NetTecture
|
|
July 16, 2011, 07:49:46 AM |
|
Some comments and questions * What happened to options like AGGRESSION etc.? I can not enter them anymore. * Can we as next step get multi GPU support? So that one can define multiple GPU or just something like -d all and it uses all GPU? Quite a lot of people I think run multi card configs with the same card everywhere, makes everything easier. * Is there a way to get rid of the new measurement output and put it back to "line by line"? Easier for those starting the miner from a control program that then parses the output Not sure I overlooked that. Besides that - good thing.
|
|
|
|
brunoshady
Sr. Member
Offline
Activity: 262
Merit: 250
Dubs Get
|
|
July 17, 2011, 05:50:04 PM |
|
Some comments and questions * What happened to options like AGGRESSION etc.? I can not enter them anymore. * Can we as next step get multi GPU support? So that one can define multiple GPU or just something like -d all and it uses all GPU? Quite a lot of people I think run multi card configs with the same card everywhere, makes everything easier. * Is there a way to get rid of the new measurement output and put it back to "line by line"? Easier for those starting the miner from a control program that then parses the output Not sure I overlooked that. Besides that - good thing. aggression = frames and the line by line, try the --verbose mode
|
😆
|
|
|
SeW900
Jr. Member
Offline
Activity: 90
Merit: 1
|
|
July 18, 2011, 04:11:43 PM |
|
Hey guys, the new Catalyst drivers come with support for OpenCL and AMD APP. So is the installation of the SDK still required, or is it not?
|
|
|
|
|
OCedHrt
Member
Offline
Activity: 111
Merit: 10
|
|
July 20, 2011, 12:14:53 AM |
|
Does the current git run? Traceback (most recent call last): File "C:\Users\Michael Hsu\Desktop\m0mchil-poclbm-1b5ec3e\HttpTransport.py", l ine 45, in loop self.queue_work(work) File "C:\Users\Michael Hsu\Desktop\m0mchil-poclbm-1b5ec3e\Transport.py", line 122, in queue_work self.process(work) File "C:\Users\Michael Hsu\Desktop\m0mchil-poclbm-1b5ec3e\Transport.py", line 81, in process self.set_difficulty(work.difficulty) File "C:\Users\Michael Hsu\Desktop\m0mchil-poclbm-1b5ec3e\Transport.py", line 76, in set_difficulty self.true_target = np.array(unpack('IIIIIIII', true_target.decode('hex')), d type=np.uint32) File "c:\Python27\lib\encodings\hex_codec.py", line 42, in hex_decode output = binascii.a2b_hex(input) TypeError: Odd-length string
|
|
|
|
Luke-Jr
Legendary
Offline
Activity: 2576
Merit: 1186
|
|
July 20, 2011, 12:42:18 AM |
|
|
|
|
|
nebiki
|
|
July 20, 2011, 03:35:38 AM |
|
It prints newlines on every hash rate update when run in lxterminal (linuxcoin).
Anyone experience the same? Know why?
EDIT: was told this is because of the width of the terminal window. There is some extra whitespace after the hash rate text which wraps around to form blank lines if the the terminal window isn't wide enough.
so how did you fix it? i want a clean command prompt @ windows
|
|
|
|
LehmanSister
Member
Offline
Activity: 68
Merit: 10
High Desert Dweller-Where Space and Time Meet $
|
|
July 20, 2011, 04:00:44 AM |
|
Once again, Luke Jr. saves the day.Here's the patch for the actually curious: --- a/Transport.py +++ b/Transport.py @@ -62,7 +62,7 @@ class Transport(object):
def set_difficulty(self, difficulty): self.difficulty = difficulty - bits = '%08x' % (difficulty,) + bits = hex(difficulty) bits = bits[2:len(bits) - 1] bits = ''.join(list(chunks(bits, 2))[::-1]) true_target = hex(int(bits[2:], 16) * 2 ** (8 * (int(bits[:2], 16) - 3)))
|
ISO: small island nations with large native populations excited to pay tribute to flying gods, will trade BTC.
|
|
|
OCedHrt
Member
Offline
Activity: 111
Merit: 10
|
|
July 20, 2011, 05:03:33 AM |
|
Didn't click through But, shouldn't also: - bits = bits[2:len(bits) - 1] In that case? This is to remove the 0x and L.
