ciocgun
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
March 05, 2015, 04:10:07 PM |
|
HI, i got a problem today
i've deleted a buy offer of an asset and got an error. Some of the amount as returned into my balance, and some not.. what can i do?
do you see any asset balance for which you placed a buy order? i think your buy order got partially filled? ![Huh](https://bitcointalk.org/Smileys/default/huh.gif) if i sum the total mined amound it's more then the total amount spent to buy asset
|
|
|
|
bensam1231
Legendary
Offline
Activity: 1750
Merit: 1024
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
March 05, 2015, 04:46:39 PM |
|
So I have a theory as to part of what is causing the fluctuations in the network (the saw tooth pattern). After a fast block, Blagos miner doesn't error out on the block and proceed to the next one. This means all machines currently mining on the last block will continue mining on a block that is already past. They do this till they finish scanning the disks as they normally would. Meanwhile the fresh block on the network will not be mined by the full network as everyone running Blagos miner is still working on the last block until they finish as if a new block hadn't appeared.
When multiple fast blocks happen, the next block will not receive the full mine, let alone the one after that, which can cause the network hash power to stagger downhill.
Obviously there are still people not using Blagos miner and may be using Pocminer, Urays, or Rurus where this does not seem to be a issue, so it only effects about 30% of the network. It also seems as though this does not happen if you mine on the Dev pool, but I've witnessed this behavior across multiple machines solo and on other pools.
It's not hard to see this. Look through the log of your miners, when you see a 'fast block or corrupt' or a incomplete mine (% isn't 100 before the next block), you can look at the difference in time between when that event happened and when the miner continues on to the next block and it's the difference of time that was remaining before it completed the mine of the 'fast block'.
Now, I'm not exactly sure why this hasn't been fixed yet. I brought this up multiple times including in a PM with Blago and this definitely looks like something that could easily be fixed and it's not hard to reproduce, so I assume it's to give fast machines a leg up on slow ones.
If someone wants to hack a fix, a program that monitors either Blagos miner or the log file and simply restarts the miner when either 'fast block' appears in the log or when the wallet pops a new block that would effectively fix this.
not so sure about this. the default setting in the config file is 2 seconds (going by memory) so at most, and if default value is present, it will take 2 seconds for Blago's miner to give up a block and start mining the next one. I reduced my setting to 750 ms for ex. but I have been thinking about this: let's say I have 200TB, atm the calculator says that should be 32386 burst or /7350 = 4.4 blocks. if I see that since 00:00:00.000 I have found 5 blocks, and let's say we have another 12 hours to go in that day, then it's very likely I shouldn't find any more blocks for the remainder of the 12 hours. so I switch to mine against a pool and rackup another 10000-15000 burst. then when midnight hits, I switch to solo. if next day I have fast luck, I repeat, if not, I keep mining solo until next day and check if I should repeat. and so on. this is kind of like pool hoping to increase your reward? this is a crude description and you can substitute any numbers with whatever you want to test this scenario... does this make sense, is it possible to "game" burst mining like this for a bit? pool operators should change their reward algos to prevent this? There isn't a setting in the config for this, unless you have a different one then me. There is two settings ""SendInterval": 200 and "UpdateInterval": 2000. One reflects the frequency of sending updates to the wallet and receiving them, neither address this bug. The values above are mine as well. 'Luck' doesn't work that way. There is not a cap on how many blocks you can find in a day. Every block is a lottery. There is no limit on the number of blocks you can find. It would be different if Burst intelligently distributed blocks, but it doesn't, which is why we get fucked on some days and not on others. That is also why if you don't have enough 'tickets' you shouldn't be mining on your own or there wont be enough potential finds to balance things out.
|
I buy private Nvidia miners. Send information and/or inquiries to my PM box.
