[ANN] [MINT] Mintcoin (POS / 5%) [NO ICO] [Fair distro, community maintained] - page 112.

kiklo

legendary

Activity: 1092

Merit: 1000

Quote from: someguy4 on August 25, 2015, 08:13:30 PM

What would be in this genesis block, all the coin values/ages that the previous blocks amounted to?

That and more, but the devs will have to answer if anything like it is possible at this stage of crypto development. If not that just leaves a chain swap and block time and algo change to fix the memory/cpu issues which would be easier to implement.

Cool

someguy4

newbie

Activity: 32

Merit: 0

What would be in this genesis block, all the coin values/ages that the previous blocks amounted to?

kiklo

legendary

Activity: 1092

Merit: 1000

Quote from: Derek492 on August 25, 2015, 07:30:46 PM

We have already decided to keep the faster block time. I think the idea of a new Genesis block the way you describe sounds good. How hard would that be to implement? It would have to be generated from a timestamp of say 1000 blocks in the past though right? Long enough to ensure chain consensus and such. Is that how it works?

In theory ,
yes it would have to be generated far enough back that their is no issue of consensus,
So after 1000 blocks it would become Block 0 of a new block chain, for anyone downloading a new wallet.

Implementation difficulty and precise details, the devs will have to answer once they explore it.

FYI:
A running Genesis block that cuts off the old section of blockchain is still theory, so whoever designs it will be the 1st.

Derek492

sr. member

Activity: 356

Merit: 250

Quote from: kiklo on August 25, 2015, 04:43:30 PM

Quote from: Derek492 on August 25, 2015, 11:33:16 AM

But changing algorithm doesn't change anything long-term, does it? Eventually the RAM keeps rising?. I think there is some degree of benefit using scypt.

Ram will keep rising , however scrypt was designed for the memory requirements to grow faster than other algos , originally to keeps asics from being able to mine it. Tongue

https://bitcointalksearch.org/topic/scrypt-is-memory-intensive-therefore-no-asics-but-how-213872

If the goal is to run mint on low powered devices , then groestl algo is a better choice.
If low powered devices are not a concern stay with scrypt as it is more widely used.

Cool

FYI:
If someone gets a genesis block system working, it would keep the memory requirements lower.
Basically every 6 months worth of blocks , a new genesis block is created which has all of the account balances, that way new blockchains are never older than 6 months, keeping the memory requirements down and speeding up install for new users.

Only other option is to change block speed after a chain swap takes place.
Where Mint & Hobo take forever to sync after being off for a few days, Sprouts sync in seconds since they have a 2.5 minute block speed verses a 30 second block.

We have already decided to keep the faster block time. I think the idea of a new Genesis block the way you describe sounds good. How hard would that be to implement? It would have to be generated from a timestamp of say 1000 blocks in the past though right? Long enough to ensure chain consensus and such. Is that how it works?

kiklo

legendary

Activity: 1092

Merit: 1000

Quote from: Derek492 on August 25, 2015, 11:33:16 AM

But changing algorithm doesn't change anything long-term, does it? Eventually the RAM keeps rising?. I think there is some degree of benefit using scypt.

Ram will keep rising , however scrypt was designed for the memory requirements to grow faster than other algos , originally to keeps asics from being able to mine it. Tongue

https://bitcointalksearch.org/topic/scrypt-is-memory-intensive-therefore-no-asics-but-how-213872

If the goal is to run mint on low powered devices , then groestl algo is a better choice.
If low powered devices are not a concern stay with scrypt as it is more widely used.

Cool

FYI:
If someone gets a genesis block system working, it would keep the memory requirements lower.
Basically every 6 months worth of blocks , a new genesis block is created which has all of the account balances, that way new blockchains are never older than 6 months, keeping the memory requirements down and speeding up install for new users.

Only other option is to change block speed after a chain swap takes place.
Where Mint & Hobo take forever to sync after being off for a few days, Sprouts sync in seconds since they have a 2.5 minute block speed verses a 30 second block.

Derek492

sr. member

Activity: 356

Merit: 250

On a side note, is everything ready to go as far as the block reward reduction coming up? Last yea, the drop from 20% to 15% resulted in the network freezing. Is that expected to happen again? According to my calculations we only have about 1-2 months until the drop to 10%.