|
|
|
|
Luke-Jr
Legendary
Offline
Activity: 2576
Merit: 1186
|
|
July 20, 2011, 05:37:50 AM |
|
Didn't click through But, shouldn't also: - bits = bits[2:len(bits) - 1] In that case? This is to remove the 0x and L. Good catch, also sanitized the code immediately following it and repushed...
|
|
|
|
bitcoindaddy
|
|
July 20, 2011, 05:22:15 PM |
|
Luke-Jr, what is the comand if you are just starting out fresh with your special poclbm and you don't already have it? (I'm no git expert). Does it depend on the normal poclbm (does that need to be installed first)?
|
|
|
|
Luke-Jr
Legendary
Offline
Activity: 2576
Merit: 1186
|
|
July 20, 2011, 05:51:13 PM |
|
Luke-Jr, what is the comand if you are just starting out fresh with your special poclbm and you don't already have it? (I'm no git expert). Does it depend on the normal poclbm (does that need to be installed first)? git clone git://github.com/m0mchil/poclbm.git && cd poclbm && git fetch https://git.gitorious.org/~Luke-Jr/bitcoin/luke-jrs-poclbm.git bugfix_difficulty_error && git merge FETCH_HEAD
|
|
|
|
ivank2139
Newbie
Offline
Activity: 27
Merit: 0
|
|
July 21, 2011, 04:32:53 AM |
|
This fix worked for me and got me back to mining at full complement. I sent you some bitcoins. thanks.
|
|
|
|
NickW
Newbie
Offline
Activity: 27
Merit: 0
|
|
July 24, 2011, 12:01:08 AM |
|
The current version is great, but could we get more significant figures on the stale share percentage please? I know I can easily put in to a calculator myself, but it would be nice to see straight away rather then seeing 0% most of the time . Keep up the good work though.
|
|
|
|
bitcoindaddy
|
|
July 24, 2011, 12:14:27 PM |
|
The newest version already shows percentage out to the thousandth's place.
pit.deepbit.net:8332 [321.472 MH/s (~305 MH/s)] [Rej: 8/2430 (0.33%)]
|
|
|
|
zubatej
Newbie
Offline
Activity: 8
Merit: 0
|
|
July 26, 2011, 06:27:36 PM |
|
Switching to backup pools fails because of variable name mismatch. Rename either all occurences of backup_server_index to backup_pool_index or other way around. Can be done by sed -i 's/backup_server_index/backup_pool_index/' Transport.py
|
|
|
|
Luke-Jr
Legendary
Offline
Activity: 2576
Merit: 1186
|
|
July 29, 2011, 03:27:58 AM |
|
Here's a new bugfix: http://tinyurl.com/3qo2j3oSynopsis: Don't turn off X-Roll-Ntime just because the work submit doesn't include the header. (it really should associate it to the specific work, but this is good enough)
|
|
|
|
|
super6
Newbie
Offline
Activity: 28
Merit: 0
|
|
July 30, 2011, 03:52:46 PM |
|
I'm getting "poclbm.exe has stopped working" errors in guiminer. Will this patch fix it and how do I go about doing that in windows?
|
|
|
|
Luke-Jr
Legendary
Offline
Activity: 2576
Merit: 1186
|
|
July 30, 2011, 09:12:12 PM |
|
The following branches are stable for merging to mainline: - bugfix_workspecific_rollntime -- Don't disable rollntime if the header is missing on share submissions (only on new work)
- extensions_header -- Send X-Mining-Extensions and X-Mining-Hashrate headers
- logformat -- Split off log-style printing into a separate --logformat option
- efficiency -- Display work efficiency (accepted shares / works)
- submit_retry -- Retry submitting works after network errors
The above are all merged into my branch named combo. I also have a extended_timeout branch that people with poor network connectivity can merge to get a more reliable mining experience. This one is not part of combo, and probably not suitable for merging to mainline. To pull any of the above, run: git fetch git://gitorious.org/~Luke-Jr/bitcoin/luke-jrs-poclbm.git PutBranchNameHere && git merge FETCH_HEAD
|
|
|
|
|