blocks only for example once every hour from same address

u would make GPU mining senceless he would just mine to many blocks and all beside the first each hour would be rejected by network

nice sideeffect u would have a lot nodes in network because mining would always required to have a node active too

Hi cryptonit, thanks for being around for advices; I've been being out of the time on the things we discussed, really tighten on the things.

I'm wondering how solo mining blocks gpu mining, not quite getting this "u would make GPU mining senceless he would just mine to many blocks and all beside the first each hour would be rejected by network". I may read through their page.

One thing I now very much concern is that, every time we create a new mechanism that is not working the way we're 100% sure, it creates hidden facts.

Actually that "technology" was already circumvented several times, I have (had) a pool up for SPR but closed it due to no interest: https://spr.suprnova.cc  I might reopen it if interest comes back.

You can mine every block to a new address without problems.

Hi ocminer, people have no interest to the pool, is that because they like to sole mining? Also, could provide deail information about "mine every block to a new address"?

[joelao95]

To force the mindset change I have an idea that could be benefit (or not) to Magi : why not open a pool that only confirmed Magi member could join ? That could spread the hashes a bit more and if we have a total pool power bigger than big miners, their rewards would be decreased a lot and maybe they will give up.

It seems an idea, however, there are other pools they can register and even they can use different IP.

[joelao95]

testcc90cc         KH/s-----23,657      XMG/Day-----    2,726.157

Wonder what they are testing  

Now 44,231 KH/s, that's not normal.

This is a concern...

smells like a gpus on it..

Maybe that explains why the hashrate has been so high recently =(

Despite this is a botnet or GPU miner, I hope people are not scared of it, and actually we could be at easy with his number. Here are the reason and some numbers.

To get over this concern, two situations:

If it is a botnet, then solution is to go higher hashrate to much exceed botnet's hash, which quite needs GPU mining to come in; this is not what magi intends to be, I meant the high hashrate.

If it is a GPU miner, then solution is to go a different algo. Well this is tough. The said getting a new mechanism will lead to a possibility of hidden facts unless we're 100% sure about the solution and no one can tackle it. After a brief discussion with Wolf0 (mostly I was asking), to resist GPU, a reasonable amount of memory needs to be plugged into the algo, and it simply exhausts the memory of GPU. For this solution to work, CPU must have a significant amount of memory than GPU. However, the nowaday's technology has pushed even higher and cheap memory; huge amount of GPU won't be unlikely in the future, so it isn't absolutely safe with such an algo. There is still another concern, CPU working with higher amount of memory automatically disables the mining capability of those low-end / old pc, at least suppressing their hashing significantly; for example, Raspberry Pi will be out of the scene. This is very much contrast with magi's aim at low cost device mining. What I want is to get magi mining on cheap devices; I don't think we can get to anywhere and I am afraid we're just a regular coin without this purpose. So I'd hesitate to get on an algo like yescript or any others with a large number of memory. But I don't think this is a simple rejection, rather we'll consider once something better comes up.

Reading the above by far, you can see the solutions are very contradictory. You want "a cheap mining" like CPU 100% assurance (well nothing 100%), use a memory enhanced algo (unless there is approach other than memeory), which is actually not "cheap"; also there is the botnet concern. You want "GPU mining", surely no where cheap and no where cheap mining.

Except the above two, what I can see is the idea we have here, the variable PoW block.

http://xmg.makejar.com/xmg-per-day/

Let's take the amount of XMG generated on "17.10.2015" which is quite a normal day, 4 668.85913172 / 444 =10.5 XMG / block. The "18.10.2015" is quite much less than "17.10.2015" (the big miner wasn't online whole day). I might be wrong; what I've seen when the big hashrate online is mostly <1 XMG / block; if it stays a day, the total XMG would be quite less than 444 XMG (taking 444 blocks / a day, same as "17.10.2015"). It's sure that when a big hashrate coming online, the total amount of XMG will be reduced significantly. Let's say the big hash is 40 Mh/s and the rest of the network is 60 Mh/s, then what he gets is 444 * 0.4 * 0.012962 = $ 2.3 in a day (I believed this is quite a reasonable estimate, and actually this miner isn't in the pool for a day in which case the reward may drop down further). So this is $2.3 for a miner with 40 Mh/s.

Speaking of the profits, when the big miner isn't getting profits, the other miners (regular) are receiving loss too. The point is that the big miner has to carry the loss on his own, while the loss is much distributed among the regular miners, so individual miner won't worry too much about that in my opinion (say there are 1000 regular miners over one or few big miners).

This is my philosophy about the idea behind scene as I mentioned; and this reminds me an experiment in physics "Schrodinger's cat" in a sealed box with a flask of poison (https://en.wikipedia.org/wiki/Schr%C3%B6dinger%27s_cat); we don't know the cat is alive or dead unless we observe it by openning up the box; at the moment when we open up the box, we trigger a signal that may cause the poison released to kill the cat, so we don't know the dead cat was killed by the poison or the observer (indirectly). So a miner is wondering if he can catch up the whole block rewards; the difference here is that, once he comes along, he actually pull the trigger, and causes the collapse of the block rewards.