|
|
|
daWallet
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
March 05, 2015, 04:51:42 PM |
|
but I have been thinking about this:
let's say I have 200TB, atm the calculator says that should be 32386 burst or /7350 = 4.4 blocks. if I see that since 00:00:00.000 I have found 5 blocks, and let's say we have another 12 hours to go in that day, then it's very likely I shouldn't find any more blocks for the remainder of the 12 hours. so I switch to mine against a pool and rackup another 10000-15000 burst. then when midnight hits, I switch to solo. if next day I have fast luck, I repeat, if not, I keep mining solo until next day and check if I should repeat. and so on. this is kind of like pool hoping to increase your reward? this is a crude description and you can substitute any numbers with whatever you want to test this scenario... does this make sense, is it possible to "game" burst mining like this for a bit? pool operators should change their reward algos to prevent this?
This doesn't make sense because if you're very lucky for some time solo, then switch to a pool... your deadlines still will be statistically worse for some time. The pool doesn't know you have 200 TB, the pool sees only your submitted deadlines and these will be quite bad in your scenario.
|
github/dawallet Burst Client for Win & Burstcoin.biz
|
|
|
mczarnek
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
March 05, 2015, 05:33:02 PM |
|
Yea, the hard part about this is my day trader cash. You kinda have to keep some there. Agreed.. just have to be careful about how much is on the exchange versus in your wallet and ideally split it between exchanges.
|
|
|
|
dvdelkin
Newbie
Offline
Activity: 5
Merit: 0
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
March 05, 2015, 06:43:13 PM Last edit: March 05, 2015, 06:56:10 PM by dvdelkin |
|
Dear Sirs Please answer my questions 1) why my nvidia gtx275 doesn't work for plotting? My driver is 314.22. When i just start plotting an error appears [ERROR][-36][CL_INVALID_COMMAND_QUEUE] Error in step2 kernel finish. What's wrong? Or i tried geforce 580 or 750mobile and i got the same error 2) i previously successfully plotted one plot with the size of 1.74Tb. But currently i tried several times to plot just about 400Gb and in several minutes i got an error "GPU plotter closed". What happens? Plotting now doesn't come to the end, it stops in several minutes after the beginning 3) Can i use simultaneously Radeon 290 and gtx275 for plotting? 4) What are correct settings, url for V2 Burst Pool? On official page it's written " http://127.0.0.1:8125 http://thispoolsip", but that doesn't work. I use Uray miner. 5) Why during mining my Disk C is periodically loaded for 100% and internet sites are loaded very slow? It happens only during mining, My plots are on the disk K. 6) I've been mining for about 12 hours on tompool.org. On http://mining.tompool.org:82/ it was written that my round share was about 144. But in my wallet i didn't get NOTHING from tompool.org. I have one plot for 1.74Tb 7) In comparison then I've been mining about 5 hours at burst.ninja:8124 and already got 16bursts. Where are my bursts from tompool.org?
|
|
|
|
Irontiga
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
March 05, 2015, 06:57:01 PM |
|
Dear Sirs Please answer my questions 1) why my nvidia gtx275 doesn't work for plotting? My driver is 314.22. When i just start plotting an error appears [ERROR][-36][CL_INVALID_COMMAND_QUEUE] Error in step2 kernel finish. What's wrong? Or i tried geforce 580 or 750mobile and i got the same error 2) i previously successfully plotted one plot with the size of 1.74Tb. But currently i tried several times to plot just about 400Gb and in several minutes i got an error "GPU plotter closed". What happens? Plotting now doesn't come to the end, it stops in several minutes after the beginning 3) Can i use simultaneously Radeon 290 and gtx275 for plotting? 4) What are correct settings, url for V2 Burst Pool? On official page it's written " http://127.0.0.1:8125 http://thispoolsip", but that doesn't work. I use Uray plotter. 