Derek492

sr. member

Activity: 356

Merit: 250

Quote from: kiklo on August 24, 2015, 08:40:03 PM

Quote from: sambiohazard on August 24, 2015, 08:03:11 PM

Quote from: Fuzzbawls on August 24, 2015, 06:11:20 PM

Lot of talk about RAM usage going on. While I do agree that it can become an issue down the road, it isn't as "game breaking" as the excessive CPU usage currently present in MINT. RAM, much like HD space, is a fairly inexpensive thing to acquire more of (with certain limits obviously). But processor inefficiency is a killer in that end users can't just add more CPU power to their computers.

I agree, but if you are not mining on CPU or running other CPU intensive apps then Mint uses only 9-10% of a i5 quad core which is manageable, although it would be really nice to have it reduced. Also RAM can be upgraded but it will become an issue for adoption, a skeptic/cautious user would ditch something that uses RAM in GBs as opposed to bitcoin that only uses something like 200-400 MB. When competing in field of 10s of good altcoins these inefficiencies will pull us down.

Mint PoS currently uses scrypt algorithm,
https://en.wikipedia.org/wiki/Scrypt
to have a permanent lower CPU & memory usage, would require switching to another algorithm .

Cool

But changing algorithm doesn't change anything long-term, does it? Eventually the RAM keeps rising?. I think there is some degree of benefit using scypt.

JohnSmithy

member

Activity: 87

Merit: 10

From my standpoint the cpu and ram usage of mintcoin are too high, i have a sub-average computer however.

30-50% cpu usage of a dual core and 700 MB on a 3 GB physical ram system (while regularly using up to 1.5 GB virtual ram, which gives massive lag when it is being assigned).

sambiohazard

hero member

Activity: 840

Merit: 1000

Quote from: Fuzzbawls on August 24, 2015, 11:27:49 PM

I've spent most of my time overhauling the GUI interface. This past week was spent ironing out the syncing/connection issues and making sure that is stable as well as working towards the removal of orphan blocks in user's local wallets.

Last night a crash issue was reported regarding the importing of wallets so I spent some time looking into that. And now...well...now i'm out of beer Sad

I will try to get some beer for you by this weekend. I am having lots of power cuts & internet outage this so my mining income is all too shitty. I hope someone else can donate some beers till then.

Fuzzbawls

hero member

Activity: 750

Merit: 500

I've spent most of my time overhauling the GUI interface. This past week was spent ironing out the syncing/connection issues and making sure that is stable as well as working towards the removal of orphan blocks in user's local wallets.

Last night a crash issue was reported regarding the importing of wallets so I spent some time looking into that. And now...well...now i'm out of beer Sad

Flyskyhigh

sr. member

Activity: 291

Merit: 250

Ezekiel 34:11, John 10:25-30

The CPU is somewhat "high", but at least it seems to be constant, and not really increasing.... (at least from what I can see, am i wrong?) If so, I don't see that as much of a threat.
On the other-hand, the RAM is rising noticeably as months go by...

But maybe working on the CPU issues will also help solve some RAM issues?

EDIT: Thanks Fuzzballs and Supasonic, and all contributors, you are making a mintcoin the best!

FYI: I just want to say, that I personally know of some big investment getting ready to flow in to mintcoin as these issues get resolved.

sambiohazard

hero member

Activity: 840

Merit: 1000

Thats great to hear!! Keep forging your way, i will keep donating to dev bounty when i can. I just keep nudging when an issue comes up. I have no real skills that i can help with so i just keep talking about what can be done. heh. Grin

Fuzzbawls

hero member

Activity: 750

Merit: 500

Quote from: sambiohazard on August 24, 2015, 08:55:48 PM

Quote from: kiklo on August 24, 2015, 08:40:03 PM

Quote from: sambiohazard on August 24, 2015, 08:03:11 PM

Quote from: Fuzzbawls on August 24, 2015, 06:11:20 PM

Lot of talk about RAM usage going on. While I do agree that it can become an issue down the road, it isn't as "game breaking" as the excessive CPU usage currently present in MINT. RAM, much like HD space, is a fairly inexpensive thing to acquire more of (with certain limits obviously). But processor inefficiency is a killer in that end users can't just add more CPU power to their computers.

