Chris, > Yes, and then you backpedalled furiously when I showed that > proper transactions prevent this.
You're a fool, out for a fight. /You/ might know exactly how to handle a database to make sure its /transactions/ will not leave the database in a corrupt state, but as I mentioned a few posts back: [quote] I think that a database is /definitily/ overcomplicating stuff, especially when looking at the OP who's supposed to write and maintain it. [/quote] Mind the part after the comma. Also, you happily ignored the part in my second post where I mentioned that corruption can be of a different kind - one that you have absolutily /no/ control over: [quote] An unreadable file is often described as being corrupt [/quote] > I lost track of what you were actually trying to say. Well, all that I wanted to say - to the OP - is in my first reply. You could always try to re-read it. The other posts where just responses to your "that can't be!" challenges to me. > So either justify this position, showing that a database can > become corrupted the same way a flat file can, :-) In the /same/ way ? Thats rather multi-interpretable, don't you think ? Are you trying to set me up ? My point was that the /effect/ of a same corruption is different for both: A flat file is /much/ easier to rescue. But, do you remember what the OP said ? [quote] want to download these as a 'background task'. ... you can CTRL-C out, [/quote] Why now do I think that, when such a backgroud process is forgotten and the 'puter switched off, the file, database or otherwise, could easily get damaged ? And that a ctrl-c at the wrong moment could cause the same if the cleanup (of the database, which could easily still be, in its own thread, busy housekeeping) isn't done correctly. Also, have you /never/ encountered a corrupted file, database or otherwise ? You must not have done much with 'puters at all ... Though I think I'm going to end our conversation here, as you're way more agressive than the subject calls for. Goodbye chris. Have a good life. Regards, Rudy Wieser -- https://mail.python.org/mailman/listinfo/python-list
