Hacker News Reader

Comment by arethuza | Hacker News Reader

I wouldn't have expected something as slow and clunky as an ATM to actually enlist in a distributed transaction - so it hardly surprises me that they might not.

There is a moderarely interesting article linked to in one of the comments that sort of sums my expectations up "Accountants Don't Use Erasers":

http://blogs.msdn.com/b/pathelland/archive/2007/06/14/accoun...

View on HN

"Moderately interesting"? This article is a gold mine on the subject!

I'd like to read his take on idempotency in logging system, as this has always been among the harder problems to solve when designing IT architectures for me.

Edit: Here are two more papers from the author, Pat Helland.

* Building on Quicksand - more about eventual consistency - http://www-db.cs.wisc.edu/cidr/cidr2009/Paper_133.pdf

* Life beyond Distributed Transactions: an Apostate’s Opinion - http://www.cidrdb.org/cidr2007/papers/cidr07p15.pdf

Most modern ATMs are actually PCs running Windows. Not sure if that qualifies as "slow and clunky".

I was thinking about the actual mechanism that counts and ejects the notes - which is pretty slow by computing standards!

I heard an urban myth the other day that the cash-counting noises are fake and are computer generated to cover the time it takes the ATM to communicate with the mainframe.

I spoke to our ATM dev and test team here, and they told me this is not true.

Reply on news.ycombinator.com