What happens is that this block reward adjustment process must take place real quick, and the regular miners must stay mining as normal (no compromise at all). And there is another condition, the normal miners must outnumber those weird ones significantly.

Another thing is about low hashrate which could cause 51% attack; this seems fine so far for magi, a PoW/PoS hybrid coin, and you know our PoS-II is safe. Either way, this is another point we need to take into consideration in order to improve this coin. Let me know any ideas you could have.

[trihex]

testcc90cc         KH/s-----23,657      XMG/Day-----    2,726.157

Wonder what they are testing  

Now 44,231 KH/s, that's not normal.

This is a concern...

smells like a gpus on it..

Maybe that explains why the hashrate has been so high recently =(

Despite this is a botnet or GPU miner, I hope people are not scared of it, and actually we could be at easy with his number. [...]

i think it probably has more to do with the fpga that came out for the coin than gpu mining. i tried getting information out of the guy that posted a question about his on reddit but he never responded to me.

https://www.reddit.com/r/coinmagi/comments/3ob60w/question_about_fpga/

[joelao95]

NEW Wallets (v1.2.3.1)
Official site, Sourceforge, Github release, Windows, Mac OS X, Linux

Thanks to lightsplasher to total balance function on the Qt interface.

Please make sure you have a backup of the wallet before launching the new app. All versions (Win, Mac, Linux) are with db4.8, and this will be the version we should use. There was a mistake in my linux machine which built the linux wallet with db5, and I made sure the linux version is a statically linked version (compiled by linux mint, it'll be fine with all debian based variants, but I'm unsure other linux version). Again please make a copy of wallet.dat before launching the new app.

Regarding the launch time of the current wallet, I believe this is not an issue of the db version, as I have used higher version in my linux box. I have the inclination of migrating latest bitcoin source code to magi, but this would be a long journey to go with.

@lightsplasher and all others, please feel free to make changes and submit merge request; I like to see changes to improve the Qt miner too if some one is interested. Currently I have tighten schedule on my personal stuffs, and I almost spent a week on the stuffs that's so expensive ; this thing is really time consuming. Here is the repo.

https://github.com/magi-project

[Mr_Bomberman]

Still upgrading like never ending story

You do it proper way

[111magic]

NEW Wallets (v1.2.3.1)
Official site, Sourceforge, Github release, Windows, Mac OS X, Linux

Thanks to lightsplasher to total balance function on the Qt interface.

Please make sure you have a backup of the wallet before launching the new app. All versions (Win, Mac, Linux) are with db4.8, and this will be the version we should use. There was a mistake in my linux machine which built the linux wallet with db5, and I made sure the linux version is a statically linked version (compiled by linux mint, it'll be fine with all debian based variants, but I'm unsure other linux version). Again please make a copy of wallet.dat before launching the new app.

Regarding the launch time of the current wallet, I believe this is not an issue of the db version, as I have used higher version in my linux box. I have the inclination of migrating latest bitcoin source code to magi, but this would be a long journey to go with.

@lightsplasher and all others, please feel free to make changes and submit merge request; I like to see changes to improve the Qt miner too if some one is interested. Currently I have tighten schedule on my personal stuffs, and I almost spent a week on the stuffs that's so expensive ; this thing is really time consuming. Here is the repo.

https://github.com/magi-project

Awesome job! Well done and special thanks to Lightsplasher!

[111magic]

testcc90cc         KH/s-----23,657      XMG/Day-----    2,726.157

Wonder what they are testing  

Now 44,231 KH/s, that's not normal.

This is a concern...

smells like a gpus on it..

Maybe that explains why the hashrate has been so high recently =(

Despite this is a botnet or GPU miner, I hope people are not scared of it, and actually we could be at easy with his number. Here are the reason and some numbers.

To get over this concern, two situations:

If it is a botnet, then solution is to go higher hashrate to much exceed botnet's hash, which quite needs GPU mining to come in; this is not what magi intends to be, I meant the high hashrate.

If it is a GPU miner, then solution is to go a different algo. Well this is tough. The said getting a new mechanism will lead to a possibility of hidden facts unless we're 100% sure about the solution and no one can tackle it. After a brief discussion with Wolf0 (mostly I was asking), to resist GPU, a reasonable amount of memory needs to be plugged into the algo, and it simply exhausts the memory of GPU. For this solution to work, CPU must have a significant amount of memory than GPU. However, the nowaday's technology has pushed even higher and cheap memory; huge amount of GPU won't be unlikely in the future, so it isn't absolutely safe with such an algo. There is still another concern, CPU working with higher amount of memory automatically disables the mining capability of those low-end / old pc, at least suppressing their hashing significantly; for example, Raspberry Pi will be out of the scene. This is very much contrast with magi's aim at low cost device mining. What I want is to get magi mining on cheap devices; I don't think we can get to anywhere and I am afraid we're just a regular coin without this purpose. So I'd hesitate to get on an algo like yescript or any others with a large number of memory. But I don't think this is a simple rejection, rather we'll consider once something better comes up.