I agree, but if you are not mining on CPU or running other CPU intensive apps then Mint uses only 9-10% of a i5 quad core which is manageable, although it would be really nice to have it reduced. Also RAM can be upgraded but it will become an issue for adoption, a skeptic/cautious user would ditch something that uses RAM in GBs as opposed to bitcoin that only uses something like 200-400 MB. When competing in field of 10s of good altcoins these inefficiencies will pull us down.

Mint PoS currently uses scrypt algorithm,
https://en.wikipedia.org/wiki/Scrypt
to have a permanent lower CPU & memory usage, would require switching to another algorithm .

Cool

I am not sure about that. Archcoin also uses Scrypt algo but it hardly uses any CPU & 200-300 MB RAM and it is almost a year old too. I think we need a developer who knows what he is doing or study other coins to get best bits together. My best guess is that most of the issues are related to faster block times/lots of blocks to process. I see same problem with UTC/Ultracoin as well. This why downloading the blockchain takes too long, lots of headers to process.

Archcoin has 278K blocks in 1 year but we have 1.82M blocks in 1.5 years. Faster block time has a trade off of having lots of blocks. But as i said we need to stop guessing and hire a dev that can study the code and work out some solutions. Do we have no devs in the community who can help?

Both Supasonic and myself are actively working on updates and there is a lot of ground to cover. My previous post wasn't intended to come off as a request for somebody else to handle this, just a statement of fact that CPU utilization is higher than it needs to be. It is one item on the list of things to do.

kiklo

legendary

Activity: 1092

Merit: 1000

Quote from: sambiohazard on August 24, 2015, 08:55:48 PM

I am not sure about that. Archcoin also uses Scrypt algo but it hardly uses any CPU & 200-300 MB RAM and it is almost a year old too. I think we need a developer who knows what he is doing or study other coins to get best bits together. My best guess is that most of the issues are related to faster block times/lots of blocks to process. I see same problem with UTC/Ultracoin as well. This why downloading the blockchain takes too long, lots of headers to process.

Archcoin has 278K blocks in 1 year but we have 1.82M blocks in 1.5 years. Faster block time has a trade off of having lots of blocks. But as i said we need to stop guessing and hire a dev that can study the code and work out some solutions. Do we have no devs in the community who can help?

Arch currently only has 16,408,824 coins and their block speed is 1 minute.
Mint has a 30 second block speed , but if you watch the block explorer , it average 3 or 4 blocks per minute where as Arch averages 1 or 2 minutes before finding a block.

Was'nt Supasonic supposed to be the new Mint Dev?

FYI:
Sprouts has a 2.5 minute block time and a different algo and they sync very fast.
HoboNickel is 30 second block time with scrypt algo and they also use ~1 gig of ram.

Cool

sambiohazard

hero member

Activity: 840

Merit: 1000

Quote from: kiklo on August 24, 2015, 08:40:03 PM

Quote from: sambiohazard on August 24, 2015, 08:03:11 PM

Quote from: Fuzzbawls on August 24, 2015, 06:11:20 PM

Lot of talk about RAM usage going on. While I do agree that it can become an issue down the road, it isn't as "game breaking" as the excessive CPU usage currently present in MINT. RAM, much like HD space, is a fairly inexpensive thing to acquire more of (with certain limits obviously). But processor inefficiency is a killer in that end users can't just add more CPU power to their computers.

I agree, but if you are not mining on CPU or running other CPU intensive apps then Mint uses only 9-10% of a i5 quad core which is manageable, although it would be really nice to have it reduced. Also RAM can be upgraded but it will become an issue for adoption, a skeptic/cautious user would ditch something that uses RAM in GBs as opposed to bitcoin that only uses something like 200-400 MB. When competing in field of 10s of good altcoins these inefficiencies will pull us down.

Mint PoS currently uses scrypt algorithm,
https://en.wikipedia.org/wiki/Scrypt
to have a permanent lower CPU & memory usage, would require switching to another algorithm .

