While I think the tone of this post is a bit...aggressive, I do agree with the point.  The lack of any GUI interface is a bit troubling, especially for people who are mining NMC with merged mining.  For linux users its pretty easy, most Linux users are used to running things via a command line.  But on windows, its a complete mess trying to do anything with it.

Sounds like you're having a routing issue, 100%.  On my end (and I'm halfway across the country, so it's not a local test), I've yet to have errors and the website has been loading instantly.  Nothing I can really do about that.  Could be local on your end, could be your ISP, or could be a random router between you and the server acting up.

All I can guess is some kind of routing issue.  I'm connecting from California, and I've been testing out poclbm for NMC stale rates.  The most my logs show are a single comm error that sometimes shows up a second before a longpoll, never has to do more than 1 retry.

Did one restart just to see if anything changes for the few people having issues.

Pool running fine all night based on my miners logs.

Hall of Fame now has the 26th row for your personal rank if you do not rank in the top 25.

Sorry about that (~3 minutes of downtime).  When moving over to the new PoolServerJ, there is one modification I forgot to copy, some anti-botnet code.  I was monitoring the pool and noticed a significant increase in long polls over the last few days since the MM switched, but the speed increase was nowhere near proportional.

So the pool is back up, and stales should see a slight reduction now that we don't have ~40% of the longpoll connections being taken up by botnet miners (~15-20 GH/s and nearly half the LP connections, god I hate CPU botnets).

Yes, soon the Hall of Fame will include a 26th row for your rank & count on Speed/Blocks Found/Shares Submitted, assuming you're not in the Top 25.  The blocks found counter has been reset quite a few times though, so just because it shows 0 doesn't mean you've never found one.

Fixed, thanks for reporting it.

Just for reference, I tested the merged mining using guiminer (default poclbm kernel).  guiminer is at 99.1-99.2% accepted rate on NMC side (99.7-99.8% on BTC).  I'll be adding back the ability to reset your stats on workers this afternoon, and you can reset them independently by NMC/BTC side, which can help users test out which miner gives them the best results for either side.

If you are seeing NMC reject rates in excess of 5% (over the course of 1-2k+ shares), something is absolutely wrong -on your end-.


UPDATE:  You can reset stats on workers again (separately between NMC and BTC).

I refuse to implement paid services on top of the fee I already charge.  It just feels completely wrong.

However, it did give me an idea on how to roll out idle miner emails again.  I'll be adding idle miners again soon, but it will be restricted to miners that have earned a certain amount of BTC since joining the pool.  I'll slowly slide that required miner earnings down as I am able to gauge how much "junk mail" is being sent for idles to people that are simply turning off their systems.  In addition to that, the idle miner warnings will be disable themselves after an email is sent, and not turn back on until the user logs in and turns the warnings back on.

Those two requirements should allow me to keep using Google Apps for email (fewer bounced emails/never received emails), while bringing back that idle function.

PPLNS is a fine system.  Over the long run a user gets the same rewards under PPLNS as they would under proportional, while having a system that can't be hopped.  Intermittent miners will receive a higher than normal variance, but it can swing either way (massively positive if they stop just as the pool hits a string of luck, massively less if they stop and the pool never finds a block before they expire).

I'm currently looking for a solution to the email issues.  The problem is we have so many users and many of them go on/off unrelated to pool problems.  I moved to GMail for handling our email because it reduced the number of emails that never made it to people's inboxes, but idle miner emails were often knocking us above the 500 emails/day limit and stopping other more important emails from getting through.  We have such a high volume of emails that its very hard to avoid getting flagged as a spammer, and in some cases the emails were dropped entirely rather than just sent to spam folders.

The new PSJ performs better than ever, even with the added overhead of merged mining.  It's an impressive update, and this is the reason I took so long to put in merged mining. I knew such a major change was coming ahead of time, so I didn't want to go through the migration to merged mining until this change was released.

New guides are starting to show up, How to Connect now has a few links and a guide on GUI Miner.  Writing guides for Diablo & CGMiner now, then I will make a guide for setting up a namecoin wallet.

I'm working on a few guides today to help people out with the Namecoin side of things, plus a better How to Connect page in general.

Thanks, forgot about the sidebar links (I always accessed them through the My Account page buttons).  Took the opportunity to remove the Change Email/Wallet links on the sidebar if you have them locked as well.

BTC Guild now has complete merged mining support.

You can now set your namecoin wallet, lock your namecoin wallet, set your namecoin auto payout threshold, and request namecoin payouts.

NMC wallet address entry, payouts, and API data will likely be up in about 12 hours.  I've been doing some database cleanup and monitoring for the last few hours before I go to bed, making sure its not going to crash 5 minutes after I shut my eyes.

Bitcoin 0.5 is the one BTC Guild is using.  4diff is completely pointless with the new PSJ (4diff was almost entirely used for massive getwork optimization, which PSJ doesn't even use anymore).

For namecoind, we used ArtForz's branch: https://github.com/ArtForz/namecoin

The only patches applied were the changes to the debug.log logging on rpc.cpp / bitcoinrpc.cpp to stop it from being too spammy.

Yeah, hash rate is down because downtime was almost 2 hours (and website down for an extra 1 hour roughly).  Wouldn't be surprised to see hash rate higher than before by the end of the weekend thanks to the addition of merged mining.