Pages:
Author

Topic: [ANN] [SUP] Supcoin, PLUCK algorithm, Just Launched, Mine Now! - page 57. (Read 62226 times)

legendary
Activity: 910
Merit: 1000
0.9.3 versions solomines and connects just fine with any miner.
but what has changed with 10.++ versions, those wont get miners connected anymore??
sr. member
Activity: 439
Merit: 250
mmmmmm
For a second, I want to talk a bit about why I like PLUCK as it is designed.

As you can see, it is quite simple. It is a couple dozen lines of code (less if you don't count salsa), terse, but pretty easy to understand, very easy to configure, and defers all actual cryptographic properties to SHA256. SHA256 is very easy to understand, and can be reasonably contained within about 50 or so lines of code as well.

Try to name another hashing algorithm that is so simple. The only one that comes to mind for me is SHA256 itself. With this simplicity, it is of course easy to understand. And with it's extremely memory-hard properties, there are not many optimization techniques that can be done. What this boils down to is that after a somewhat optimized GPU miner comes out, there shouldn't be anything like X11, where some kernels can double the hashrate. There will always be microoptimizations and ways to do things a bit faster, but due to it's simplicity, I don't expect huge hashrate increases with optimizations.

This helps avoid the X11 GPU kernel race. X11 is actually extremely complicated. It is 11 hashing algorithms, each with their own potential optimizations, and their interactions can be optimized. Scrypt is also relatively complex, and if you've ever looked at it's implementation, it's not too easy to understand or get the whole picture. The numerous number of parameters also allow for special-case optimization pretty easily. Whereas PLUCK is designed to be as simple as possible with just 1 parameter controlling memory usage, meaning that the first implementation is probably not too far off from a significantly optimized version. PLUCK is not designed to be a password hashing algorithm, or anything like that. It is designed as a PoW algorithm, and I personally think it fits that quite well.

PLUCK is of course not ASIC "proof", which is impossible. However, I estimate GPU performance gains to be only 4-6x over CPU, and because it is heavy on memory access, rather than computation, it has potential to be a "cool" algorithm. GPUs get hot from computation, not memory access.. So, since this is memory access intense, the bottleneck should lie there, and thus bringing cooler CPU temps and lower power draw. This is of course my best guess though.
sr. member
Activity: 439
Merit: 250
mmmmmm
In order to raise tangible funds to motivate GPU miner devs, I am now hosting a donation address for Bitcoin.

1K6cUbtsn5xinD8XfjnTn7UnsMriLhZaGD

The first GPU miner to be created (AMD or Nvidia) and published under an open source license will receive all the money from this fund.

I'm putting in 0.25 BTC of my own money. I encourage everyone to donate what you can to help make this a lucrative bounty.
sr. member
Activity: 439
Merit: 250
mmmmmm
I finally have some compiled versions of the TEST-ONLY version of the code.

This wallet will NOT WORK when teh code is actually released and any coins mined on it will have no value and will not exist when the actual coin is released.

Note, this hasn't been themed yet, so it still shows bitcoin's logos and such.

Anyway, I've only got two version for now due to lack of time

http://earlz.net/static/supcoin-TEST-win32.zip
http://earlz.net/static/supcoin-TEST-linux64.tar.gz
sr. member
Activity: 257
Merit: 250
BUY BUY CELL CELL
Looks like a solid coin, definitely going to throw as much hashing power as I can get a hold of at this Grin
newbie
Activity: 24
Merit: 0
so we are still waiting for the launch of the coin!
sr. member
Activity: 439
Merit: 250
mmmmmm
I now have #supcoin registered on Freenode, for all you IRCers
sr. member
Activity: 439
Merit: 250
mmmmmm
Just to keep people from working on more logos, I'll decide the winner within 24 hours.
member
Activity: 81
Merit: 1002
It was only the wind.
Fair mining can only happen if the fastest kernels for the available technologies are shared from block number 1.

