Pages:
Author

Topic: [ANN][XEL] Elastic Project - The Decentralized Supercomputer - page 28. (Read 450520 times)

hero member
Activity: 742
Merit: 501
Coralreefer, thanks for your effort! You are my personal hero! Sorry for not responding you yet, I will catch up this night.

So if you both working hard I feel useless now so I'll go to Elastic source and change old logo to new one. It's always something Wink

It's already done! ;-) Just wait a bit until I commit the new client.

legendary
Activity: 1260
Merit: 1168
Coralreefer, thanks for your effort! You are my personal hero! Sorry for not responding you yet, I will catch up this night.

So if you both working hard I feel useless now so I'll go to Elastic source and change old logo to new one. It's always something Wink

It's already done! ;-) Just wait a bit until I commit the new client.
sr. member
Activity: 464
Merit: 260
Awesome work, guys!

EK, coralreefer, unvoid. Awesome!

Elastic is about to get real.  Cool

Indeed. Long time since I've seen all devs on one page at BCT thread.

I agree, hopefully we'll gain some momentum...still lots of work to be done.

btw...nice work on the node guide unvoid...
hero member
Activity: 535
Merit: 500
Awesome work, guys!

EK, coralreefer, unvoid. Awesome!

Elastic is about to get real.  Cool

Indeed. Long time since I've seen all devs on one page at BCT thread.
legendary
Activity: 952
Merit: 1000
Awesome work, guys!

EK, coralreefer, unvoid. Awesome!

Elastic is about to get real.  Cool

Thanks for all your hard work guys:)
hero member
Activity: 840
Merit: 500
Risk taker & Black Swan farmer.
Awesome work, guys!

EK, coralreefer, unvoid. Awesome!

Elastic is about to get real.  Cool
hero member
Activity: 994
Merit: 513
(…)

     2) How to store data between iterations and distribute to miners
    
(…)



I just asked myself this very question. We'd need practically virtual RAM, right?

I don't know, but would something like IPFS be an option? Or maybe SNs provide storage? Maybe a job issuer could provide storage on their own for a start?
hero member
Activity: 535
Merit: 500
hero member
Activity: 535
Merit: 500
Coralreefer, thanks for your effort! You are my personal hero! Sorry for not responding you yet, I will catch up this night.

So if you both working hard I feel useless now so I'll go to Elastic source and change old logo to new one. It's always something Wink
hero member
Activity: 900
Merit: 500
Coralreefer, thanks for your effort! You are my personal hero! Sorry for not responding you yet, I will catch up this night.
ek coralreefer,our hero.
thx.
sr. member
Activity: 464
Merit: 260
another quick update...

When EK wrote his BTC example, he identified 2 key issues.  1) We needed a more flexible memory model, and 2) Incorporating Functions into the language was needed (it was originally not incorporated due to issues w/ recursion).  I am about done with the upgrades to the ElasticPL engine which addresses these 2 issues as well as incorporates several other fixes to prepare for SN integration.

To see an example of how much more user friendly the language is now, I rewrote EK's BTC example using the language upgrades:

     https://github.com/sprocket-fpga/xel_miner/blob/master/examples/SHA256_BTC.epl

Here's what it looked like before the changes to ElasticPL:

     https://github.com/OrdinaryDude/elastic_bitcoin_miner/blob/master/test

I am still a couple weeks away from wrapping up all the changes to ElasticPL, but they should be ready to test soon.  However, there are still other key issues that still need to be addressed before everything is ready:

     1) Will we have POW, and if so what logic is needed to ensure the miner actually performed the work
     2) How to store data between iterations and distribute to miners
     3) SN Integration

These are complex issues that will take time to solve.  It will require quite a bit of change to both the Core Server and Miner and will require quite a bit of coding from both EK and myself to complete.



Well done Coralreefer.
sha256 proof-of-work example implemented in ElasticPL language - awesome!

Did you guys think about having some kind of shared state between workers/nodes per job?
I wonder how would you divide a task of some sort between workers in the network, in such a way that they do it in parallel?
(I'm not sure if my question is perhaps to generic)

Thanks.



Ida1000, item #2 above on the todo list will pretty much address this.  EK has already done this is his github, but the issue is how to make it more generic / robust so that it can handle different use cases.  Some jobs may just want be able to store their state in just a few ints, but other may require 1MB or more of data....still working through this.

Also, EK no rush getting back to me...I still have plenty to do and I know you're pretty busy right now.
sr. member
Activity: 581
Merit: 253
another quick update...

When EK wrote his BTC example, he identified 2 key issues.  1) We needed a more flexible memory model, and 2) Incorporating Functions into the language was needed (it was originally not incorporated due to issues w/ recursion).  I am about done with the upgrades to the ElasticPL engine which addresses these 2 issues as well as incorporates several other fixes to prepare for SN integration.

