When was transaction malleability first identified as a potential problem?

[pythonscript]

In the Bitcoin Foundation's statement on transaction malleability it says that

Transaction malleability has been known about since 2011.

I see that the wiki article was created in January 2013 (i.e. well before the current situation with Mt. Gox, Bitstamp, etc.), but a search of bitcointalk.org didn't turn up any posts about the subject before 2013. Were there mailing list/forum posts, commit messages, etc. that stated that developers of custom wallets shouldn't use txid's as unique references to transactions before those transactions are confirmed in the blockchain (and should use the reference implementation instead)?

[Holliday]

Here is an old thread which was recently bumped.

https://bitcointalk.org/index.php?topic=8392.0

[pythonscript]

Here is an old thread that I saw was recently bumped.

https://bitcointalk.org/index.php?topic=8392.0

Thanks for pointing that out. Not sure why that didn't come up in my search of the forum, but it's a very enlightening thread.