5) Why during mining my Disk C is periodically loaded for 100% and internet sites are loaded very slow? It happens only during mining, My plots are on the disk K. 6) I've been mining for about 12 hours on tompool.org. On http://mining.tompool.org:82/ it was written that my round share was about 144. But in my wallet i didn't get NOTHING from tompool.org. I have one plot for 1.74Tb 7) In comparison then I've been mining about 5 hours at burst.ninja:8124 and already got 16bursts. Where are my bursts from tompool.org? 1)Dunno 2)^^^ Try a CPU plotter 3)^^^ But i think U can ![Cheesy](https://bitcointalk.org/Smileys/default/cheesy.gif) 4)Plotter and miner are different....I think u meant uray miner, and it doesn't support dev v2 pool. It does support burst.ninja tho(actually, burst.ninja supports it), so u can just mine there. For dev v2 u need blago miner or the old javaminer from the dev(don't go there). 5)Plotting is where u hardout, full time make plots. This is essentially just storing your hashrate One it is done, it is done. Mining reads a chunk of your plots(based on the scoop the current block needs, it is always 1/4096 of your plots), and then calculates deadlines, and submits them to pool or network. 6) You probably didn't hit min payout or the pool didn't find a block...i'm guessing it's the latter. 7) Burst.ninja is just awesome ![Cheesy](https://bitcointalk.org/Smileys/default/cheesy.gif)
|
|
|
|
coinits
Legendary
Offline
Activity: 1582
Merit: 1019
011110000110110101110010
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
March 05, 2015, 07:30:43 PM |
|
Dear Sirs Please answer my questions 1) why my nvidia gtx275 doesn't work for plotting? My driver is 314.22. When i just start plotting an error appears [ERROR][-36][CL_INVALID_COMMAND_QUEUE] Error in step2 kernel finish. What's wrong? Or i tried geforce 580 or 750mobile and i got the same error 2) i previously successfully plotted one plot with the size of 1.74Tb. But currently i tried several times to plot just about 400Gb and in several minutes i got an error "GPU plotter closed". What happens? Plotting now doesn't come to the end, it stops in several minutes after the beginning 3) Can i use simultaneously Radeon 290 and gtx275 for plotting? 4) What are correct settings, url for V2 Burst Pool? On official page it's written " http://127.0.0.1:8125 http://thispoolsip", but that doesn't work. I use Uray plotter. 5) Why during mining my Disk C is periodically loaded for 100% and internet sites are loaded very slow? It happens only during mining, My plots are on the disk K. 6) I've been mining for about 12 hours on tompool.org. On http://mining.tompool.org:82/ it was written that my round share was about 144. But in my wallet i didn't get NOTHING from tompool.org. I have one plot for 1.74Tb 7) In comparison then I've been mining about 5 hours at burst.ninja:8124 and already got 16bursts. Where are my bursts from tompool.org? 1)Dunno 2)^^^ Try a CPU plotter 3)^^^ But i think U can ![Cheesy](https://bitcointalk.org/Smileys/default/cheesy.gif) 4)Plotter and miner are different....I think u meant uray miner, and it doesn't support dev v2 pool. It does support burst.ninja tho(actually, burst.ninja supports it), so u can just mine there. For dev v2 u need blago miner or the old javaminer from the dev(don't go there). 5)Plotting is where u hardout, full time make plots. This is essentially just storing your hashrate One it is done, it is done. Mining reads a chunk of your plots(based on the scoop the current block needs, it is always 1/4096 of your plots), and then calculates deadlines, and submits them to pool or network. 6) You probably didn't hit min payout or the pool didn't find a block...i'm guessing it's the latter. 7) Burst.ninja is just awesome ![Cheesy](https://bitcointalk.org/Smileys/default/cheesy.gif) I could never get NVIDIA to plot. Always used CPU.
|
Jump you fuckers! | The thing about smart motherfuckers is they sound like crazy motherfuckers to dumb motherfuckers. | My sig space for rent for 0.01 btc per week.
|
|
|
bensam1231
Legendary
Offline
Activity: 1750
Merit: 1024
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
March 05, 2015, 09:45:51 PM |
|
ROI on just a hard drive is up to 16 months (and that's if miners decide not to dump). Why is the network size still growing? D:
|
I buy private Nvidia miners. Send information and/or inquiries to my PM box.
