>>>>>>>>Medical science is not an exact science -- something people often forget (or don't want to know. :)
>>>>>>
>>>>>>This also makes it difficult for the public to assess doctors and health advice, when this inexact science can mean that "the best lack all conviction, while the worst >>No worries -- computers would become so powerful that it would make up the inefficiency of crappy code... wouldn't they? 😜😜😜
>>
>>Nope. We have fast computers, yet we still get data corruption because there are no transactions. Speed is not the only solution.
>
>In terms of data corruption, Transactions are just a tool to avoid certain types of data corruption but isn't the silver magical bullet either. VFP transactions only play a minor role in preventing data corruptions as it does not conform to the ACID properties. SQL server transactions are much better however preventing data corruption goes way beyond just doing transactions.
>
>The real reason we still get data corruption is that nothing is infallible and many programmers do not have the knowledge and insights to minimize the risks. Its a very complicated topic, especially if you deal with financial systems that need to adhere to the utmost highest standards.

There is no need to go to 'perfect'. Transactions prevent the stupid errors where 1 or 2 out of 3 records are saved. That is not anything more than having reasonable standards which are woefully lacking as is your inability to reason.