Reading the above by far, you can see the solutions are very contradictory. You want "a cheap mining" like CPU 100% assurance (well nothing 100%), use a memory enhanced algo (unless there is approach other than memeory), which is actually not "cheap"; also there is the botnet concern. You want "GPU mining", surely no where cheap and no where cheap mining.

Except the above two, what I can see is the idea we have here, the variable PoW block.

http://xmg.makejar.com/xmg-per-day/

Let's take the amount of XMG generated on "17.10.2015" which is quite a normal day, 4 668.85913172 / 444 =10.5 XMG / block. The "18.10.2015" is quite much less than "17.10.2015" (the big miner wasn't online whole day). I might be wrong; what I've seen when the big hashrate online is mostly <1 XMG / block; if it stays a day, the total XMG would be quite less than 444 XMG (taking 444 blocks / a day, same as "17.10.2015"). It's sure that when a big hashrate coming online, the total amount of XMG will be reduced significantly. Let's say the big hash is 40 Mh/s and the rest of the network is 60 Mh/s, then what he gets is 444 * 0.4 * 0.012962 = $ 2.3 in a day (I believed this is quite a reasonable estimate, and actually this miner isn't in the pool for a day in which case the reward may drop down further). So this is $2.3 for a miner with 40 Mh/s.

Speaking of the profits, when the big miner isn't getting profits, the other miners (regular) are receiving loss too. The point is that the big miner has to carry the loss on his own, while the loss is much distributed among the regular miners, so individual miner won't worry too much about that in my opinion (say there are 1000 regular miners over one or few big miners).

This is my philosophy about the idea behind scene as I mentioned; and this reminds me an experiment in physics "Schrodinger's cat" in a sealed box with a flask of poison (https://en.wikipedia.org/wiki/Schr%C3%B6dinger%27s_cat); we don't know the cat is alive or dead unless we observe it by openning up the box; at the moment when we open up the box, we trigger a signal that may cause the poison released to kill the cat, so we don't know the dead cat was killed by the poison or the observer (indirectly). So a miner is wondering if he can catch up the whole block rewards; the difference here is that, once he comes along, he actually pull the trigger, and causes the collapse of the block rewards.

What happens is that this block reward adjustment process must take place real quick, and the regular miners must stay mining as normal (no compromise at all). And there is another condition, the normal miners must outnumber those weird ones significantly.

Another thing is about low hashrate which could cause 51% attack; this seems fine so far for magi, a PoW/PoS hybrid coin, and you know our PoS-II is safe. Either way, this is another point we need to take into consideration in order to improve this coin. Let me know any ideas you could have.

Good text Joe.
One other thing we as awesome Magi community must continue. Stay strong and keep working together. As Magi is growing with new followers, members, miners and other people who joining the Crypto-world there is always attention from competition. And we all noticed before this is not always in a good way. Magi respects other coins and will never try to grow bigger with negative attention towards other coins. Magi likes a fair grow and it might take more time it will give stronger results.
So all keep work together to stay strong and work with our dev to achieve improvements.
Magi will be the next generation digital currency and will enter the top! This month the Magi android app will be released. The PoM campaign will finish this month (31-october-2015) All rewards will be paid asap after closing. New campaign will start but more info about that later!
Thanks for reading!
 

[111magic]

i think spreadcoin invented a technology that only allow solo mining

now if u combine that with a check to accept new blocks only for example once every hour from same address

u would make GPU mining senceless he would just mine to many blocks and all beside the first each hour would be rejected by network

nice sideeffect u would have a lot nodes in network because mining would always required to have a node active too

Hi cryptonit, thanks for being around for advices; I've been being out of the time on the things we discussed, really tighten on the things.

I'm wondering how solo mining blocks gpu mining, not quite getting this "u would make GPU mining senceless he would just mine to many blocks and all beside the first each hour would be rejected by network". I may read through their page.

One thing I now very much concern is that, every time we create a new mechanism that is not working the way we're 100% sure, it creates hidden facts.

Actually that "technology" was already circumvented several times, I have (had) a pool up for SPR but closed it due to no interest: https://spr.suprnova.cc  I might reopen it if interest comes back.

You can mine every block to a new address without problems.

Hi ocminer, people have no interest to the pool, is that because they like to sole mining? Also, could provide deail information about "mine every block to a new address"?

This is an interesting subject.
Btw hashrate is little lower again.