|
|
|
fabula
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
March 05, 2015, 09:49:40 PM |
|
ROI on just a hard drive is up to 16 months (and that's if miners decide not to dump). Why is the network size still growing? D:
And why price is 100 satoshi when on agoust it reached 800? The main problem here is max supply. 2 billions. Too much.
|
|
|
|
Yanakitu Tenatako
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
March 05, 2015, 10:00:34 PM |
|
ROI on just a hard drive is up to 16 months (and that's if miners decide not to dump). Why is the network size still growing? D:
Price is not high it should be. There are ppl adjusting price by their needs.
|
|
|
|
riskyfire
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
March 05, 2015, 10:09:32 PM |
|
ROI on just a hard drive is up to 16 months (and that's if miners decide not to dump). Why is the network size still growing? D:
My guess is that casual users are plotting their free hard drive space and likely not looking for a full return as the money is costed into the computer purchase.
|
|
|
|
daWallet
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
March 05, 2015, 10:16:46 PM |
|
Price is in a stalemate because there are no announcements for weeks now. It would be nice to hear how development is going... The Crowdfund was successful, so some further information is appropriate. "ATDevelopmentFund The purpose of this case is to provide funding to help extend the technical work and general product development and marketing for AT inside the burst platform."
|
github/dawallet Burst Client for Win & Burstcoin.biz
|
|
|
xizmax
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
March 05, 2015, 11:05:34 PM |
|
Price is in a stalemate because there are no announcements for weeks now. It would be nice to hear how development is going... The Crowdfund was successful, so some further information is appropriate. "ATDevelopmentFund The purpose of this case is to provide funding to help extend the technical work and general product development and marketing for AT inside the burst platform." Work is progressing well, however like I said on several occasions, there are no hard dates to report. Just a few hints: crazyearner is almost ready with his ultimate mining guide cryo has an interesting new development in the pipeline Byte enterprises crew is hard at work doing their own things, all of use to community vbcs is busy developing the "tag" system for AT So, no worries. People aren't asleep ![Smiley](https://bitcointalk.org/Smileys/default/smiley.gif)
|
|
|
|
haitch
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
March 05, 2015, 11:58:50 PM |
|
Hey guys, is burst.ga down? ![Huh](https://bitcointalk.org/Smileys/default/huh.gif) burst.ga became burst.ninja due to DNS provider issues. H.
|
███ ███ ███ ███ ███ ███ ███ ███ ███ ███ ███ ███ █████████████████████████ ███ ███ ███ ███ ███ ███ | IRELINE |
██████ ██████ ██████ ██████ ██████ ██████ ██████ ██████
| Largest Fund worldwide for distributed application makers ███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ ███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ wireline.io - facebook.com/wirelineio - @wirelineio |
██████ ██████ ██████ ██████ ██████ ██████ ██████ ██████
| ●⚫⦁ ICO ⦁⚫● September 1 |
|
|
|
FakeAccount
Full Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif) ![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 248
Merit: 100
I'm not real
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
March 06, 2015, 04:37:19 AM |
|
but I have been thinking about this:
let's say I have 200TB, atm the calculator says that should be 32386 burst or /7350 = 4.4 blocks. if I see that since 00:00:00.000 I have found 5 blocks, and let's say we have another 12 hours to go in that day, then it's very likely I shouldn't find any more blocks for the remainder of the 12 hours. so I switch to mine against a pool and rackup another 10000-15000 burst. then when midnight hits, I switch to solo. if next day I have fast luck, I repeat, if not, I keep mining solo until next day and check if I should repeat. and so on. this is kind of like pool hoping to increase your reward? this is a crude description and you can substitute any numbers with whatever you want to test this scenario... does this make sense, is it possible to "game" burst mining like this for a bit? pool operators should change their reward algos to prevent this?
