Could wild swings in difficulty happen? (see namecoin)

[grod]

Looking at what happened to namecoin got me thinking about long term future of bitcoin.

The premise of bitcoin is it's worth mining even when there's nearly no profit in it because the win comes from hoarding most of your product and selling just enough to cover expenses.  Miners bank on rising difficulty and bitcoin deflation, in other words.

The exact opposite happened to namecoin.  As it became vastly more profitable to mine namecoins than buy them the amount of hashing power in that network grew by 400% overnight.  Difficulty went from 15k to 55k in about 2 days.  At that point it was no longer advantageous to mine namecoins, but the added risk remained so everyone shifted back to BTC.  The hashing power fell right back to what it was a week ago, and at the moment namecoin network is slogging along with about a 1 block/hour rate.  Difficulty should adjust down in about 3-4 more weeks at this rate.

Could that happen to bitcoins?  At some point it would be advantageous to turn off your rig for a month or two and let the patsies mine at $1/day.  Then the difficulty plunges, everyone dogpiles on and makes two weeks worth of coins in a day or two.  Turns off rigs, lets patsies mine for a month+... Repeat.

Could be a fun alternative to an 8 million difficulty and $100 btc price.

[JohnDoe]

A simple fix for this problem has been suggested a few times: retarget should occur either after 2016 blocks or after 14 days, whichever happens first.

[just_someguy]

A simple fix for this problem has been suggested a few times: retarget should occur either after 2016 blocks or after 14 days, whichever happens first.

Without having all of the network synced to the same clock doesn't this have the potential to fork the block chain?
If my 14 days is just slightly different then your 14 days and a block comes out in that gap of time it may be valid to me but not to you.

[JohnDoe]

Without having all of the network synced to the same clock doesn't this have the potential to fork the block chain?
If my 14 days is just slightly different then your 14 days and a block comes out in that gap of time it may be valid to me but not to you.

I'm not knowledgeable on the technicalities but I think timestamps take care of that problem. Even if you are out of sync when a block with new difficulty arrives you would see something like "time : 1412542299" and go "oh ok, that block was created at that time so it is valid". Could be wrong though, we'll have to wait for confirmation from someone who knows for certain how it works.