Cool

I am not sure about that. Archcoin also uses Scrypt algo but it hardly uses any CPU & 200-300 MB RAM and it is almost a year old too. I think we need a developer who knows what he is doing or study other coins to get best bits together. My best guess is that most of the issues are related to faster block times/lots of blocks to process. I see same problem with UTC/Ultracoin as well. This why downloading the blockchain takes too long, lots of headers to process.

Archcoin has 278K blocks in 1 year but we have 1.82M blocks in 1.5 years. Faster block time has a trade off of having lots of blocks. But as i said we need to stop guessing and hire a dev that can study the code and work out some solutions. Do we have no devs in the community who can help?

kiklo

legendary

Activity: 1092

Merit: 1000

Quote from: sambiohazard on August 24, 2015, 08:03:11 PM

Quote from: Fuzzbawls on August 24, 2015, 06:11:20 PM

Lot of talk about RAM usage going on. While I do agree that it can become an issue down the road, it isn't as "game breaking" as the excessive CPU usage currently present in MINT. RAM, much like HD space, is a fairly inexpensive thing to acquire more of (with certain limits obviously). But processor inefficiency is a killer in that end users can't just add more CPU power to their computers.

I agree, but if you are not mining on CPU or running other CPU intensive apps then Mint uses only 9-10% of a i5 quad core which is manageable, although it would be really nice to have it reduced. Also RAM can be upgraded but it will become an issue for adoption, a skeptic/cautious user would ditch something that uses RAM in GBs as opposed to bitcoin that only uses something like 200-400 MB. When competing in field of 10s of good altcoins these inefficiencies will pull us down.

Mint PoS currently uses scrypt algorithm,
https://en.wikipedia.org/wiki/Scrypt
to have a permanent lower CPU & memory usage, would require switching to another algorithm .

Cool

sambiohazard

hero member

Activity: 840

Merit: 1000

Quote from: Fuzzbawls on August 24, 2015, 06:11:20 PM

Lot of talk about RAM usage going on. While I do agree that it can become an issue down the road, it isn't as "game breaking" as the excessive CPU usage currently present in MINT. RAM, much like HD space, is a fairly inexpensive thing to acquire more of (with certain limits obviously). But processor inefficiency is a killer in that end users can't just add more CPU power to their computers.

I agree, but if you are not mining on CPU or running other CPU intensive apps then Mint uses only 9-10% of a i5 quad core which is manageable, although it would be really nice to have it reduced. Also RAM can be upgraded but it will become an issue for adoption, a skeptic/cautious user would ditch something that uses RAM in GBs as opposed to bitcoin that only uses something like 200-400 MB. When competing in field of 10s of good altcoins these inefficiencies will pull us down.

Fuzzbawls

hero member

Activity: 750

Merit: 500

Lot of talk about RAM usage going on. While I do agree that it can become an issue down the road, it isn't as "game breaking" as the excessive CPU usage currently present in MINT. RAM, much like HD space, is a fairly inexpensive thing to acquire more of (with certain limits obviously). But processor inefficiency is a killer in that end users can't just add more CPU power to their computers.

Derek492

sr. member

Activity: 356

Merit: 250

Quote from: duplan on August 24, 2015, 04:16:17 AM

XDN — aka duckNote » DarkNote » DigitalNote — moved its chain out of RAM ; their approach works flawless .
The code is open source , MIT style , the principles and workings could be potentially useful to adjourn ours .

I think we should study a way to lower the memory footprint , close all memory leaks as soon as possible ,
it is essential we adapt our technology to fit the needs of the more wholesome vision MINT wants to meet .

How does it work?

If it works flawless we should take a solid look at implementation. The RAM issue going forward is the biggest threat I see to Mintcoin imo. Hard drive space is not as big of an issue, the blockchain is only about 2GB and hard drive space is cheaper and much easier to accomplish.

crzybilly

sr. member

Activity: 703

Merit: 250

Mintcoin faucet works again.

Topic: [ANN] [MINT] Mintcoin (POS / 5%) [NO ICO] [Fair distro, community maintained] - page 112. (Read 1369788 times)