This doesn't make sense because if you're very lucky for some time solo, then switch to a pool... your deadlines still will be statistically worse for some time. The pool doesn't know you have 200 TB, the pool sees only your submitted deadlines and these will be quite bad in your scenario. not really. pools will pay for deadlines below their threshold. my example applies to specific day, the lucky ones. I think it's possible to increase mining returns this way. this isn't that much different than pool hoping in the early days of bitcoin mining, until pools adjusted their reward algorithms, which isn't the case with existing burst pools, at least not yet.
|
|
|
|
Irontiga
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
March 06, 2015, 04:44:36 AM |
|
but I have been thinking about this:
let's say I have 200TB, atm the calculator says that should be 32386 burst or /7350 = 4.4 blocks. if I see that since 00:00:00.000 I have found 5 blocks, and let's say we have another 12 hours to go in that day, then it's very likely I shouldn't find any more blocks for the remainder of the 12 hours. so I switch to mine against a pool and rackup another 10000-15000 burst. then when midnight hits, I switch to solo. if next day I have fast luck, I repeat, if not, I keep mining solo until next day and check if I should repeat. and so on. this is kind of like pool hoping to increase your reward? this is a crude description and you can substitute any numbers with whatever you want to test this scenario... does this make sense, is it possible to "game" burst mining like this for a bit? pool operators should change their reward algos to prevent this?
This doesn't make sense because if you're very lucky for some time solo, then switch to a pool... your deadlines still will be statistically worse for some time. The pool doesn't know you have 200 TB, the pool sees only your submitted deadlines and these will be quite bad in your scenario. not really. pools will pay for deadlines below their threshold. my example applies to specific day, the lucky ones. I think it's possible to increase mining returns this way. this isn't that much different than pool hoping in the early days of bitcoin mining, until pools adjusted their reward algorithms, which isn't the case with existing burst pools, at least not yet. As much as u would get from the pool the last wee while, you'd have a chance of finding a block. Pool hopping is 100% pointless. The network doesn't care about 24hour periods
|
|
|
|
Blago
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
March 06, 2015, 07:35:21 AM |
|
So I have a theory as to part of what is causing the fluctuations in the network (the saw tooth pattern). After a fast block, Blagos miner doesn't error out on the block and proceed to the next one. This means all machines currently mining on the last block will continue mining on a block that is already past. They do this till they finish scanning the disks as they normally would. Meanwhile the fresh block on the network will not be mined by the full network as everyone running Blagos miner is still working on the last block until they finish as if a new block hadn't appeared.
When multiple fast blocks happen, the next block will not receive the full mine, let alone the one after that, which can cause the network hash power to stagger downhill.
Obviously there are still people not using Blagos miner and may be using Pocminer, Urays, or Rurus where this does not seem to be a issue, so it only effects about 30% of the network. It also seems as though this does not happen if you mine on the Dev pool, but I've witnessed this behavior across multiple machines solo and on other pools.
It's not hard to see this. Look through the log of your miners, when you see a 'fast block or corrupt' or a incomplete mine (% isn't 100 before the next block), you can look at the difference in time between when that event happened and when the miner continues on to the next block and it's the difference of time that was remaining before it completed the mine of the 'fast block'.
Now, I'm not exactly sure why this hasn't been fixed yet. I brought this up multiple times including in a PM with Blago and this definitely looks like something that could easily be fixed and it's not hard to reproduce, so I assume it's to give fast machines a leg up on slow ones.
If someone wants to hack a fix, a program that monitors either Blagos miner or the log file and simply restarts the miner when either 'fast block' appears in the log or when the wallet pops a new block that would effectively fix this.
