But some of us already have BTC that we mined.  You buy these things in BTC so it makes sense to consider ROI in BTC.  If you guys are talking about buying BTC with dollars and then trying to ROI in dollars that is not the same conversation that other people are having here.

That garage S9 hit another one for holyscott and the pool! That's a lucky machine right there.

If you want to get shipping cost I can do it.

https://bitcointalk.org/index.php?topic=1533501.0

On the occasion when my transaction did not confirm for two weeks it had disappeared at one point from both blockchain.info and blocktrail.com only to reappear a couple days later.  I thought it had dropped off and I was going to give it a couple more days and then resend it but the transaction was apparently rebroadcast by someone because it reappeared on both of those block explorers and then was picked up in a block soon after.

But you're right Philip, we don't really know if that is the issue without the txid.  But based on what bitmain told him I am guessing that is the issue.

I think all you can do is wait for the transaction to get picked up in a block.  This is why I always include an over-size fee when buying a miner from bitmain, to make sure it gets confirmed quickly.  I have waited nearly two weeks in the past for a transaction to get picked up that had too small of a fee on it.  In that case I actually thought it wasn't ever going to happen but it eventually did.

bitmainwarranty.com

Despite the name they do non-warranty repairs for a fee.  I had them fix an S7 board that wouldn't hash due to a temp sensor problem.  Cost $100 and had it back in about a week including shipping both ways.

I have only heard good things about Toomim hosting but that's five star service right there.

That's a pretty fast turn around for bitmain.  Did you have to send it to China?

It is a bit disturbing how it failed in a "burn down the house" sort of way, but hardware is going to fail.  All of it.  No matter how well made too, it will eventually fail.  The real warning here I think is the inability to provide the warranty and the cavalier manner in which they treated the whole warranty claim.  That's my 2 bits anyway.

EDIT: I understand also the idea that they don't want to warranty burnt boards because they can't be sure just what people are doing in terms of what PSU's they use, how many connectors, and so on.  But in this case I think there is a strong enough case for the failure being the miner and not the PSU or the end user.

Nobody?

If you start talking about token ring and vampire taps and shit we are going to put you out to pasture.

Oh that was your script.  I didn't notice that the link in this thread and the other post were both the same person.

Yeah I figured it was a local copy, I was just commenting about how I would have tackled it.  The older antminers had a cron tab where you could put commands and schedule them.  I used to schedule cgminer restarts in there.  That could have been used to accomplish this too but alas, not on a new antminer.  I suppose doing the copy from the PC might wind up being better also if the local versions of the cgminer.conf were wiped out by a reboot.  I know most of the file system changes get lost when you reboot the miners, can't remember which ones exactly.  I thought I saw someone post that there was at least one dir where changes persisted on reboots but not sure which.

Ahh I was actually going to recommend something very like that but someone else did all the work.  plink comes with putty and I have used it for automated ssh tasks in order to connect to a linux box, and execute commands based on a windows bat file.  I am not familiar with pscp which that guy uses but I would just have created two cgminer.conf files on the miner, like cgminer.conf.lo and cgminer.conf.hi and use plink to do some sort of copy and rename on the conf files on the miner rather than copy them onto the miner each time.

I wonder why the default freq on these is only 550.  Same number of chips as previous batches, so no doubt the same board.  Why lower the freq and thereby the hashrate?  Are these QA failures on some level?

Added my second S9 to the pool.  On the down side I have an S7 with nowhere to plug in.  Just sitting on the floor now...

notlist3d did some testing on a bunch of pools with a B1.

http://s100.photobucket.com/user/notlist3d/library/S9?sort=3&page=1

I went back and looked and antpool hashrate was slightly better than some of the others but not drastically and certainly not 20%.

Here's the specs on this one:

https://shop.bitmain.com/productDetail.htm?pid=00020151212042549033J27u55CO0610

Not including the PSU.

Not really looking to ship outside the lower 48 but we can talk if you want to pay additional for shipping elsewhere.

I pay shipping but buyer pays for escrow.

It OC's a little and I have been running it this way.

Reset it using the button.  You'll have to reconfigure everything but it looks like that would be the next logical step here since nothing else jumps out as obviously wrong.

B4 available again.  I do eventually want 1 more to replace my last S7 but I just can't conceive of paying more than 2k for one with the halving this close.

Well I just got an arrival scan for Anchorage although it wasn't before 1:30 it says 1:37.

So these clearance agency items are definitely appearing after the fact.  It has me wondering if they just didn't get around to entering it in the computer until later or if they just gathered info on the packages and then flip the coin to decide if there is a fee after the package has been carried on it's way.  So far mine does not say there is a fee but I still don't have the post about having been released from the clearance agency despite the fact that it is now in Anchorage.