pywallet hasn't been updated since 0.3.23, and I wouldn't trust it not to mangle or misread the new wallet database introduced in 0.4.0.
Note, also, there are some funky things that an upgrade does to pool keys (namely creating new ones). It's worth reading more in the 0.5.0 testing thread to see what the pool key rewriting does on an upgrade. Glancing at some comments, it seems that if you had a never-used private key/address, it may have been discarded during the upgrade.
If you have the private key for the bitcoin address in any form, it is still redeemable. If you backed up your wallet six months ago or such and can find the matching key pair, you can restore that (or export/import the pool key) into a fresh 0.3.24 installation (perhaps in a VirtualBox VM), and rescan for transactions and send the money.
Thanks very much for the reply, unfortunately the key is gone (I couldn't find it in any backups) and the Bitcoins with it.
There is something else very strange however, I just found out. According to blockexplorer:
http://blockexplorer.com/address/13RVBjpo3xLeDBkB2NM64N8sWK4fariZUuThe 7.17 bitcoins which I sent to this address on December 1st, were SPENT yesterday, obviously not by me. Also, I never recall having small transactions like the ones in this addresses' transaction list above. It also looks very much like a small-donation address (receiving lots of 0.02 BTC transactions) which does not look like me either, I never had a donation address.
Could anyone clarify what the output of "getaddressesbyaccount" produces? Are these addresses usable as receive addresses or not? If not what on earth are they? I could not make sense out of the Bitcoin code, as I'm far from being a cpp expert.
How is it possible that someone else has the private key for 13RVBjpo3xLeDBkB2NM64N8sWK4fariZUu and I have this address appear in my list of addresses? I suppose I am missing something really important here, so could someone help?
I consider the bitcoins gone, but I want to know what is going on, in order to avoid such mistakes in the future.