1. Took the data from my log file. 2. I have received data from blokchain. 3. Subtract the time between the change of the blocks in the miner and blokchain 97 blocks: block 73036, 03.03.2015 18:59:29 GMT <--> block 73154 04.03.2015 01:40:51 GMT) "UpdateInterval": 2000, so permissible deviation +-2 seconds X - #block Y - time, seconds RED - blockchain have new block, miner have old block GREEN - miner have new block, but blockchain have old block (fork) ![](https://ip.bitcointalk.org/?u=https%3A%2F%2Fi.imgur.com%2FQAWAvCa.png&t=665&c=a9LmiXsWnoSwFQ) ![](https://ip.bitcointalk.org/?u=https%3A%2F%2Fi.imgur.com%2FqgSrVA9.png&t=665&c=S13IrcUBQ9SbMg)
|
Relax, I’m russian!... BURST-B2LU-SGCZ-NYVS-HZEPK
|
|
|
bobafett
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
March 06, 2015, 08:24:53 AM |
|
Update and News: www.burstcoin.deNew Feature German Plotting Guide now available. ------ Help me to keep up the page and enhance the information for the german community. 220k to go in the Crowfund. Please do some donations!!!!
|
|
|
|
bensam1231
Legendary
Offline
Activity: 1750
Merit: 1024
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
March 06, 2015, 10:42:11 AM |
|
So I have a theory as to part of what is causing the fluctuations in the network (the saw tooth pattern). After a fast block, Blagos miner doesn't error out on the block and proceed to the next one. This means all machines currently mining on the last block will continue mining on a block that is already past. They do this till they finish scanning the disks as they normally would. Meanwhile the fresh block on the network will not be mined by the full network as everyone running Blagos miner is still working on the last block until they finish as if a new block hadn't appeared.
When multiple fast blocks happen, the next block will not receive the full mine, let alone the one after that, which can cause the network hash power to stagger downhill.
Obviously there are still people not using Blagos miner and may be using Pocminer, Urays, or Rurus where this does not seem to be a issue, so it only effects about 30% of the network. It also seems as though this does not happen if you mine on the Dev pool, but I've witnessed this behavior across multiple machines solo and on other pools.
It's not hard to see this. Look through the log of your miners, when you see a 'fast block or corrupt' or a incomplete mine (% isn't 100 before the next block), you can look at the difference in time between when that event happened and when the miner continues on to the next block and it's the difference of time that was remaining before it completed the mine of the 'fast block'.
Now, I'm not exactly sure why this hasn't been fixed yet. I brought this up multiple times including in a PM with Blago and this definitely looks like something that could easily be fixed and it's not hard to reproduce, so I assume it's to give fast machines a leg up on slow ones.
If someone wants to hack a fix, a program that monitors either Blagos miner or the log file and simply restarts the miner when either 'fast block' appears in the log or when the wallet pops a new block that would effectively fix this.
1. Took the data from my log file. 2. I have received data from blokchain. 3. Subtract the time between the change of the blocks in the miner and blokchain 97 blocks: block 73036, 03.03.2015 18:59:29 GMT <--> block 73154 04.03.2015 01:40:51 GMT) "UpdateInterval": 2000, so permissible deviation +-2 seconds X - #block Y - time, seconds RED - blockchain have new block, miner have old block GREEN - miner have new block, but blockchain have old block (fork) Maybe you have a really awesome pool or something like dev pool? It may look like the issue isn't happening, but it definitely is for me and I've posted photos of the stalls already. Try mining solo or on Urays and you should run into it, I know it happens under those conditions. Something also worth noting, the closer you are to 0 mine time, the less prominent this issue will be as it has less of a impact as the time spent in limbo between the fast block and the new block will be closer to 0. It's really easy to see this happening if you have like a 2 minute mine time. It will freeze on the fastblock, but continue to mine it for the full duration even though the wallet will show the blocks done.
|
I buy private Nvidia miners. Send information and/or inquiries to my PM box.
|
|
|
mczarnek
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
March 06, 2015, 02:42:26 PM |
|
Price is in a stalemate because there are no announcements for weeks now. It would be nice to hear how development is going... The Crowdfund was successful, so some further information is appropriate. "ATDevelopmentFund The purpose of this case is to provide funding to help extend the technical work and general product development and marketing for AT inside the burst platform." Making it a good time to buy... better to buy when nobody else is buying and sell when they are buying ![Smiley](https://bitcointalk.org/Smileys/default/smiley.gif)
|
|
|
|
|