Topic: artforz and coblee gpu mining litecoin since the start? - page 11. (Read 32574 times)

I believe the demo will only work on 5800 and 6900 series cards. Maybe others will work. mtrlt needs to release the opencl source instead of binaries to get most cards working good.

I think someone has now managed to get 250KH/s on a 5830 overclocked.

Doesn't work:

Kernel build not successful: -46
2012-02-29 14:53:44 Error: Error creating OpenCL kernel

Yeah that's why usually no one has heard of them.
So who are these "best developers"
Seriously I had heard of NONE of them until this thread was posted about someone referred to as "mtrlt"

As for running a windows executable written by some solidcoin unknown with no source code released - I'll wait until someone else tries that
(I wouldn't want to infect my kids computer ... ... I use linux)

Though I will add yet again - I was the first to suggest that GPU should be able to handle scrypt with the small settings used.

And lastly - yeah that amazing new word you've never heard of - "decompiler" - come on man - upgrade to at least ... the 90's
Why not just release the source.
If someone actually did give a shit about the program existing they'd just decompile it anyway.
Or is the source just do fucking crappy that it's best hidden away?
Hell I don't write the best looking code in the world, but it certainly isn't so shitty that I'm scared to let others see any of it.
... and why does the executable have the text "bitcoin" in it?

And ... why is it doing something that cgminer does and no other miner does in the CL compiling code ... ... ... ... ...

And I hope that "Copyright (c) 1992-2004 by P.J. Plauger, licensed by Dinkumware, Ltd. ALL RIGHTS RESERVED."
wasn't pirated and you have the rights to distribute it ...

... and since I know little about windows specific code, why does it try to get administrator privileges?

Goes to show where the best developers are... working at SolidCoin.

http://solidcoin.info/reaper.html

"GPU resistant" . Nice work artforz and coblee, fooled your few fans for some months. Reaper gets 200KH/s on a 5830 . artfoz and coblee went around telling people gpu mining was impossible because they only got 9KH/s on a 6870.... believable with hindsight?

Once I read the CH's messages about thousands of modifications in solidcoin, and this project very interested me. I've downloaded the source code and a well-studied it. What should I say ... Yes, It's true. There are thousands of meaningless changes, and the creation of work visibility. Applause.

If anything about mrlt's code is true to the SC norm, it will be released late; it will be glamorously called open source, but won't be open; and will be nothing but a rework of somebody else's work.

Bet is still open, yet there is bias for mtrlt having the miner. Is the source posted yet?

I think, that you misunderstood me. I mean that this topic is associated with this video:

http://www.youtube.com/watch?v=9SboyoievYU

 

I have been running "a few" computers from a few labs on Litecoin getting several hundred KH and was mining solo... if enough people do that you get a large unknown without large botnets or GPU miners..

The total network speed is probably computed using the "getnetworkhashps" RPC call, which can be very inaccurate after difficulty retargets, even when the difficulty doesn't change significantly.

In addition, the 15 MH/s miner that made an appearance at litecoinpool.org a few days ago has now (quite rightly) switched to solo mining. By the way, I have talked with this miner, and I am now pretty sure that this is neither a botnet nor a GPU farm.

But the graph in http://pool-x.eu/net says 67.2 % litecoin mined by unknown



51% attack going on  ??

None that was ever released... mtrlt claims to have one.. buuuut...

I wasted some time to read this topic from beginning, so... There is the miner? 

Please see: https://bitcointalksearch.org/topic/thread-about-gpu-mining-and-litecoin-64239

Well my node has 287 connections right now (and for unrelated purposes I have a instance of bitcoind running on amazon EC 2000 miles away) so it is unlikely (although not impossible).  Still even if isolated my node will still treat EVERYTHING as untrusted until verified internally which is why your claim is full of shit.

So it isn't possible for poison node to provide fake proof of a transaction, or provide fake blocks.  The poison nodes (if they have 100% control of information to my node) could prevent me from seeing information but that is about it.  Even in a rare instance of information blackout the fact that the client/node DOESN'T trust anything it gets from the network limits the scope of any attacks that poison node could make.

No you don't.  The client DISTRUSTS every node it meets on the network.  That is why it downloads a complete copy of the blockchain to ensure it can trace every transaction back to an origination block.

Every communication, transaction, node list, or block the client receives is verified before acting upon.  Each node will refuse to forward information it considers invalid.  The node strives to get large number of geo-diverse nodes to connect with to ensure it has clear access to the network at large.


TL/DR:
You aren't just wrong you are 180 deg wrong.
The Bitcoin client treats every other node as potentially hostile and looks at all data received from the network with distrust.

If you're willing to swallow the pill of centralized control, then are protocol-breaking changes really a problem? You shouldn't have to worry about being on the wrong side of a fork if your client gets its confirmation from a trusted source instead of its own calculations.

Begs the question, if he has more knowledge of how a true p2p currency should work, why has it taken him 4 protocol revs so far, and he still hasn't gotten it right ?

Soon to be 4th.

OMFG Smoothie out of nowhere!  Where have you been?