2 reasons for me:
1. compared to all other pools I tried, that's the one earning me the most, even considering the 3% fee
2. there's no correct PPLNS alternative IMO. There's an okay PPLNS pool, but I find its N too small: by the time I used to mine there, I could only mine during the day, and the small N meant that even though it was PPLNS I still wasn't paid if a block was found during the night. Plus it went down a couple of times. So, too much variability (paradoxically, much more variability than with deepbit).
The only pool I'd consider trying now would be a stable enough (uptime) PPLNS pool with N=at least 24h worth of shares.

Additionnally:
Slush has higher fee if you factor invalid blocks, and I really hate that score-based system that was terrible when I was mining with a very unpowerful GPU.
BTC Guild has a history of suspiciously terrible luck. Okay, now that's outdated since they enforce PPS. My new reason for not going there is thus: I just don't like PPS, pool owner takes away the "risk" of bad (and good!) luck at the cost of a huge fee: I prefer keeping the luck for a much lower fee.

You when you don't have enough time I wish I could still play...

Hm, let's try to explain that more simply:
- if CH explains a great-looking design, then it's worth checking the source just to check that he actually implemented this great design and not something else
- if CH explains a crappy-looking design, then we can safely assume that this design is _at best_ what we'll find when reading the source (since he's not here to undermine his own currency, we can assume he exposes his design to make it look as good as possible). Or to put it in other words: it's not worth checking if it's not as good as described, if it's not good enough in the description in the first place.

Plus what LoupGaroux and DeathAndTaxes said.

What you want to check in the source is whether or not you're told the truth about the "greatness" of a design. So the idea is, if the design sounds incoherent in the first place, there's much less of an incentive to check the source...

okay, it's centralized on 10 accounts. What difference does it make anyway? I mean, obviously a same account is controlled by a same person (no matter how many nodes it has), and my concern is about people centralization, not machine centralization.

1: No: everyone had the same power
2: Somewhat, but reversibly: people give hashing power to pools, but they're not forced to

Indeed, saying people "go read the source" isn't nearly as efficient as explaining in a convincing and coherent manner how things work. Speaking of coherence (or lack of thereof), cf posts by DeathAndTaxes and Bobnova.

Try again, it's centralized on 10 (or so) trusted nodes. Doesn't sound like 5% economic centralization.

Define "many"...

It's part of surgery training, duh

Are you kidding? The network shutdown was announced barely a few hours before it occurred. It was so fast that a big pool owner didn't even notice it until several hours after the shutdown!
And when the shutdown was announced, it was just "to fix the 50%+1 attack". Only MUCH after came the CPU-biased mining and the tax. And only after the SC2 release we were made aware of the shitloads of post-pre-mined coins.

So no, we didn't have plenty of time to get out. And no we had no idea of the mess it would become before it shut down. Get some memory.

Oh, wait, I get it. The network isn't 50%+1 attackable simple because... it already is 50%+1 owned

Yeah, I noticed that too, so I also checked in my account transactions: no credit for this block here either.
Maybe the code that notes whether a block is confirmed or not is bugged? I've never seen any block marked as unconfirmed yet.

One of the reasons why I stopped playing DRM-ed games. They abuse the DRMs to enforce censorship. So they can eat dust and drop dead, as far as I'm concerned.

Well, the higher N the weirder it makes that I missed that block

I just noticed that I didn't get any reward for Block 23,377. Was there an issue with this block? It shows as "confirmed" and I'm fairly sure I had valid shares around it (we're still on PPLNS with N=20k, right?) since I got paid for the blocks before and after at my usual rate.

Thanks for the direct links, I just hate how AMD makes us download (and install!) a downloader program first.

Edit: 100% CPU bug still present with my dual GPU setup.

Well, in this case okay, but what you describe here isn't a "comparable benefit" between GPU and CPU (top of the range GPU=somewhere around 800MH, not 100).

Nice DES cracking board, btw

Doesn't work: CPU remains efficient and cheap, only we can expect to have maybe twice the hashing power because people will mine both with CPU and GPU. That is, supposing they're willing to drop BTC mining in favor of this new chain where GPU doesn't have an advantage over CPU.

Actually, the TaxCoin fanboy has a point there. This transcript just reflects the nursery school maturity level of some discussions on the SC IRC, not a real attack intent.

Personally, I find this much less discouraging than 10000 blocks being distributed withing the first 24 hours, or than 13M premined coins. Even though botnets are clearly not legitimate, at least they're playing "by the rules" (kind of) instead of profiting from a very very misadjusted difficulty or from being the network admin (or his pals): unlike the latest, they do have a (fairly important) cost for the botnet. Plus they are actually helping in protecting the network against 50%+1 attacks, until you piss them off at least.
And no, I don't own a botnet (I wish ).

Lol, that's another possibility indeed