What I don't get is why coin devs don't work together with the mining kernel devs and make just that happen ...
And yes a 0.5 to 1 % premine would definatley justify that. The alternative is centralized huge mining farms with optimized kernels, your choice ...


Alternative sounds good.  Grin
legendary
Activity: 1400
Merit: 1050
earlz is starting to remind me a lot of r3wt...not a good sign

Just cause I don't like the premise of your coins, doesn't mean I'm a scammer Wink

any giveaway from this coin?

There is no giveaway other than the bounties
hope you're actually pushing to gh and not leaving that fresh repo empty to mess with us ;-)

Lol I see I'm being watched closely Wink

I've made a TEST-ONLY copy of the code and uploaded it to github. ONLY USE THIS FOR GPU MINER DEVELOPMENT. When the real coin is released, this code WILL NOT sync or work and it's blocks will be incompatible. This is only for GPU miner developers to test their code against a real coin. The github repo for it is at https://github.com/supcoin/supcoin-TEST-ONLY and it should sync automatically with 1 node and 2 or 3 blocks.

Again, DEVELOPMENT ONLY. Do not download this and try to mine it because you think these coins will be worth something. The real release will be later on
so exactly when does it launch ?
Can you make windows binaries for the lazy devs ?

I finally got Gitian working last night, so yea I can try and make a Windows wallet tonight for this code.

For reference, since I have Gitian working, at launch pre-compiled Linux, Windows, and Mac wallets will be available
Have you been able to make a win wallet ?

I've been trying to get Gitian working which makes that work, but ran into some mystery issue I can't figure out.

If you're committing to actually use it, I might try doing a one-off special Windows build for you... but otherwise, I'd rather wait and figure out this Gitian issue
take your time, I will look into but I don't know exactly when (and I can always compile it myself... was just to trying avoid the trouble)
legendary
Activity: 1134
Merit: 1000
www.cryptodesign.cc
Corporate style:



Coinshape style



3D style


small icon





sr. member
Activity: 439
Merit: 250
mmmmmm
earlz is starting to remind me a lot of r3wt...not a good sign

Just cause I don't like the premise of your coins, doesn't mean I'm a scammer Wink

any giveaway from this coin?

There is no giveaway other than the bounties
hope you're actually pushing to gh and not leaving that fresh repo empty to mess with us ;-)

Lol I see I'm being watched closely Wink

I've made a TEST-ONLY copy of the code and uploaded it to github. ONLY USE THIS FOR GPU MINER DEVELOPMENT. When the real coin is released, this code WILL NOT sync or work and it's blocks will be incompatible. This is only for GPU miner developers to test their code against a real coin. The github repo for it is at https://github.com/supcoin/supcoin-TEST-ONLY and it should sync automatically with 1 node and 2 or 3 blocks.

Again, DEVELOPMENT ONLY. Do not download this and try to mine it because you think these coins will be worth something. The real release will be later on
so exactly when does it launch ?
Can you make windows binaries for the lazy devs ?

I finally got Gitian working last night, so yea I can try and make a Windows wallet tonight for this code.

For reference, since I have Gitian working, at launch pre-compiled Linux, Windows, and Mac wallets will be available
Have you been able to make a win wallet ?

I've been trying to get Gitian working which makes that work, but ran into some mystery issue I can't figure out.

If you're committing to actually use it, I might try doing a one-off special Windows build for you... but otherwise, I'd rather wait and figure out this Gitian issue
legendary
Activity: 1400
Merit: 1050
earlz is starting to remind me a lot of r3wt...not a good sign

Just cause I don't like the premise of your coins, doesn't mean I'm a scammer Wink

any giveaway from this coin?