To see an example of how much more user friendly the language is now, I rewrote EK's BTC example using the language upgrades:

     https://github.com/sprocket-fpga/xel_miner/blob/master/examples/SHA256_BTC.epl

Here's what it looked like before the changes to ElasticPL:

     https://github.com/OrdinaryDude/elastic_bitcoin_miner/blob/master/test

I am still a couple weeks away from wrapping up all the changes to ElasticPL, but they should be ready to test soon.  However, there are still other key issues that still need to be addressed before everything is ready:

     1) Will we have POW, and if so what logic is needed to ensure the miner actually performed the work
     2) How to store data between iterations and distribute to miners
     3) SN Integration

These are complex issues that will take time to solve.  It will require quite a bit of change to both the Core Server and Miner and will require quite a bit of coding from both EK and myself to complete.



Well done Coralreefer.
sha256 proof-of-work example implemented in ElasticPL language - awesome!

Did you guys think about having some kind of shared state between workers/nodes per job?
I wonder how would you divide a task of some sort between workers in the network, in such a way that they do it in parallel?
(I'm not sure if my question is perhaps to generic)

Thanks.

legendary
Activity: 1260
Merit: 1168
Coralreefer, thanks for your effort! You are my personal hero! Sorry for not responding you yet, I will catch up this night.
member
Activity: 122
Merit: 10
another quick update...

When EK wrote his BTC example, he identified 2 key issues.  1) We needed a more flexible memory model, and 2) Incorporating Functions into the language was needed (it was originally not incorporated due to issues w/ recursion).  I am about done with the upgrades to the ElasticPL engine which addresses these 2 issues as well as incorporates several other fixes to prepare for SN integration.

To see an example of how much more user friendly the language is now, I rewrote EK's BTC example using the language upgrades:

     https://github.com/sprocket-fpga/xel_miner/blob/master/examples/SHA256_BTC.epl

Here's what it looked like before the changes to ElasticPL:

     https://github.com/OrdinaryDude/elastic_bitcoin_miner/blob/master/test

I am still a couple weeks away from wrapping up all the changes to ElasticPL, but they should be ready to test soon.  However, there are still other key issues that still need to be addressed before everything is ready:

     1) Will we have POW, and if so what logic is needed to ensure the miner actually performed the work
     2) How to store data between iterations and distribute to miners
     3) SN Integration

These are complex issues that will take time to solve.  It will require quite a bit of change to both the Core Server and Miner and will require quite a bit of coding from both EK and myself to complete.



Great work and sharpness!
member
Activity: 122
Merit: 10
Almost done with the light thing! Expect a finished prototype tomorrow!

damn nice!
sr. member
Activity: 448
Merit: 250
Ben2016
Two awesome news from two great devs . Thank you guys !
sr. member
Activity: 464
Merit: 260
another quick update...

When EK wrote his BTC example, he identified 2 key issues.  1) We needed a more flexible memory model, and 2) Incorporating Functions into the language was needed (it was originally not incorporated due to issues w/ recursion).  I am about done with the upgrades to the ElasticPL engine which addresses these 2 issues as well as incorporates several other fixes to prepare for SN integration.

To see an example of how much more user friendly the language is now, I rewrote EK's BTC example using the language upgrades:

     https://github.com/sprocket-fpga/xel_miner/blob/master/examples/SHA256_BTC.epl

Here's what it looked like before the changes to ElasticPL:

     https://github.com/OrdinaryDude/elastic_bitcoin_miner/blob/master/test

I am still a couple weeks away from wrapping up all the changes to ElasticPL, but they should be ready to test soon.  However, there are still other key issues that still need to be addressed before everything is ready:

     1) Will we have POW, and if so what logic is needed to ensure the miner actually performed the work
     2) How to store data between iterations and distribute to miners
     3) SN Integration

These are complex issues that will take time to solve.  It will require quite a bit of change to both the Core Server and Miner and will require quite a bit of coding from both EK and myself to complete.

legendary
Activity: 1260
Merit: 1168
Almost done with the light thing! Expect a finished prototype tomorrow!
member
Activity: 122
Merit: 10
I can setup a permanent node on a VPS if I have clear instructions, Im not great with *nix.

https://talk.elasticexplorer.org/t/elastic-fullnode-for-dummies/231

thanks! it is very easy. I just added a node.
cant wait for mainnet
legendary
Activity: 1708
Merit: 1000
Reality is stranger than fiction
Whoever runs full nodes will forge new coins? Also the more coins a node holds, the more coins it forges?
Pages:
Jump to: