Ufasoft Miner - Windows/Linux, x86/x64, SSE2/OpenCL, Open Source

[rchapoteau]

Is there a way to make this undetectable by virus scanners?  I can get it to work when I disable my virus scanners but I really don't like running my machine without them.

EDIT - Nevermind, I took a hexedit to the miner.dll and now symantec doesn't have a shitfit.

[1715171694]

1715171694

[1715171694]

1715171694

[d3m0n1q_733rz]

Hey, a small request for an extremely handy feature addition here.  Maybe you could implement a packet compression algorithm like gz to send and receive blocks much quicker.  I've been stuck at about 264 days for a few days now and block downloads are taking forever and a day.  It's one thing to compress the database (which causes a dramatic increase in data space used while doing so resulting in a self-defeat), but packets need to be transmitted a little faster.  It wouldn't be too difficult to implement a check to see if it is supported by other connected software.  And maybe compressing the database on-the-fly continuously could be helpful so as to avoid significant space requirements and time to do so.

Thanks!

[ufasoft]

Hey, a small request for an extremely handy feature addition here.  Maybe you could implement a packet compression algorithm like gz to send and receive blocks much quicker.  I've been stuck at about 264 days for a few

Block data  are uncompressible because composed of pubkeys/hash-values. They have very high level of information entropy

[d3m0n1q_733rz]

Hey, a small request for an extremely handy feature addition here.  Maybe you could implement a packet compression algorithm like gz to send and receive blocks much quicker.  I've been stuck at about 264 days for a few

Block data  are uncompressible because composed of pubkeys/hash-values. They have very high level of information entropy

That would be true if the block data weren't mostly plain text.  However, even using NTFS compression on the data causes the downloaded blocks to compress to less than 1/3 of their original size.  It's when it's utilizing characters outside of plain text that compression becomes difficult.   

Give it a try via a VPN or something if you like.  It might be easier than programming the transport compression into Ufasoft's Coin program.  However, the efficiency would be greatly increased if this function were included to reduce the plain text into other characters via compression.  The gz algorithm is tried and true for HTTP sessions, it could be of significant use.  Alternatively, there's the PPMD compression algorithm, but it's not really meant for quick packet compression and decompression.  Though, it would see the greatest benefits as far as packet size is concerned.

[ufasoft]

That would be true if the block data weren't mostly plain text.  However, even using NTFS compression on the data causes the downloaded blocks to compress to less than 1/3 of their original size.  It's when it's utilizing characters outside of plain text that compression becomes difficult.   

Block data of Bitcoin protocol is not a Plain Text. It is Binary.

Mining protocol is text JSON-RPC, but for mining we already have Stratum protocol to lower Internet Traffic.

[d3m0n1q_733rz]

That would be true if the block data weren't mostly plain text.  However, even using NTFS compression on the data causes the downloaded blocks to compress to less than 1/3 of their original size.  It's when it's utilizing characters outside of plain text that compression becomes difficult.  

Block data of Bitcoin protocol is not a Plain Text. It is Binary.

Mining protocol is text JSON-RPC, but for mining we already have Stratum protocol to lower Internet Traffic.

I'm not talking about the mining transport protocol.  I'm talking about the block download traffic of already solved blocks that need to be downloaded any time someone starts a new bitcoin tracking program such as Coin.exe.  Bitcoin has been around for over a year and with several blocks being solved each day, there are several GB of block data that must be downloaded to be up-to-date to the current block.  And, while the Stratum protocol only sends JSON messages for block updates, it can be significantly compressed upon transport so that over three times as much data on previous blocks can be sent at any given point in time.  This will help Bitcoin to become the truly universal currency that it was intended to be by cutting down on the required time and bandwidth required to be up-to-date for new Bitcoin users.

If you don't believe that the data can be compressed due to its being of encrypted data, try compressing the blocks you already have downloaded as a proof of concept.  Though I may have read my data wrong on the exact amount of space saved by which type of coin, it's still a significant amount of space saved by simple compression.  Just look at the results for yourself.  That's all I'm asking.

Here, look at this to see what I'm talking about:  http://web-resource-optimization.blogspot.com/2011/06/json-compression-algorithms.html

[ufasoft]

If you don't believe that the data can be compressed due to its being of encrypted data, try compressing the blocks you already have downloaded as a proof of concept.  Though I may have read my data wrong on the exact amount of space saved by which type of coin, it's still a significant amount of space saved by simple compression.  Just look at the results for yourself.  That's all I'm asking.

Let me explain:
One block is incomressible.
But if we download 100 000 blocks, they will have much duplicated data, so entire database is compressible.
Coin.exe does "Database normalization", if removes duplicated data during saving of blocks.

So we can compress blockchain download only when downloading entire blockchain as single file.
But P2P clients should update their databases block-by-block. In this case updates are not compressible.

[ufasoft]

0.39 released.
http://ufasoft.com/coin

ADDED:
  BIP22 protocol support
  Stratum protocol support.

[joshstandalone]

What means
New Stratum data with Clean
12/23/2012 5:43:28 AM Result: b889f31c accepted
12/23/2012 5:45:22 AM Result: 2116cc65 accepted


Processed: 2598 Mhash, 149 s with average Rate: 10.63707 MHash/s
Accepted: 0, average: 0 shares/min

why i see here accepted: 0 is this normal or some bug?
does he really accept shares or no ?

[ufasoft]

Accepted: 0, average: 0 shares/min

why i see here accepted: 0 is this normal or some bug?
does he really accept shares or no ?

It is bug, accepted-counter is not incremented.
Really the shares accepted at the Pool.

Thank you, will be fixed in next release.

[joshstandalone]

I am glad to help

[noncecents]

The installer cannot be ran in Windows 8 because of Windows SmartScreen, which cannot be fully disabled.

Additionally, if 7zip is used to unzip the installer, Windows 8 deletes the unzipped files and then informs you that SmartScreen has protected you from what you want.

To get the files out of the archive in Windows 8, you have to open the archive in 7zip, browse into the "coin" folder, and then drag and drop the files out of the archive into a folder on your destination media.

[ufasoft]

The installer cannot be ran in Windows 8 because of Windows SmartScreen, which cannot be fully disabled.

You can select "More info" in the SmartScreen and run the installer.

[K1773R]

The installer cannot be ran in Windows 8 because of Windows SmartScreen, which cannot be fully disabled.

You can select "More info" in the SmartScreen and run the installer.

ah win8, win is already a pain but win8? im sry 4 u

[turboNOMAD]

Was this miner ever tested on non-x86 hardware?
I tried to build ufasoft-0.39 on Toshiba AC100 with Ubuntu and compilation fails because of undefined "_lrotr".

[os2sam]

Was this miner ever tested on non-x86 hardware?
I tried to build ufasoft-0.39 on Toshiba AC100 with Ubuntu and compilation fails because of undefined "_lrotr".

This miner is written in assembler which is CPU specific.  It cannot be run on non x86 hardware.

[ufasoft]

Was this miner ever tested on non-x86 hardware?
I tried to build ufasoft-0.39 on Toshiba AC100 with Ubuntu and compilation fails because of undefined "_lrotr".

This miner is written in assembler which is CPU specific.  It cannot be run on non x86 hardware.

No, the miner has portable code and optional x86-specific code.

GCC 4.6 or 4.7 required to build on ARM architecture (_lrotr instruction)

[os2sam]

Was this miner ever tested on non-x86 hardware?
I tried to build ufasoft-0.39 on Toshiba AC100 with Ubuntu and compilation fails because of undefined "_lrotr".

This miner is written in assembler which is CPU specific.  It cannot be run on non x86 hardware.

No, the miner has portable code and optional x86-specific code.

GCC 4.6 or 4.7 required to build on ARM architecture (_lrotr instruction)

Ah, OK.  Never mind then.

[Kryptox]

I downloaded this for my system - Win7 x64 SP1 and it wouldn't run so I uninstalled it.  Now whenever I try to run Litecoin-qt.exe I get a Runtime error even after re-downloading the Litecoin client and wiping out my %appdata%.

[ufasoft]

I downloaded this for my system - Win7 x64 SP1 and it wouldn't run so I uninstalled it.  Now whenever I try to run Litecoin-qt.exe I get a Runtime error even after re-downloading the Litecoin client and wiping out my %appdata%.

What is the Runtime error? You can see it in the %APPDATA%\Litecoin\debug.log file.