There is no giveaway other than the bounties
hope you're actually pushing to gh and not leaving that fresh repo empty to mess with us ;-)

Lol I see I'm being watched closely Wink

I've made a TEST-ONLY copy of the code and uploaded it to github. ONLY USE THIS FOR GPU MINER DEVELOPMENT. When the real coin is released, this code WILL NOT sync or work and it's blocks will be incompatible. This is only for GPU miner developers to test their code against a real coin. The github repo for it is at https://github.com/supcoin/supcoin-TEST-ONLY and it should sync automatically with 1 node and 2 or 3 blocks.

Again, DEVELOPMENT ONLY. Do not download this and try to mine it because you think these coins will be worth something. The real release will be later on
so exactly when does it launch ?
Can you make windows binaries for the lazy devs ?

I finally got Gitian working last night, so yea I can try and make a Windows wallet tonight for this code.

For reference, since I have Gitian working, at launch pre-compiled Linux, Windows, and Mac wallets will be available
Have you been able to make a win wallet ?
legendary
Activity: 952
Merit: 1000
Sup! Still Looking for Submissions for earlz to decide on.  This is just some fun for testing, and we could all use some fun. Check the below submissions.  If you believe you have something better, post it. This is a contest and there will be a winner.   Grin  Please retweet to help raise awareness.  Earlz is working hard on making sure this is a solid test of the codebase.  Lets show him some support.   Grin






full member
Activity: 224
Merit: 100
is this coin traded on ANY exchange?

What coin?  It's not even released, so, NO.
full member
Activity: 178
Merit: 100
is this coin traded on ANY exchange?
legendary
Activity: 952
Merit: 1000
Forum in beta http://forum.supcoin.com/ Please test make sure things are working.   Grin
sr. member
Activity: 439
Merit: 250
mmmmmm
Fair mining can only happen if the fastest kernels for the available technologies are shared from block number 1.

What I don't get is why coin devs don't work together with the mining kernel devs and make just that happen ...
And yes a 0.5 to 1 % premine would definatley justify that. The alternative is centralized huge mining farms with optimized kernels, your choice ...


Besides that I would chose multi algo support like myriad does ...

you need one cpu algo (15% chance to mint a block)

and you need one gpu algo which is very memory intensive so that low core high memory devices can be profitable as well (see high n factor coins) (50% chance to mint a block)

or split it up in 2 gpu algos one high memory algo and one standard algo (25% chance each)

hdd algo (burst) (20 % chance to mint a block)

and an asic algo (15 % chance to mint a block)

% numbers are a subject to discuss of course  ...

cpu/gpu algos need to be adjusted if workarounds are developed to gpu/asic mine them.
(there is no way around a hard fork every once in a while)



Eh, this is just a coin to test my codebase. The end goal for the real coin is 3 different mining mechanisms which should alleviate these concerns some.
sr. member
Activity: 506
Merit: 252
Fair mining can only happen if the fastest kernels for the available technologies are shared from block number 1.

What I don't get is why coin devs don't work together with the mining kernel devs and make just that happen ...
And yes a 0.5 to 1 % premine would definatley justify that. The alternative is centralized huge mining farms with optimized kernels, your choice ...


Besides that I would chose multi algo support like myriad does ...

you need one cpu algo (15% chance to mint a block)

and you need one gpu algo which is very memory intensive so that low core high memory devices can be profitable as well (see high n factor coins) (50% chance to mint a block)

or split it up in 2 gpu algos one high memory algo and one standard algo (25% chance each)

hdd algo (burst) (20 % chance to mint a block)

and an asic algo (15 % chance to mint a block)

% numbers are a subject to discuss of course  ...

cpu/gpu algos need to be adjusted if workarounds are developed to gpu/asic mine them.
(there is no way around a hard fork every once in a while)

sr. member
Activity: 249
Merit: 250
earlz is starting to remind me a lot of r3wt...not a good sign

Dude get your head out of your ass and stop insulting others because your services get criticized for being outright dogshit right out of the gate. earlz is one of the most trusted guys in this industry.
Pages:
Jump to: