I think there can be a major problem if Deepbit agrees to work with another pool or gets 51%
It will be possible for them to set the transaction fees to 1-5BTC/kb and reject any block that was not created by them.
I think this is a lot worse than the 51% double spending problem Satoshi described
And the miners might not leave - they will be making extra money from those fees.
This is sort of the reason monopolies are illegal in most countries. when a large group agrees on a price - that group can get almost twice the profit - at the expense of the users.
Even if tycho won't do something like this, somebody needs to hack just ~2 servers to achieve this goal.
Bitcoin will become regulated...

Edit:
just to be clear. i dont think the pool managers are the problem. the miners are.

its a problem of notation
when you write sqrt(1) it is accepted that you mean the positive root
but there is a negative root as well (-1)*(-1)=1 -> sqrt(1)=+/-1
similarly  sqrt(-1)=-/+i

i have a core i5 750 and it also happens to me once in a while

how about this:

sig1 OP_0 sig1 sig2 SEPARATOR 2 pub1 pub2 2 checkmultisig pub1

hash <hash(pub1)> compare checksig <hash(x)> OP_CHECKHASHVERIFY OP_DROP

x is all the mess from the separator up to the hash(x)
the pub will be twice as long but at a const. length (the complex script can be as long is it needs)
an old client will always verify at least the first sig

psychological reasons.
he's the "leader" of a big chunk of the bitcoin network
people will assume that he knows what he's doing and that after having 40% of the hashing power the full support for bip16 will be soon enabled.

thats a deadlock
the others are waiting for you

how do you buy from china? ebay?
a specific seller/store?

well if you say so....
Thats already 40% of the hashing power.
What exactly are you waiting for before you enable the support in your pool?

supports here is "technically supports" as in "able to accept and process such transactions". not support like in politics

Edit: the concern was that most people wont update, because even now more than half of the clients are below version 0.5

this was not intended to be a vote...
people started interpreting it as a vote when luke came up with BIP17

I think i had enough of this. Let gavin and luke fight to the death. the winner writes the P2SH implementation

The regular pools should have at least made a poll on their site. so its not the manager who decides, but the miners of that pool

No tea shops in CZ?
there should be...
maybe you should check some UK websites as well.
Good luck

is there already an implementation of both BIPs?

http://www.palaisdesthes.co.il/
i think their HQ is in france or something: http://us.palaisdesthes.com/en_us/

make it pink!

Earl grey rulz
also rooibos

AFAIK bitcoin still does exactly that.
puts the signatures together and executes the result
both signatures contain both data and code
my objection is not to push data, but that this data is being executed at a later stage

https://en.bitcoin.it/wiki/Transaction
also look at the examples there

bummer

you dont need computing power, you just send a competing transaction, and wichever gets first into a block will be accepted

BTW, why is the address of BIP17 longer than BIP16?

look at how Gavin himself wrote it in his post:
https://bitcointalk.org/index.php?topic=60433.0

OP_HASH160 <hash> OP_EQUAL
OP_0 <signature> OP_PUSHDATA(2 <pubkey1> <pubkey2> 2 OP_CHECKMULTISIG)

the code is "OP_PUSHDATA"
"2 <pubkey1> <pubkey2> 2 OP_CHECKMULTISIG" is data
then, this previously pushed data gets executed

while in BIP17:
<hash> OP_CODEHASHVERIFY OP_POP
OP_0 <signature> OP_CODESEPARATOR 2 <pubkey1> <pubkey2> 2 OP_CHECKMULTISIG
no tricks are used...

my problem with the idea of executing data is that it is the basis of a lot of hacks in other software